SMELTING OPTIMIZATION METHOD AND DEVICE OF COPPER-CONTAINING CONCENTRATE

Abstract

Provided are smelting optimization methods and devices of copper-containing concentrate. The smelting optimization method includes: obtaining a water content and physical parameter of copper-containing concentrate powder; in response to the water content meeting a first preset condition, mixing the copper-containing concentrate powder with oxygen-rich gas to obtain a mixture, spraying the mixture into a smelting furnace by a nozzle; determining a first production parameter for producing matte based on the physical parameter, an oxygen content of the oxygen-rich gas and a gas temperature of the oxygen-rich gas; determining a second production parameter for producing matte by processing the physical parameter, the oxygen content of the oxygen-rich gas and the gas temperature of the oxygen-rich gas using a prediction model, wherein the prediction model is a machine learning model; and determining a target production parameter based on the first production parameter and the second production parameter.

Claims

1. A smelting optimization method of copper-containing concentrate, which is executed by at least one processing device, comprising: obtaining a water content and physical parameter of copper-containing concentrate powder; in response to the water content meeting a first preset condition, mixing the copper-containing concentrate powder with oxygen-rich gas to obtain a mixture, spraying the mixture into a smelting furnace by a nozzle; determining a first production parameter for producing matte based on the physical parameter, an oxygen content of the oxygen-rich gas and a gas temperature of the oxygen-rich gas; determining a second production parameter for producing matte by processing the physical parameter, the oxygen content of the oxygen-rich gas and the gas temperature of the oxygen-rich gas using a prediction model, wherein the prediction model is a machine learning model; and determining a target production parameter based on the first production parameter and the second production parameter.

2. The smelting optimization method according to claim 1, wherein the determining a first production parameter for producing matte based on the physical parameter, the oxygen content of the oxygen-rich gas and the gas temperature of the oxygen-rich gas comprises: determining a target vector based on the physical parameter, the oxygen content of the oxygen-rich gas and the gas temperature of the oxygen-rich gas; retrieving in a vector database based on the target vector to determine a matching vector and a production parameter corresponding to the matching vector; and determining the production parameter corresponding to the matching vector as the first production parameter.

3. The smelting optimization method according to claim 1, further comprising: obtaining a fracture surface image of hot matte based on an image recognition device; determining a measured matte grade through a grade determination model based on the fracture surface image of hot matte, wherein the grade determination model is a machine learning model; and updating the target production parameter based on the measured matte grade.

4. The smelting optimization method according to claim 3, wherein the grade determination model comprises an image extraction layer, a correction layer and a judgment layer; an input of the image extraction layer includes the fracture surface image of the hot matte, and an output includes a fracture surface feature; an input of the correction layer includes the fracture surface feature, a difference between a matte tapping time and an image acquisition time, and an output includes a corrected fracture surface feature; and an input of the judgment layer includes the corrected fracture surface feature, and an output includes the measured matte grade.

5. The smelting optimization method according to claim 3, further comprising: judging whether the measured matte grade, a measured matte temperature and a measured slag iron-silicon ratio meet a second preset condition; and in response to the judgement that the measured matte grade, the measured matte temperature and the measured slag iron-silicon ratio do not meet the second preset condition, updating a frequency of obtaining the water content and a frequency of obtaining the physical parameter.

6. The smelting optimization method according to claim 5, wherein the updating the frequency of obtaining the water content and the frequency of obtaining the physical parameter comprises: updating the frequency of obtaining the water content and the frequency of obtaining the physical parameter based on a detected difference value and a number of occurrences of abnormal differences; wherein the detected difference value includes at least one of a difference between the measured matte temperature and a preset matte temperature, a difference between the measured matte grade and a preset matte grade, and a difference between the measured slag iron-silicon ratio and a preset slag iron-silicon ratio; the number of occurrences of the abnormal differences includes a number of times the difference between the measured matte temperature and the preset matte temperature does not meet the first threshold, a number of times the difference between the measured matte grade and the preset matte grade does not meet the second threshold, and a difference between the measured slag iron-silicon ratio and the preset slag iron-silicon ratio does not meet the third threshold.

7. The smelting optimization method according to claim 5, further comprising: judging whether the measured matte grade, the measured matte temperature and the measured slag iron-silicon ratio meet a third preset condition; and updating the target production parameter in response to the judgement that the measured matte grade, the measured matte temperature and the measured slag iron-silicon ratio do not meet the third preset condition.

8. The smelting optimization method according to claim 1, further comprising: determining a first update parameter based on a measured value and a preset value, wherein the measured value comprises a measured matte temperature, a measured matte grade and a measured slag iron-silicon ratio, and the preset value comprises a preset matte temperature, a preset matte grade and a preset slag iron-silicon ratio; and updating the target production parameter based on the first update parameter.

9. The smelting optimization method according to claim 8, wherein the determining a first update parameter based on the measured value and the preset value comprises: determining at least one group of candidate production parameters based on the target production parameter; determining at least one predicted value based on the at least one group of candidate production parameters, wherein the predicted value comprises a predicted matte temperature, a predicted matte grade and a predicted slag iron-silicon ratio; and determining the first update parameter based on the at least one predicted value and a symmetrical value, wherein the symmetrical value and the measured value are symmetrical based on the predicted value, and the symmetrical value includes a symmetrical matte temperature, a symmetrical matte grade and a symmetrical slag iron-silicon ratio.

10. The smelting optimization method according to claim 9, wherein the updating the target production parameter based on the first update parameter comprises: determining a second update parameter based on the first production parameter and the first update parameter; and updating the target production parameter based on the second update parameter.

11. The smelting optimization method according to claim 8, wherein the measured matte grade is determined based on a grade determination model.

12. A smelting optimization system for copper-containing concentrate, comprising: at least one storage medium comprising an instruction set for smelting optimization of copper-containing concentrate; at least one processor in communication with the at least one storage medium, wherein when executing the instruction set, the at least one processor is configured to: obtain a water content and physical parameter of copper-containing concentrate powder; in response to the water content meeting a first preset condition, mix the copper-containing concentrate powder with oxygen-rich gas to obtain a mixture, spray the mixture into a smelting furnace by a nozzle; determine a first production parameter for producing matte based on the physical parameter, an oxygen content of the oxygen-rich gas and a gas temperature of the oxygen-rich gas; determine a second production parameter for producing matte by processing the physical parameter, the oxygen content of the oxygen-rich gas and the gas temperature of the oxygen-rich gas using a prediction model, wherein the prediction model is a machine learning model; and determine a target production parameter based on the first production parameter and the second production parameter.

13. The smelting optimization system according to claim 12, wherein the determining a first production parameter for producing matte based on the physical parameter, the oxygen content of the oxygen-rich gas and the gas temperature of the oxygen-rich gas comprises: determining a target vector based on the physical parameter, the oxygen content of the oxygen-rich gas and the gas temperature of the oxygen-rich gas; retrieving in a vector database based on the target vector to determine a matching vector and a production parameter corresponding to the matching vector; and determining the production parameter corresponding to the matching vector as the first production parameter.

14. The smelting optimization system of claim 12, wherein the at least one processor is further configured to: obtain a fracture surface image of hot matte based on an image recognition device; determine a measured matte grade through a grade determination model based on the fracture surface image of hot matte, wherein the grade determination model is a machine learning model; and update the target production parameter based on the measured matte grade.

15. The smelting optimization system according to claim 14, wherein the grade determination model comprises an image extraction layer, a correction layer and a judgment layer; an input of the image extraction layer includes the fracture surface image of the hot matte, and an output includes a fracture surface feature; an input of the correction layer includes the fracture surface feature, a difference between a matte tapping time and an image acquisition time, and an output includes a corrected fracture surface feature; and an input of the judgment layer includes the corrected fracture surface feature, and an output includes the measured matte grade.

16. The smelting optimization system of claim 14, wherein the at least one processor is further configured to: judge whether the measured matte grade, a measured matte temperature and a measured slag iron-silicon ratio meet a second preset condition; and in response to the judgement that the measured matte grade, the measured matte temperature and the measured slag iron-silicon ratio do not meet the second preset condition, update a frequency of obtaining the water content and a frequency of obtaining the physical parameter.

17. The smelting optimization system according to claim 16, wherein the updating the frequency of obtaining the water content and the frequency of obtaining the physical parameter comprises: updating the frequency of obtaining the water content and the frequency of obtaining the physical parameter based on a detected difference value and a number of occurrences of abnormal differences; wherein the detected difference value includes at least one of a difference between the measured matte temperature and a preset matte temperature, a difference between the measured matte grade and a preset matte grade, and a difference between the measured slag iron-silicon ratio and a preset slag iron-silicon ratio; the number of occurrences of the abnormal differences includes a number of times the difference between the measured matte temperature and the preset matte temperature does not meet the first threshold, a number of times the difference between the measured matte grade and the preset matte grade does not meet the second threshold, and a difference between the measured slag iron-silicon ratio and the preset slag iron-silicon ratio does not meet the third threshold.

18. The smelting optimization system of claim 16, wherein the at least one processor is further configured to: judge whether the measured matte grade, the measured matte temperature and the measured slag iron-silicon ratio meet a third preset condition; and update the target production parameter in response to the judgement that the measured matte grade, the measured matte temperature and the measured slag iron-silicon ratio do not meet the third preset condition.

19. A non-transitory computer-readable storage medium storing computer instructions, and after the computer reads the computer instructions in the storage medium, the computer implements the smelting optimization method of copper-containing concentrate according to claim 1.

20. A smelting optimization device for copper-containing concentrate, comprising a water content monitoring device, a detection device, a conveying device, a smelting furnace and a processing device; wherein the water content monitoring device is used for obtaining a water content of copper-containing concentrate powder; the detection device is used for obtaining a physical parameter of the copper-containing concentrate powder; the conveying device is used for, in response to the water content meeting a first preset condition, mixing the copper-containing concentrate powder with oxygen-rich gas to obtain a mixture, spraying the mixture into a smelting furnace by a nozzle; the smelting furnace is used for smelting the mixture; and the processing device is used for: determining a first production parameter for producing matte based on the physical parameter, an oxygen content of the oxygen-rich gas and a gas temperature of the oxygen-rich gas; determining a second production parameter for producing matte by processing the physical parameter, the oxygen content of the oxygen-rich gas and the gas temperature of the oxygen-rich gas using a prediction model, wherein the prediction model is a machine learning model; and determining a target production parameter based on the first production parameter and the second production parameter.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0009] The present disclosure will be further illustrated by way of exemplary embodiments, which will be described in detail with the accompanying drawings. These embodiments are not limiting, and in these embodiments, like numerals refer to like structures, and in which:

[0010] FIG. 1 is a schematic diagram of an application scenario of a smelting optimization system for copper-containing concentrates according to some embodiments of the present disclosure;

[0011] FIG. 2 is an exemplary block diagram of a smelting optimization system of copper-containing concentrate according to some embodiments of the present disclosure;

[0012] FIG. 3 is an exemplary flow chart for determining target production parameters according to some embodiments of the present disclosure;

[0013] FIG. 4 is an exemplary flowchart of updating target production parameters according to some embodiments of the present disclosure;

[0014] FIG. 5 is an exemplary schematic diagram of a grade determination model according to some embodiments of the present disclosure;

[0015] FIG. 6 is an exemplary flowchart of updating a target production parameter based on a first update parameter according to some embodiments of the present disclosure;

[0016] FIG. 7 is an exemplary flowchart of determining a first update parameter based on a detected difference value according to some embodiments of the present disclosure.

DETAILED DESCRIPTION

[0017] In order to explain the technical scheme of the embodiment of the present disclosure more clearly, the following will briefly introduce the drawings needed in the description of the embodiment. Obviously, the attached drawings in the following description are only some examples or embodiments of the present disclosure. For ordinary technicians in this field, the present disclosure may be applied to other similar situations according to these drawings without creative work. Unless it is obvious from the linguistic context or otherwise stated, the same reference numerals in the drawings represent the same structure or operation.

[0018] It should be understood that system, device, unit and/or module as used herein is a method for distinguishing different components, elements, parts, parts or assemblies at different levels. However, if other words may achieve the same purpose, they may be replaced by other expressions.

[0019] As shown in the present disclosure and claims, the words a, an, one and/or the do not refer to the singular form, but may also include the plural form. Generally speaking, the terms including and comprising only imply the inclusion of clearly identified steps and elements, and these steps and elements do not constitute an exclusive list, and a method or device may also include other steps or elements.

[0020] Flowcharts are used in the present disclosure to explain the operations performed by the system according to the embodiments of the present disclosure. It should be understood that the preceding or following operations are not necessarily performed accurately in order. Instead, the steps may be processed in reverse order or simultaneously. At the same time, other operations may be added to these processes, or one or more steps may be removed from these processes.

[0021] FIG. 1 is a schematic diagram of an application scenario 100 of a smelting optimization system for copper-containing concentrate according to some embodiments of the present disclosure. In some embodiments, the application scenario 100 may include a water content monitoring device 110, a detection device 120, a conveying device 130, a smelting furnace 140, a network 150 and a processing device 160.

[0022] The water content monitoring device 110 may refer to a device for monitoring the water content of copper-containing concentrate powder entering the smelting furnace 140. In some embodiments, the water content monitoring device 110 may include a microwave water separator, a near infrared water separator, and the like. In some embodiments, the water content of the copper-containing concentrate powder is obtained based on the water content monitoring device 110. When the water content meets the first preset condition, the copper-containing concentrate powder is mixed with oxygen-rich gas, and the mixture is sprayed into the smelting furnace from the nozzle, so that the water content of the copper-containing concentrate powder entering the smelting furnace 140 meets the smelting requirements, and it is avoided that the water content in the copper-containing concentrate is too high, which leads to concentrate adhesion and affects the batching accuracy. For the specific content of water content, see FIG. 3 and its related description.

[0023] The detection device 120 may refer to a device for detecting the physical parameter of copper-containing concentrate powder. In some embodiments, the detection device 120 may include a wall-mounted melting thermometer, a particle size distribution detector, an ore grade detector, and the like. In some embodiments, the physical parameter of the copper-containing concentrate powder is obtained based on the detection device 120, thereby determining the target production parameter(s). For details of the physical parameter, see FIG. 3 and its related descriptions.

[0024] The conveying device 130 may refer to a device for conveying copper-containing concentrate powder and oxygen-rich gas into a smelting furnace. In some embodiments, when the water content meets the first preset condition, the conveying device 130 injects the mixture of copper-containing concentrate powder and oxygen-rich gas into the smelting furnace from the nozzle. For details about oxygen-rich gas and water content meeting the first preset condition, see FIG. 3 and its related description.

[0025] The smelting furnace 140 may refer to an apparatus for smelting the mixture. In some embodiments, the smelting furnace 140 may include an Outokumpu flash furnace, an Inco flash furnace, and the like.

[0026] The network 150 may connect various components of the system and/or connect the system with external resource parts. The network 150 enables communication among various components and with other parts outside the system, and facilitates the exchange of data and/or information. For example, the processing device 160 may obtain the physical parameter of copper-containing concentrate powder from a storage device through the network 150.

[0027] In some embodiments, the network 150 may be any one or more of a wired network or a wireless network. For example, the network 150 may include a cable network, an optical fiber network, a telecommunication network, the Internet, a local area network (LAN), a wide area network (WAN), a wireless local area network (WLAN), a metropolitan area network (MAN), a public switched telephone network (PSTN), a Bluetooth network, a ZigBee network, a near field communication (NFC), an in-device bus, an out-device bus, etc. The network connection between each part may be in one of the above ways or in many ways. In some embodiments, the network may be point-to-point, shared, centralized and other topologies or a combination of multiple topologies.

[0028] Processing device 160 may be used to perform one or more functions disclosed in one or more embodiments in the present disclosure. For example, the processing device 160 may determine a first production parameter for producing matte (also referred to as copper matte) based on the physical parameter, oxygen content of the oxygen-rich gas and gas temperature of the oxygen-rich gas. For another example, the processing device 160 may determine a second production parameter for producing matte based on the processing of the physical parameter, the oxygen content of the oxygen-rich gas and the gas temperature of the oxygen-rich gas by the prediction model. For another example, the processing device 160 may determine a target production parameter based on the first production parameter and the second production parameter.

[0029] In some embodiments, the processing device 160 may include one or more processing engines (e.g., a single-chip processing engine or a multi-chip processing engine). By way of example only, the processing device 160 may include a central processing unit (CPU), an application specific integrated circuit (ASIC), an application specific instruction processor (ASIP), a graphics processing unit (GPU), a physics processing unit (PPU), a digital signal processor (DSP), a field programmable gate array (FPGA), a programmable logic device (PLD), a controller, a microcontroller unit, and a reduced instruction set computer.

[0030] In some embodiments, the application scenario 100 of the smelting optimization system of copper-containing concentrate may also include one or more other devices, such as storage devices and/or user terminals.

[0031] The storage device may be used to store data and/or instructions related to the application scenario 100 of the smelting optimization system copper-containing concentrate. In some embodiments, the storage device may store data and/or information obtained from the processing device 160, the user terminal, and the like. For example, the storage device may store the physical parameter and production parameters of copper-containing concentrate powder.

[0032] A storage device may include one or more storage components, and each storage component may be an independent device or a part of other devices. In some embodiments, the storage device may include random access memory (RAM), read-only memory (ROM), mass storage, mobile storage, volatile read-write memory, etc. or any combination thereof. Illustratively, the mass storage may include magnetic disks, optical disks, solid-state disks, and the like. In some embodiments, the storage device may be implemented on a cloud platform.

[0033] A user terminal may refer to one or more terminal devices or software used by a user. Users may refer to individuals or groups related to smelting optimization of copper-containing concentrates. For example, smelting workers, smelting engineers. In some embodiments, the user terminal may include a mobile device, a tablet computer, a notebook computer, a desktop computer, etc. or any combination thereof. In some embodiments, the processing device 160 may interact with a user through a user terminal. The above examples are only illustrative, which are not limited.

[0034] It should be noted that the application scenario 100 of the smelting optimization system of copper-containing concentrate is provided for illustration purposes only and is not intended to limit the scope of this present disclosure. Many modifications or changes may be made to those skilled in the art according to the description in the present disclosure. For example, the application scenario 100 of the smelting optimization system of copper-containing concentrate may realize similar or different functions on other devices. However, these changes and modifications do not depart from the scope of this application.

[0035] FIG. 2 is an exemplary module diagram 200 of a smelting optimization system for copper-containing concentrate according to some embodiments of the present specification.

[0036] In some embodiments, the module diagram 200 may include an acquisition module 210, a mixing module 220, a first determination module 230, a second determination module 240 and a third determination module 250. In some embodiments, the module diagram 200 may further include an update module 260.

[0037] In some embodiments, the acquisition module 210 may be used to obtain the water content and physical parameter of copper-containing concentrate powder.

[0038] In some embodiments, the mixing module 220 may be used to mix copper-containing concentrate powder with oxygen-rich gas in response to the water content meeting the first preset condition, and then spray the mixture into the smelting furnace from the nozzle.

[0039] In some embodiments, the first determination module 230 may be used to determine a first production parameter for producing matte based on the physical parameter, oxygen content of the oxygen-rich gas and gas temperature of the oxygen-rich gas.

[0040] In some embodiments, the first determining module 230 may be further used to determine the target vector based on the physical parameter, the oxygen content of the oxygen-rich gas and the gas temperature of the oxygen-rich gas; based on the target vector, retrieve in the vector database to determine the matching vector and a production parameter corresponding to the matching vector; and determine the production parameter corresponding to the matching vector as the first production parameter.

[0041] In some embodiments, the second determination module 240 may be used to determine the second production parameters for producing matte based on the processing of the physical parameter, oxygen content of oxygen-rich gas and gas temperature of oxygen-rich gas by a prediction model, wherein the prediction model is a machine learning model.

[0042] In some embodiments, the third determining module 250 may be used to determine a target production parameter based on the first production parameter and the second production parameter.

[0043] In some embodiments, the updating module 260 may be used to obtain a fracture surface image of hot matte based on an image recognition device; determine a measured matte grade through a grade determination model based on the fracture surface image of hot matte, wherein the grade determination model is a machine learning model; update the target production parameter based on the measured matte grade.

[0044] In some embodiments, the grade determination model includes an image extraction layer, a correction layer and a judgment layer. The input of the image extraction layer includes a fracture surface image of hot matte, and the output includes a fracture surface feature. The input of the correction layer includes the fracture surface feature, a difference between a matte discharge time and an image acquisition time, and an output includes a corrected fracture surface feature. An input of the judgment layer includes the corrected fracture surface feature, and an output includes the measured matte grade.

[0045] In some embodiments, the updating module 260 may be further used to judge whether the measured matte grade, the measured matte temperature and the measured slag iron-silicon ratio meet a second preset condition; and in response to the judgement that the measured matte grade, the measured matte temperature and the measured slag iron-silicon ratio do not meet the second preset condition, updating a frequency of obtaining the water content and a frequency of obtaining the physical parameter.

[0046] In some embodiments, the updating module 260 may be further used to update the frequency of obtaining the water content and the frequency of obtaining the physical parameter based on a detected difference value and an abnormal difference occurrence time. The detected difference value includes at least one of a difference between the measured matte temperature and a preset matte temperature, a difference between the measured matte grade and a preset matte grade, and a difference between the measured slag iron-silicon ratio and a preset slag iron-silicon ratio. The number of occurrences of the abnormal differences includes at least one of a number of times the difference between the measured matte temperature and the preset matte temperature does not meet the first threshold, a number of times the difference between the measured matte grade and the preset matte grade does not meet the second threshold, and a difference between the measured slag iron-silicon ratio and the preset slag iron-silicon ratio does not meet the third threshold.

[0047] In some embodiments, the updating module 260 may be further used to judge whether the measured matte grade, the measured matte temperature and the measured slag iron-silicon ratio meet a third preset condition; and update the target production parameter in response to the judgement that the measured matte grade, the measured matte temperature and the measured slag iron-silicon ratio do not meet the third preset condition.

[0048] In some embodiments, the updating module 260 may be further used to determine a first update parameter based on a measured value and a preset value, wherein the measured value comprises a measured matte temperature, a measured matte grade and a measured slag iron-silicon ratio, and the preset value comprises a preset matte temperature, a preset matte grade and a preset slag iron-silicon ratio; and update the target production parameter based on the first update parameter.

[0049] In some embodiments, the update module 260 may be further used to determine at least one group of candidate production parameters based on the target production parameter; determine at least one predicted value based on the at least one group of candidate production parameters, wherein the predicted value comprises a predicted matte temperature, a predicted matte grade and a predicted slag iron-silicon ratio; and determine the first update parameter based on the at least one predicted value and a symmetrical value, wherein the symmetrical value and the measured value are symmetrical based on the predicted value, and the symmetrical value includes a symmetrical matte temperature, a symmetrical matte grade and a symmetrical slag iron-silicon ratio.

[0050] In some embodiments, the update module 260 may be further configured to determine a second update parameter based on the first production parameter and the first update parameter; and update the target production parameter based on the second update parameter.

[0051] In some embodiments, the measured matte grade is determined based on a grade determination model.

[0052] For more details about the acquisition module 210, the mixing module 220, the first determining module 230, the second determining module 240, the third determining module 250 and the updating module 260, please refer to FIGS. 3-6 and their related descriptions.

[0053] It should be understood that the system shown in FIG. 2 and its modules may be implemented in various ways. It should be noted that the above description of the smelting optimization system for copper-containing concentrate and its modules is only for the convenience of description, and this description cannot be limited to the scope of the cited embodiments. It may be understood that for those skilled in the field, after understanding the principle of the system, it is possible to combine various modules or form a subsystem to connect with other modules without deviating from this principle. In some embodiments, the obtaining module 210, the mixing module 220, the first determining module 230, the second determining module 240, the third determining module 250 and the updating module 260 disclosed in FIG. 2 may be different modules in a system, or one module may realize the functions of two or more modules mentioned above. For example, each module may share a storage module, and each module may also have its own storage module. Such variations are within the protection scope of the present disclosure.

[0054] FIG. 3 is an exemplary flowchart for determining target production parameters according to some embodiments of the present disclosure. As shown in FIG. 3, the process 300 includes the following steps. In some embodiments, process 300 may be performed by processing device 160.

[0055] step 310, obtaining the water content and physical parameter of copper-containing concentrate powder. In some embodiments, step 310 may be performed by the acquisition module 210.

[0056] The water content (also referred to as moisture content) may refer to an index of water contained in the copper-containing concentrate powder. In some embodiments, the water content of copper-containing concentrate powder may be expressed as water-containing ratio. For example, the water content of copper-containing concentrate powder may be 0.4%. In some embodiments, the water content of copper-containing concentrate powder may be detected by water content monitoring devices such as microwave moisture meters and near infrared moisture meters.

[0057] The physical parameter may refer to parameters that may represent the physical properties of copper-containing concentrate powder. In some embodiments, the physical properties of copper-containing concentrate powder may include ore powder temperature, particle size distribution, ore powder grade, etc. For example, the physical properties of copper-containing concentrate powder may be expressed by the physical parameters that the ore powder temperature is 40 C., the particle size distribution is D50 and the ore powder grade is 0.8%. In some embodiments, the physical parameters of the copper-containing concentrate powder may be automatically detected by the detection device, displayed via the interactive terminal and input into the processor device after confirmation.

[0058] Step 320, mixing the copper-containing concentrate powder with oxygen-rich gas to obtain a mixture, spraying the mixture into a smelting furnace by a nozzle. In some embodiments, step 320 may be performed by the mixing module 220.

[0059] The first preset condition may refer to a condition under which the copper-containing concentrate powder can be mixed with oxygen-rich gas. The first preset condition may include that the water content of copper-containing concentrate powder is less than a water content threshold. For example, the water content threshold may be 0.3%, and the first preset condition may be that the water content of copper-containing concentrate powder is less than 0.3%.

[0060] In some embodiments, when the copper-containing concentrate powder enters the reaction shaft (furnace) during the flash smelting process, the water content of the copper-containing concentrate powder needs to meet the first preset condition. Specifically, the temperature in the reaction shaft (furnace) is relatively high, and if the water content of copper-containing concentrate powder does not meet the first preset condition, the water will form a vapor film on the surface of copper-containing concentrate powder particles, which not only affects the heat transfer of copper-containing concentrate powder during flash smelting, but also hinders the contact between oxygen and copper-containing concentrate powder, so that copper-containing concentrate powder which is not completely reacted falls into the settler to form raw material accumulation, leading to the deterioration of furnace conditions. In some embodiments, the first preset condition may be that the water content of copper-containing concentrate powder is less than 0.3%.

[0061] Oxygen-rich gas (also referred to as oxygen-enriched gas) may refer to a gas whose oxygen content can make copper-containing concentrate powder fully smelting. In some embodiments, the oxygen-rich gas may include any one of air, oxygen-enriched air or industrial oxygen. For example, the oxygen-rich gas may be industrial oxygen with an oxygen content of not less than 99%.

[0062] In some embodiments, in response to the water content of the copper-containing concentrate powder meeting the first preset condition, the processing device may control the mixing device to mix the copper-containing concentrate powder with the oxygen-rich gas, and then control the conveying device to spray the mixture from the nozzle into the smelting furnace.

[0063] Step 330, determining a first production parameter for producing matte based on the physical parameter, the oxygen content of the oxygen-rich gas and the gas temperature of the oxygen-rich gas. In some embodiments, step 330 may be performed by the first determining module 230.

[0064] In some embodiments, the oxygen content of the oxygen-rich gas and the gas temperature of the oxygen-rich gas may affect the efficiency of fully smelting copper-containing concentrate powder in the reaction shaft (furnace). Specifically, the higher the oxygen content of oxygen-rich gas and the higher the gas temperature of oxygen-rich gas, the more complete the reaction between oxygen-rich gas and copper-containing concentrate powder in the reaction shaft (furnace), and the higher the efficiency of fully smelting the copper-containing concentrate powder.

[0065] The first production parameter may refer to a parameter that can meet the demand of matte production by using historical production experience. In some embodiments, the first production parameter may include air supply volume, heavy oil volume, silica flux ratio in the furnace charge, etc. For example, the first production parameter may include that the air supply volume of 0.9 m.sup.3/s, the heavy oil volume of 1.0 kg, and the silica flux ratio in the furnace charge of 0.3%.

[0066] In some embodiments, the first production parameter may be determined by looking up a table based on historical production data statistics. In some embodiments, the processing device 160 may collect the physical parameter, the oxygen content of the oxygen-rich gas, the gas temperature of the oxygen-rich gas and the first production parameter as a data comparison table, and determine the first production parameter based on the data comparison table.

[0067] In some embodiments, the processing device 160 may determine the first production parameter for producing matte based on the physical parameter, oxygen content of oxygen-rich gas and gas temperature of oxygen-rich gas. This operation further includes determining a target vector based on the physical parameter, oxygen content of oxygen-rich gas and gas temperature of oxygen-rich gas, and retrieving in a vector database based on the target vector to determine a matching vector and a production parameter corresponding to the matching vector; and determining the production parameter corresponding to the matching vector as the first production parameter.

[0068] The target vector may refer to a vector constructed based on the physical parameter of copper-containing concentrate powder, oxygen content and gas temperature of oxygen-rich gas, and other characteristics. There are many ways to construct the target vector based on the characteristics, e.g., the physical parameter of copper-containing concentrate powder, oxygen content and gas temperature of oxygen-rich gas. For example, the target vector p is constructed based on the characteristics (x0, x1, x2, y, m) of copper-containing concentrate powder and oxygen-rich gas, where x0, x1 and x2 respectively represent the mineral powder temperature, particle size distribution and mineral powder grade of the physical parameter of copper-containing concentrate powder, y represents the oxygen content of oxygen-rich gas and m represents the gas temperature of oxygen-rich gas.

[0069] In some embodiments, the processing device 160 may construct a target vector based on the detected physical parameter of the copper-containing concentrate powder, the oxygen content and the gas temperature of the oxygen-rich gas.

[0070] The vector database may contain a plurality of reference vectors, and each of the plurality of reference vectors has a corresponding reference first production parameter.

[0071] The reference vector may refer to a vector constructed based on the physical parameter of historical copper-containing concentrate powder, the oxygen content and the gas temperature of historical oxygen-rich gas. The reference first production parameter corresponding to the reference vector is the first production parameter corresponding to the historical copper-containing concentrate powder smelting. There are many ways to construct the reference vector based on the characteristics of historical copper-containing concentrate powder and historical oxygen-rich gas. For example, the reference vector p1 is constructed based on the characteristics (x01, x11, x21, y, m) of historical copper-containing concentrate powder and historical oxygen-rich gas, where x01, x11, x21 respectively represent the historical mineral powder temperature, historical particle size distribution and historical mineral powder grade of the physical parameter of historical copper-containing concentrate powder, y represents the oxygen content of historical oxygen-rich gas, and m represents the gas temperature of historical oxygen-rich gas.

[0072] In some embodiments, the processing device 160 may construct a vector database based on a large number of extensive historical production data, wherein the vector database may include a plurality of reference vectors.

[0073] In some embodiments, the processing device 160 may search in the vector database by using the target vector, and return a reference vector having higher (or highest) similarity with the target vector and its corresponding reference first production parameter, so as to determine the reference first production parameter corresponding to the reference vector as the first production parameter.

[0074] In some embodiments of the present disclosure, the processing device 160 may determine a reasonable first production parameter through the vector database determined by historical production data.

[0075] Step 340, determining a second production parameters for producing matte by processing the physical parameter, the oxygen content of the oxygen-rich gas and the gas temperature of the oxygen-rich gas using the prediction model. In some embodiments, step 340 may be performed by the second determining module 240.

[0076] The second production parameter may refer to a parameter obtained by the machine learning model that may meet the demand of matte production. In some embodiments, the second production parameter may include the air supply, the heavy oil amount, the silica flux ratio in the furnace charge, and the like. For example, the second production parameters may include that air supply is 0.8 m.sup.3/s, heavy oil amount is 1.1 kg, and silica flux ratio of 0.35%.

[0077] In some embodiments, the processing device 160 may determine the second production parameter for producing matte based on the processing of the physical parameter of copper-containing concentrate powder, the oxygen content of oxygen-rich gas and the gas temperature of oxygen-rich gas by the prediction model, which is a machine learning model.

[0078] The prediction model is a model used to predict the production parameters used in a process of producing matte by smelting the copper-containing concentrate powder. In some embodiments, the prediction model may be a machine learning model. For example, the prediction model may include recurrent neural network model (RNN), neural network model (NN), Deep Neural Networks model (DNN), etc. or any combination thereof.

[0079] In some embodiments, the input of the prediction model may include the physical parameter of copper-containing concentrate powder, oxygen content and gas temperature of oxygen-rich gas, and the output may include predicted second production parameter.

[0080] In some embodiments, the prediction model may be trained based on a large amount of historical data. Historical data may include multiple sets of training samples and their corresponding training labels. In some embodiments, historical data may be obtained based on historical production data of copper-containing concentrate powder collected manually. Each group of training samples includes the physical parameter of historical copper-containing concentrate powder, oxygen content and gas temperature of historical oxygen-rich gas. Each training label in multiple groups of training labels includes the historical production parameters of producing through the matte copper-containing concentrate powder smelting corresponding to each group of training samples. The training process includes inputting a group of training samples with training labels into the initial prediction model, update parameters of the initial prediction model through training until the trained intermediate prediction model meets preset conditions, and obtaining the trained prediction model, wherein the preset conditions may be that the loss function is less than a threshold, converges, or the training period reaches a threshold.

[0081] In step 350, determining a target production parameter based on the first production parameters and the second production parameters. In some embodiments, step 350 may be performed by the third determining module 250.

[0082] The target production parameter may refer to a parameter obtained by weighted calculation of the first production parameter and the second production parameter. For example, the first production parameter includes that the air supply is 0.9 m.sup.3/s, the heavy oil amount is 1.0 kg, the silica flux ratio in the furnace charge is 0.35%, the second production parameter includes that the air supply is 0.8 m.sup.3/s, the heavy oil amount is 1.1 kg, and the silica flux ratio in the furnace charge is 0.39%, and their weights are 50% respectively, then the processing device 160 obtains the target production parameter through weighted calculation, and the calculated target production parameter is that the air supply is 0.85 m.sup.3/s, the heavy oil amount is 1.05 kg, and the silica flux ratio in the furnace charge is 0.37%.

[0083] In some embodiments, the weight of the first production parameter and the weight of the second production parameter may be determined according to the distance between the target vector and the matching vector.

[0084] In some embodiments, the processing device 160 may collect the physical parameter, the oxygen content of the oxygen-rich gas, the gas temperature of the oxygen-rich gas, the weight of the first production parameter and the weight of the second production parameter as a data comparison table, and determine the weight of the first production parameter and the weight of the second production parameter based on the data comparison table.

[0085] In some embodiments, the weight of the first production parameter of the target production parameter may be determined according to the distance between the target vector and the matching vector. If the distance between the matching vector from the vector database and the target vector is large (that is, the similarity is low), the target production parameter mostly depend on the second production parameter generated by the prediction model, that is, the weight of the second production parameters is large; otherwise, the target production parameter mostly depend on the first production parameter obtained from the vector database, that is, the weight of the first production parameters is large. For example, the processing device 160 may collate the distance between the matching vector and the target vector, the weight of the first production parameter and the weight of the second production parameter into a data comparison table, and determine the weight of the first production parameter and the weight of the second production parameter based on the data comparison table.

[0086] In some embodiments of the present disclosure, the processing device 160 may obtain more reasonable production parameters according to the first production parameters determined by historical production data and the second production parameters predicted by the model, and then determine the final target production parameters by giving different weights to the first parameters and the second parameters respectively, thus avoiding the influence of insufficient historical production data and making the final target production parameters more reasonable.

[0087] It should be noted that the above description about smelting optimization of copper-containing concentrate is only for illustration and explanation, and does not limit the scope of application of the present disclosure. For those skilled in the art, various modifications and changes may be made to the flow 300 under the guidance of the present disclosure. However, these modifications and changes are still within the scope of the present disclosure.

[0088] In some embodiments, after determining the target production parameters, the processing device may also obtain the fracture surface image of hot matte based on the image recognition device, and determine the measured matte grade through a grade determination model based on the fracture surface image of hot matte, wherein the grade determination model is a machine learning model, and then update the target production parameters based on the measured matte grade. For more details about determining the measured matte grade and updating the target production parameter based on the measured matte grade, please refer to FIGS. 4-5 and their related descriptions.

[0089] In some embodiments, after determining the target production parameters, the processing device may also determine the first update parameter based on the measured value including the measured matte temperature, the measured matte grade and the measured slag iron-silicon ratio and the preset value including the preset matte temperature, the preset matte grade and the preset slag iron-silicon ratio; and updating the target production parameter based on the first update parameter. For more details about determining the first update parameter and updating the target production parameter based on the first update parameter, please refer to FIG. 6 and its related description.

[0090] FIG. 4 is an exemplary flowchart of updating the target production parameter according to some embodiments of the present disclosure. As shown in FIG. 4, the process 400 includes the following steps. In some embodiments, process 400 may be performed by processing device 160. In some embodiments, the process 400 may be performed by the update module 260.

[0091] Step 410, obtaining the fracture surface image of hot matte based on the image recognition device.

[0092] The image recognition device may refer to a device that may obtain the fracture surface image of hot matte. The image recognition device may include a camera device (for example, a camera), and the like.

[0093] The fracture surface image of hot matte may refer to the image which can represent the crystal shape and color of the fracture surface of hot matte. For example, the fracture surface image of hot matte may be presented that the crystal shape of the fracture surface of hot matte is a hexagonal structure and color of the fracture surface of hot matte is a gray-black color.

[0094] In some embodiments, the processing device 160 may control the image recognition device to photograph the fracture surface of hot matte to obtain a fracture surface image of hot matte. In some embodiments, the weir of the smelting furnace may be used for sampling hot matte copper.

[0095] Step 420, determining a measured matte grade through the grade determination model based on the fracture surface image of hot matte.

[0096] The measured matte grade may refer to the actually measured matte content in ore which includes matte. For example, the measured matte grade may be 25%.

[0097] In some embodiments, the processing device 160 may determine the measured matte grade through a grade determination model based on the fracture surface image of hot matte obtained by the image recognition device, and the grade determination model is a machine learning model.

[0098] For more details about the grade model, please refer to FIG. 5 and its related description.

[0099] Step 430, updating the target production parameter based on the measured matte grade.

[0100] In some embodiments, the processing device 160 may update the target production parameter based on the matte temperature monitored by the detection device 120 and the measured matte grade determined by the grade determination model. For example, when the measured matte grade is unqualified, increase the air supply (oxygen quantity) and heavy oil quantity (also referred to as heavy oil amount). For another example, based on the preset rules, when the matte grade is lower than the preset threshold (for example, the preset threshold is qualified value 90%), both the air supply and the heavy oil amount increase by 10%.

[0101] In some embodiments, the processing device 160 may determine a first update parameter based on the detected difference value, and update the target production parameter based on the first update parameter. The detected difference value comprises at least one of the difference between the measured matte temperature and the preset matte temperature, and the difference between the measured matte grade and the preset matte grade.

[0102] For more details about updating the target production parameter, please refer to FIGS. 6-7 and its related descriptions.

[0103] In some embodiments of the present disclosure, the processing device 160 determines the measured matte grade through the grade determination model to update the target production parameter, which effectively improves the production efficiency and product quality.

[0104] It should be noted that the above description of the flow 400 is only for example and explanation, and does not limit the scope of application of the present disclosure. For those skilled in the art, various modifications and changes may be made to the flow 400 under the guidance of the present disclosure. However, these modifications and changes are still within the scope of the present disclosure. In some embodiments, the processing device 160 may also judge whether to update the frequency of acquiring the water content and the frequency of acquiring the physical parameter based on the measured matte grade and other parameters.

[0105] In some embodiments, the processing device 160 may judge whether the measured matte grade, the measured matte temperature and the measured slag iron-silicon ratio meet the second preset condition. In response to the judgement that the measured matte grade, the measured matte temperature and the measured slag iron-silicon ratio do not meet the second preset condition, update the frequency of obtaining water content and the frequency of obtaining physical parameter.

[0106] In some embodiments, in response to the judgement that the measured matte grade, the measured matte temperature, and the measured slag iron-silicon ratio do not meet the second preset condition, the processing device 160 may also adjust the supply of raw materials (for example, change the batch of raw materials).

[0107] The second preset condition may refer to the condition of updating the frequency of obtaining the water content and physical parameter of copper-containing concentrate powder. In some embodiments, the second preset condition may be that the difference between matte temperature and preset matte temperature is less than a first threshold, the difference between matte grade and preset matte grade is less than a second threshold, and the difference between slag iron-silicon ratio and preset slag iron-silicon ratio is less than a third threshold.

[0108] In some embodiments, the processing device 160 may determine the range values of the most suitable matte temperature, the most suitable matte grade and the most suitable slag iron-silicon ratio based on the physical parameter of the copper-containing concentrate powder through statistics, table lookup and other methods, and determine the median of the corresponding range values as the preset matte temperature, the preset matte grade and the preset slag iron-silicon ratio respectively.

[0109] In some embodiments, the matte grade is related to whether the production requirements may be met, and the matte grade is too high or too low to meet the production requirements. If the matte grade is too low, the subsequent blowing will not meet the requirements; if the grade of the matte is too high, there will be less sulfide in matte, and the heat generated will not be enough for the subsequent blowing chemical reaction to proceed spontaneously in the furnace.

[0110] In some embodiments, the matte temperature will affect the matte grade, and when the matte temperature is low, the matte grade may be correspondingly low. In some embodiments, the iron-silicon ratio of slag (also referred to as slag iron-silicon ratio) will affect the production quality. If the iron-silicon ratio of slag is too large, the viscosity of slag will increase, and the melting point of slag will increase, which may generate refractory fouling, which may be mixed in the matte. The part mixed with the fouling will increase the melting point and density of the matte and affect the separation of matte and slag in the furnace.

[0111] In some embodiments, the first threshold, the second threshold and the third threshold may refer to thresholds preset manually based on experience.

[0112] In some embodiments, in response to the fact that the measured matte grade, the measured matte temperature and the measured slag iron-silicon ratio do not meet the second preset condition, that is, the difference between matte temperature and preset matte temperature is greater than the first threshold, the difference between matte grade and preset matte grade is greater than the second threshold, and/or the difference between slag iron-silicon ratio and preset slag iron-silicon ratio is greater than the third threshold, the processing device 160 may update the frequency of obtaining the water content and the frequency of obtaining the physical parameter.

[0113] In some embodiments, during an detection period of the copper-containing concentrate powder production, the processing device 160 may update the frequency of obtaining the water content and the frequency of obtaining the physical parameter (for example, the detection period) based on the total number of times that the difference between the matte temperature and the preset matte temperature is greater than a first threshold, the difference between the matte grade and the preset matte grade is greater than a second threshold, and the difference between the slag iron-silicon ratio and the preset slag iron-silicon ratio is greater than a third threshold. For example, if the total number of times is more than 3 times, the processing device 160 determines update detection period, that is, the frequency of acquiring water content and the frequency of acquiring the physical parameter as half of the previous detection period (for example, the previous detection period is 1 h and the update detection period is 0.5 h). For another example, when at least one of the difference between the measured matte temperature and the preset matte temperature, the difference between the measured matte grade and the preset matte grade, and the difference between the measured slag iron-silicon ratio and the preset slag iron-silicon ratio among the detected difference values exceeds a limit threshold, the processing device 160 determines the updated detection period as half of the previous detection period (for example, the previous detection period is 1 h, and the updated detection period is 0.5 h).

[0114] In some embodiments, the processing device 160 may also update the frequency of acquiring water content and the frequency of acquiring physical parameter based on the detected difference value and the number of occurrences of the abnormal difference; The detected difference value includes at least one of the difference between the measured matte temperature and the preset matte temperature, the difference between the measured matte grade and the preset matte grade, and the difference between the measured slag iron-silicon ratio and the preset slag iron-silicon ratio. The occurrence times of abnormal differences include at least one of the number of times the difference between the measured matte temperature and the preset matte temperature is greater than the first threshold, the number of times the difference between the measured matte grade and the preset matte grade is greater than the second threshold, and the number of times the difference between the measured slag iron-silicon ratio and the preset slag iron-silicon ratio is greater than the third threshold.

[0115] In some embodiments, the processing device 160 may calculate a shortening coefficient of the copper-containing concentrate powder production cycle by weighted calculation based on the detected difference value and the number of occurrences of the abnormal differences. For example, the shortening coefficient

[00001]$H={0.5}^{\frac{A+B}{T}};$

and are normalized coefficients, and A and B are the average size of the detected difference value and the frequency of occurrence of the detected difference value exceeding the corresponding threshold (for example, four times in a detection period); T is the attenuation coefficient, which may be determined according to the frequency of acquiring water content and physical parameter through historical updating (the higher the frequency of acquiring the water content and physical parameter, the greater the T); T may also be determined according to the frequency of updating the target production parameter in historical data (the higher the frequency of updating the target production parameter, the smaller the T). For more details about the frequency of updating the target production parameters, please refer to FIG. 4 and its related description.

[0116] In some embodiments, the processing device 160 may update the frequency of obtaining the water content and the frequency of obtaining the physical parameter based on the shortening factor of the copper-containing concentrate powder production cycle. For example, if the calculated shortening coefficient H of the copper-containing concentrate powder production cycle is 0.8, the processing device 160 determines the product of the previous detection period and the shortening coefficient H of the copper-containing concentrate powder production cycle as the updated detection period for acquiring water content and for acquiring physical parameter (for example, the previous detection period is 5 h and the updated detection period is 4 h).

[0117] In some embodiments of the present disclosure, the processing device 160 may improve the rationality of production monitoring, find out the problems of raw materials in time and strengthen the monitoring frequency based on the detected difference value and the frequency of abnormal differences, so as to avoid affecting subsequent production.

[0118] In some embodiments, the processing device 160 may judge whether the measured matte grade, the measured matte temperature and the measured slag iron-silicon ratio meet the third preset condition. In response to the judgement that the measured matte grade, the measured matte temperature and the measured slag iron-silicon ratio do not meet the third preset condition, the target production parameter are updated.

[0119] The third preset condition may refer to the condition of updating the target production parameter. In some embodiments, the third preset condition may be that the difference between matte temperature and preset matte temperature is less than a fourth threshold, the difference between the matte grade and preset matte grade is less than a fifth threshold, and the difference between the slag iron-silicon ratio and preset slag iron-silicon ratio is less than a sixth threshold.

[0120] In some embodiments, in response to the judgement that the measured matte grade, the measured matte temperature and the measured slag iron-silicon ratio do not meet the third preset condition, that is, the difference between the matte temperature and preset matte temperature is greater than the fourth threshold, the difference between the matte grade and preset matte grade is greater than the fifth threshold, and/or the difference between the slag iron-silicon ratio and preset slag iron-silicon ratio is greater than the sixth threshold, the processing device 160 may update the target production parameter.

[0121] In some embodiments, the processing device 160 may update the target production parameter based on the detected difference value.

[0122] For more details about updating the target production parameter, please refer to FIGS. 6-7 and their related descriptions.

[0123] In some embodiments of the present disclosure, the processing device 160 dynamically determines when to update the target production parameters based on the measured matte grade, the measured matte temperature, and the measured slag iron-silicon ratio not meeting the third preset condition, so that the target production parameter may be updated in time and the product quality may be improved.

[0124] FIG. 5 is an exemplary schematic diagram of a grade determination model according to some embodiments of the present disclosure.

[0125] As shown in FIG. 5, in some embodiments, the grade determination model 520 includes an image extraction layer 520-1, a correction layer 520-2 and a judgment layer 520-3.

[0126] In some embodiments, the image extraction layer 520-1 may be CNN, the correction layer 520-2 and the judgment layer 520-3 may be NN.

[0127] In some embodiments, the input of the image extraction layer 520-1 may include a hot matte fracture surface image 510-1 (also referred to as fracture surface image of hot matte), and the output may include a fracture surface feature 530-1.

[0128] In some embodiments, the fracture surface feature 530-1 may be represented by a fracture surface feature vector, which may refer to a vector constructed based on the fracture surface image feature information of hot matte, and the fracture surface image feature information of hot matte may include color features, crystal particle size features and crystal shape features. For example, the fracture surface feature vector p is constructed based on the features (x.sub.4, y.sub.2, m.sub.2) of the fracture surface image of hot matte, where the features (x.sub.4, y.sub.2, m.sub.2) may indicate that the color feature of the corresponding fracture surface is x.sub.4, the crystal particle size feature is y.sub.2, and the crystal shape feature is m.sub.2.

[0129] In some embodiments, the input of the correction layer 520-2 may include the fracture surface feature 530-1, the difference 510-2 between the matte discharge time and the image acquisition time, and the output may include the corrected fracture surface feature 530-2.

[0130] In some embodiments, the detection device 120 may obtain the matte discharge time and the image acquisition time, and the processing device 160 may calculate the time difference between them based on the matte discharge time and the image acquisition time. Specifically, the matte discharge time may refer to the time point when the matte product leaves from the reaction furnace after the copper-containing concentrate powder is completely smelted in the reaction furnace, and the image acquisition time may refer to the time point when the image recognition device acquires the fracture surface image of the matte. For example, when the matte leaves from the furnace at 12:00 and the image acquisition time is 12:01, the difference 510-2 between the matte discharge time and the image acquisition time is 1 minute.

[0131] In some embodiments, due to the high temperature (1000 C.-1400 C.) of matte when it is discharged from the furnace, the temperature of matte will drop sharply after it flows out of the reaction furnace. When the image recognition device obtains the fracture surface image 510-1 of hot matte, the fracture surface features of matte have changed greatly with the sharp drop of matte temperature, so the input of the correction layer 520-2 needs to include the discharge time of matte and image acquisition. By inputting the difference 510-2 between the matte discharge time and the image acquisition time to the correction layer 520-2, the corrected fracture surface feature 530-2 of matte may be output more accurately.

[0132] In some embodiments, the corrected fracture surface feature 530-2 may be represented by a corrected fracture surface feature vector, which may refer to a fracture surface feature vector modified based on the difference 510-2 between matte discharge time and image acquisition time. The corrected fracture surface feature information may include a corrected color feature, a corrected crystal particle size feature and a corrected crystal shape feature. For example, a corrected fracture surface feature vector p is constructed based on the corrected features (x.sub.5, y.sub.3, m.sub.3) of the fracture surface image matte, wherein the corrected features (x.sub.5, y.sub.3, m.sub.3) may indicate that the corresponding corrected color feature is x, the corrected crystal particle size feature is y.sub.3, and the corrected crystal shape feature is m.sub.3.

[0133] In some embodiments, the input of the judgment layer 520-3 may include a corrected fracture surface feature 530-2, and the output may include a measured matte grade 540.

[0134] In some embodiments, the output of the image extraction layer 520-1 may be the input of the correction layer 520-2, the output of the correction layer 520-2 may be the input of the judgment layer 520-3, and the image extraction layer 520-1, the correction layer 520-2 and the judgment layer 520-3 may be jointly trained. In some embodiments, the sample data of the joint training includes the image data of the sample hot matte fracture surface image data obtained from the actual production data, the difference between the sample matte discharge time and the image acquisition time, and the label is the sample matte grade determined artificially according to the actual production data (for example, the matte grade obtained artificially through experimental detection based on experimental production data). Inputting the of the sample hot matte fracture surface image data into the image extraction layer 520-1 to obtain the fracture surface features output by the image extraction layer 520-1; Inputting the fracture surface features output by the image extraction layer 520-1 and the difference between the sample matte discharge time and the image acquisition time into the correction layer 520-2 to obtain the corrected fracture surface features output by the correction layer 520-2; The corrected fracture surface features output by the correction layer 520-2 are input to the judgment layer 520-3 to obtain the measured matte grade output by the judgment layer 520-3. During the training process, the processing device 160 may construct a loss function based on the sample matte grade in the label and the measured matte grade output by the judgment layer 520-3. At the same time, the parameters of the image extraction layer 520-1, the correction layer 520-2 and the judgment layer 520-3 are updated until the preset conditions are met and the training is completed. Among them, the preset condition may be one or more of loss function less than threshold, convergence, or training period reaching threshold.

[0135] By jointly training the image extraction layer 520-1, the correction layer 520-2 and the judgment layer 520-3, it is beneficial to solve the problem that it is difficult to obtain labels when training the image extraction layer 520-1, the correction layer 520-2 and the judgment layer 520-3 separately. Secondly, jointly training the image extraction layer 520-1, the correction layer 520-2 and the judgment layer 520-3 may not only reduce the number of samples needed, but also improve the training efficiency.

[0136] In some embodiments, the traditional method for determining the matte grade by manually observing the cross section of hot matte is inefficient, the accuracy fluctuates greatly, and there may be potential safety hazards. Using the trained grade determination model to determine the matte grade may not only improve the efficiency and accuracy, but also reduce the incidence of personal injury incidents.

[0137] FIG. 6 is an exemplary flowchart of updating a target production parameter based on a first update parameter according to some embodiments of the present disclosure. As shown in FIG. 6, the flow 600 includes the following steps. In some embodiments, process 600 may be performed by processing device 160. In some embodiments, the process 600 may be performed by the update module 260.

[0138] Step 610, determining a first update parameter based on the measured value and the preset value.

[0139] The measured value includes the measured matte temperature, the measured matte grade and the measured slag iron-silicon ratio, and the preset value includes the preset matte temperature, the preset matte grade and the preset slag iron-silicon ratio. For example, the measured value may be (1100, 45%, 5), where 1100 represents the measured matte temperature, 45% represents the measured matte grade, and 5 represents the measured slag iron-silicon ratio. The preset value may be (1200, 50%, 4), where 1200 represents the preset matte temperature, 50% represents the preset matte grade, and 4 represents the preset slag iron-silicon ratio. For more information about measured matte temperature, measured matte grade, measured slag iron-silicon ratio, preset matte temperature, preset matte grade and preset slag iron-silicon ratio, see FIG. 4 and its related descriptions.

[0140] In some embodiments, the measured matte grade may be determined based on a grade determination model.

[0141] The first update parameter may refer to a production parameter determined based on a measured value and a preset value. For example, the first update parameter may include an air supply of 0.79 m.sup.3/s, a heavy oil amount of 1.01 kg, and a silica flux ratio in the charge of 0.35%.

[0142] In some embodiments, the processing device 160 may collate the difference between the measured matte temperature and the preset matte temperature, the difference between the measured matte grade and the preset matte grade, the difference between the measured slag iron-silicon ratio and the predicted slag iron-silicon ratio, and the first update parameter into a data comparison table, and determine the first update parameter based on the data comparison table.

[0143] In some embodiments, the processing device 160 may determine at least one group of candidate production parameters based on the target production parameter; determine at least one predicted matte temperature, at least one predicted matte grade and at least one predicted slag iron-silicon ratio based on at least one group of candidate production parameters; and determine a first update parameter based on at least one predicted matte temperature, at least one predicted matte grade, at least one predicted slag iron-silicon ratio, symmetrical matte temperature, symmetrical matte grade and symmetrical slag iron-silicon ratio. The symmetrical matte temperature and the measured matte temperature are symmetrical based on a preset matte temperature, the symmetrical matte grade and the measured matte grade are symmetrical based on a preset matte grade, and the symmetrical slag iron-silicon ratio and the measured slag iron-silicon ratio are symmetrical based on a preset slag iron-silicon ratio. For more details about determining the first update parameter, please refer to FIG. 7 and its related description.

[0144] Step 620, updating the target production parameter based on the first update parameter.

[0145] In some embodiments, the processing device 160 may average the first update parameter and the target production parameter to determine the updated target production parameter.

[0146] In some embodiments, the processing device 160 may determine a second update parameter based on the first production parameter and the first update parameter; Based on the second update parameter, the target production parameter is updated.

[0147] The second update parameter may refer to a production parameter determined based on the first production parameter and the first update parameter. For example, the second update parameter may include an air supply of 0.80 m.sup.3/s, a heavy oil amount of 1.02 kg, and a silica flux ratio in the charge of 0.36%.

[0148] In some embodiments, the processing device 160 may calculate the first production parameter and the first update parameter to determine the second update parameter. The weight of the first production parameter is related to the distance between the target vector and the matching vector, and the closer the distance, the greater the weight of the first production parameter; the weight of the first update parameter is also related to the distance between the predicted value and the symmetric value, and the closer the distance, the greater the weight of the first update parameter. If the distance between the target vector and the matching vector is smaller than the distance between the predicted value and the symmetric value, the second update parameter mostly depends on the first production parameter, that is, the weight of the first production parameter is larger; otherwise, the second update parameter mostly depends on the first update parameter, that is, the weight of the first update parameter is larger. In some embodiments, the processing device 160 may collate the distance between the target vector and the matching vector, the distance between the predicted value and the symmetric value, the weight of the first production parameter and the weight of the first update parameter into a data comparison table, and determine the weight of the first production parameter and the weight of the first update parameter based on the data comparison table. For more details about the distance between the predicted value and the symmetric value, please refer to FIG. 7 and its related description.

[0149] In some embodiments, the processing device 160 may determine the second update parameter as the target production parameter.

[0150] In some embodiments of the present disclosure, the second update parameter is determined by the weighted calculation of the first production parameter and the first update parameter, and then the target production parameter is further updated, so that the updated target production parameter is more reasonable and the product quality is improved.

[0151] In some embodiments of the present disclosure, the first update parameter is determined based on the detected difference value, and then the target production parameter is updated according to the first update parameter, so that the working parameters may be adjusted in time, which may avoid the shutdown caused by untimely adjustment and improve the production efficiency.

[0152] It should be noted that the above description of the flow 600 is only for illustration and explanation, and does not limit the scope of application of the present disclosure. For those skilled in the art, various modifications and changes may be made to the process under the guidance of the present disclosure. However, these modifications and changes are still within the scope of the present disclosure.

[0153] FIG. 7 is an exemplary flowchart of determining a first update parameter based on a detected difference value according to some embodiments of the present disclosure. As shown in FIG. 7, the flow 700 includes the following steps. In some embodiments, process 700 may be performed by processing device 160. In some embodiments, the process 700 may be performed by the update module 260.

[0154] Step 710, determining at least one group of candidate production parameters based on the target production parameter.

[0155] The candidate production parameters may refer to production parameters to be determined as the first update parameters. In some embodiments, the processing device 160 may generate multiple groups of candidate production parameters based on the production parameters corresponding to the measured values.

[0156] In some embodiments, the processing device 160 may generate a plurality of equally spaced parameters for each of the target production parameters, and then randomly combine these parameters to form at least one group of candidate production parameters. For example, the target production parameter is (0.85, 1.05, 0.37%), 0.85 represents the air supply of 0.85 m.sup.3/s, 1.05 represents the heavy oil amount of 1.05 kg, 0.37% represents the silica flux ratio in the charge. The first parameter is 0.85, and the values of corresponding several equally spaced parameters may be (0.75, 0.8, 0.85, 0.9, 0.95). The second parameter is 1.05, and the values of corresponding several equally spaced parameters may be (0.95, 1, 1.05, 1.1, 1.15). The third parameter is 0.37, and the values of corresponding several equally spaced parameters may be (0.33, 0.35, 0.37, 0.39, 0.41). Randomly combining a plurality of equally spaced parameters corresponding to the first parameter, a plurality of equally spaced parameters corresponding to the second parameter and a plurality of equally spaced parameters corresponding to the third parameter to form at least one group of candidate production parameters. For example, the processing device 160 may select 0.8 of a plurality of equally spaced parameters corresponding to the first parameter, 1.1 of a plurality of equally spaced parameters corresponding to the second parameter, and 0.37 of a plurality of equally spaced parameters corresponding to the third parameter to form a group of candidate production parameters (0.8, 1.1, 0.37%). In some embodiments, the number and interval of equally spaced parameter values may be set manually.

[0157] In some embodiments, some candidate production parameters may be obtained by transformation based on the above groups of candidate production parameters. Among them, transformation may refer to the exchange of parameters in the same position of two groups of candidate production parameters or the adjustment of any one of a group of candidate production parameters. For example, two groups of candidate production parameters (0.75, 1, 0.35%) and (0.8, 1.1, 0.39%) may be exchanged to obtain two new groups of candidate production parameters (0.75, 1.1, 0.35%) and (0.8, 1, 0.39%). For another example, by adjusting 0.35% of the candidate production parameters (0.75,1,0.35%) to the preset value of 0.4%, new candidate production parameters (0.75, 1,0.4%) may be obtained.

[0158] Step 720, determining at least one predicted value based on at least one group of candidate production parameters.

[0159] In some embodiments, the predicted value may include predicted matte temperature, predicted matte grade, and predicted slag iron-silicon ratio.

[0160] Predicted matte temperature may refer to matte temperature obtained by prediction, predicted matte grade may refer to matte grade obtained by prediction, and predicted slag iron-silicon ratio may refer to slag iron-silicon ratio obtained by prediction. In some embodiments, at least one predicted matte temperature, at least one predicted matte grade and at least one predicted slag iron-silicon ratio may be predicted by historical experience, and may also be retrieved in a vector database. The process of obtaining the predicted matte temperature and the predicted matte grade based on the vector database is similar to the process of obtaining the first production parameter based on the vector database, and will not be described here.

[0161] In some embodiments, the processing device 160 may input the candidate production parameters into the evaluation model to obtain a predicted matte temperature, a predicted matte grade, and a predicted slag iron-silicon ratio.

[0162] The evaluation model may refer to a machine learning model for outputting predicted matte temperature and predicted matte grade. In some embodiments, the evaluation model may be a trained machine learning model. For example, the evaluation model may include any one or combination of RNN, NN, DNN, etc.

[0163] In some embodiments, the input of the evaluation model may be candidate production parameters, physical parameter of copper-containing concentrate powder, oxygen content and gas temperature of oxygen-rich gas. For details of physical parameter of copper-containing concentrate powder, oxygen content and gas temperature of oxygen-rich gas, see FIG. 3 and its related descriptions.

[0164] In some embodiments, the output of the evaluation model may be predicted matte temperature, predicted matte grade and predicted slag iron-silicon ratio.

[0165] In some embodiments, the evaluation model may be trained based on a large number of labeled training samples. For example, the labeled training samples are input into the initial evaluation model, the loss function is constructed through the labels and the prediction results of the initial evaluation model, and the parameters of the model are iteratively updated based on the loss function. When the trained model meets the preset condition, the training ends and a trained evaluation model is obtained. The preset condition includes that the loss function converges and the number of iterations reaches the threshold.

[0166] In some embodiments, the training samples may be production parameters in historical data, physical parameter of copper-containing concentrate powder, oxygen content and gas temperature of oxygen-rich gas. The labels may be actual matte temperature, actual matte grade and actual slag iron-silicon ratio.

[0167] Step 730, determining the first update parameter based on at least one predicted value and the symmetrical value.

[0168] In some embodiments, the symmetrical values may include symmetrical matte temperature, symmetrical matte grade and symmetrical slag iron-silicon ratio.

[0169] In some embodiments, the symmetrical value and the measured value are symmetrical based on the predicted value, that is, the symmetrical matte temperature and the measured matte temperature are symmetrical based on the preset matte temperature, the symmetrical matte grade and the measured matte grade are symmetrical based on the preset matte grade, and the symmetrical slag iron-silicon ratio and the measured slag iron-silicon ratio are symmetrical based on the preset slag iron-silicon ratio. For example, the default value is (1200, 50%, 4), the measured value is (1100, 45%, 5), and the symmetrical value is (1300, 55%, 3).

[0170] In some embodiments, the processing device 160 may input the candidate production parameters into the evaluation model, determine the predicted matte grade, predicted matte temperature and predicted slag iron-silicon ratio, and determine the candidate production parameters corresponding to the predicted value closest to the symmetrical value as the first update parameters.

[0171] For example only, the preset values of matte temperature, matte grade and slag iron-silicon ratio are (1200, 50%, 4), and the measured values are (1100, 45%, 5), so the symmetrical value is (1300, 55%, 3). There are three groups of candidate production parameters A, B and C, and the predicted values (predicted matte temperature, predicted matte grade and predicted slag iron-silicon ratio) obtained by inputting them into the evaluation model are PA (1250, 49%, 3.5), PB (1350, 53%, 4) and PC (1450, 57%, 4) respectively. The vector distances between PA, PB, PC and the symmetric value (1300, 55%, 3) are calculated respectively. The smaller the vector distance, the closer it is to the symmetric value. If the vector distance between PA and the symmetric value (1300,55%, 3) is the smallest, the candidate production parameter A is taken as the first update parameter.

[0172] In some embodiments, the processing device 160 may determine the vector distance based on the predicted matte temperature, predicted matte grade, predicted slag iron-silicon ratio, symmetrical matte temperature, symmetrical matte grade and symmetrical slag iron-silicon ratio. For example, the vector distance between the predicted value PA (1250, 49%, 3.5) and the symmetric value PA (1300, 55%, 3) may be determined based on the following formula:

[00002]$\sqrt{a{\left(1250-1300\right)}^2+b{\left(49\%-55\%\right)}^2+c{\left(3.5-3\right)}^2}.$

Among them, A, B and C respectively represent the weights of matte temperature, matte grade and slag iron-silicon ratio, which are set manually and empirically.

[0173] Because the preset matte temperature, preset matte grade and preset slag iron-silicon ratio are optimal values, the measured matte temperature, measured matte grade and measured slag iron-silicon ratio have a certain gap with the optimal values. Symmetrical matte temperature and measured matte temperature are symmetrical based on the preset matte temperature, symmetrical matte grade and measured matte grade are symmetrical based on the preset matte grade, and symmetrical slag iron-silicon ratio and measured slag iron-silicon ratio are symmetrical based on the preset slag iron-silicon ratio. Therefore, it is hoped that the predicted matte temperature, predicted matte grade and predicted slag iron-silicon ratio determined based on the first update parameter may be as close as possible to the symmetrical matte temperature, symmetrical matte grade and symmetrical slag iron-silicon ratio (equivalent to adjusting the production parameters from the opposite direction), so that the matte temperature, matte grade and slag iron-silicon ratio may be closer to the preset matte temperature, preset matte grade and preset slag iron-silicon ratio after production based on the first update parameter.

[0174] In some embodiments of the present disclosure, based on the target production parameters, candidate production parameters are determined, and the first update parameters are further determined, so that the first update parameters are more reasonable and accurate, so as to ensure the production requirements and meet the quality requirements.

[0175] It should be noted that the above description of the flow 700 is only for example and explanation, and does not limit the scope of application of the present disclosure. For those skilled in the art, various modifications and changes may be made to the process breath-holding training under the guidance of the present disclosure. However, these modifications and changes are still within the scope of the present disclosure.

[0176] One embodiment of the present disclosure provides a smelting optimization device for copper-containing concentrate. The device comprises a water content monitoring device, detection device, a conveying device, a smelting furnace and processing device. The water content monitoring device is used for obtaining the water content of copper-containing concentrate powder; the detection device is used to obtain the physical parameter of copper-containing concentrate powder; the conveying device is used for mixing copper-containing concentrate powder with oxygen-rich gas in response to the water content meeting the first preset condition, and then spraying the mixture into the smelting furnace from the nozzle. The smelting furnace is used for smelting the mixture; the processing device is used for executing the first determining module, the second determining module and the third determining module.

[0177] One embodiment of the present disclosure provides a computer-readable storage medium that stores computer instructions, and when the computer reads the computer instructions in the storage medium, the computer executes the smelting optimization method of copper-containing concentrate.

[0178] The basic concepts have been described above. Obviously, for those skilled in the art, the above detailed disclosure is only an example, and does not constitute a limitation of the present disclosure. Although it is not explicitly stated here, those skilled in the art may make various modifications, improvements and corrections to the present disclosure. Such modifications, improvements and corrections are suggested in the present disclosure, so they still belong to the spirit and scope of the exemplary embodiments of the present disclosure.

[0179] Meanwhile, the present disclosure uses specific words to describe the embodiments of the present disclosure. Such as one embodiment, an embodiment and/or some embodiments mean a certain feature, structure or characteristic related to at least one embodiment of the present disclosure. Therefore, it should be emphasized and noted that references to one embodiment or one embodiment or an alternative embodiment in different places in the present disclosure do not necessarily refer to the same embodiment. In addition, some features, structures or characteristics in one or more embodiments of the present disclosure may be combined appropriately.

[0180] In addition, unless explicitly stated in the claims, the order of processing elements and sequences, the use of numbers and letters, or the use of other names mentioned in the present disclosure are not used to limit the order of the procedures and methods in the present disclosure. Although some presently considered useful embodiments of the invention have been discussed through various examples in the above disclosure, it should be understood that such details are only for illustrative purposes, and the appended claims are not limited to the disclosed embodiments. On the contrary, the claims are intended to cover all modifications and equivalent combinations that conform to the spirit and scope of the embodiments in the present disclosure. For example, although the system components described above may be realized by hardware devices, they may also be realized only by software solutions, such as installing the described system on existing servers or mobile devices.

[0181] In the same way, it should be noted that in order to simplify the expression disclosed in the present disclosure and help to understand one or more inventive embodiments, in the previous description of the embodiments of the present disclosure, sometimes a variety of features are merged into one embodiment, the drawings or the description thereof. However, this disclosure method does not mean that the object of the present disclosure needs more features than those mentioned in the claims. In fact, the features of the embodiment are less than all the features of the single embodiment disclosed above.

[0182] In some embodiments, numbers describing the number of ingredients and attributes are used. It should be understood that such numbers used in the description of embodiments are modified by the modifiers about, approximate or substantially in some examples. Unless otherwise specified, approximately, approximately or substantially means that the number allows a variation of plus or minus 20%. Accordingly, in some embodiments, the numerical parameters used in the specification and claims are approximations, which may be changed according to the required characteristics of individual embodiments. In some embodiments, numerical parameters should consider the specified significant digits and adopt the method of general digit reservation. Although the numerical fields and parameters used to confirm the range and breadth in some embodiments of the present disclosure are approximations, in specific embodiments, such numerical values are set as accurately as possible within the feasible range.

[0183] For each patent, patent application, patent application publication and other materials cited in the present disclosure, such as articles, books, specifications, publications, documents, etc., the entire contents thereof are hereby incorporated into the present disclosure by reference. Except for the application history documents that are inconsistent with or conflict with the contents of the present disclosure, and the documents that limit the widest scope of the claims in the present disclosure (currently or later attached to the present disclosure). It should be noted that in case of any inconsistency or conflict between the descriptions, definitions and/or terms used in the attached materials of the present disclosure and those described in the present disclosure, the descriptions, definitions and/or terms used in the present disclosure shall prevail.

[0184] Finally, it should be understood that the embodiments described in the present disclosure are only used to illustrate the principles of the embodiments in the present disclosure. Other variations may also fall within the scope of the present disclosure. Therefore, by way of example and not limitation, alternative configurations of embodiments of the present disclosure may be regarded as consistent with the teachings of the present disclosure. Accordingly, the embodiments in the present disclosure are not limited to those explicitly introduced and described in the present disclosure.