Disclosed herein is a method for predicting a faulty behaviour of an electrical converter. The method includes receiving an operation point indicator of the electrical converter indicative of an actual operation point of the electrical converter, where the electrical converter is connected to a rotating electrical machine; receiving a measured device temperature of a power semiconductor device of the electrical converter indicative of an actual temperature of the power semiconductor device; inputting the operation point indicator as input data into a machine learning algorithm trained with historical data comprising operation point indicators and associated device temperatures, where the historical data was recorded during normal operation of a power semiconductor device; estimating an estimated device temperature with the machine learning algorithm, where the estimated device temperature represents a device temperature during a normal operation; and predicting the faulty behaviour by comparing the estimated device temperature with the measured device temperature.

A neural network method, comprising: modeling an environment; implementing a policy based on the modeled environment, to perform an action by an agent within the environment, having at least one estimated dynamic parameter; receiving an observation and a temporally-associated cost or reward based on operation of the agent in the environment controlled according to the policy; and updating the policy, dependent on the received observation and the temporally-associated cost or reward, to improve the policy to optimize an expected future cumulative cost or reward. The policy may represent a set of parameters defining an artificial neural network having a plurality of hierarchical layers and having at least one layer which receives inputs representing aspects of the received observation indirectly from other neurons, and produce outputs to other neurons which indirectly implement the policy, the plurality of hierarchical layers being trained according to respectfully distinct training criteria.

An operation management system is disclosed. The operation management system may receive a video stream from a wearable device of a user that is performing an operation in a physical environment. The operation management system may process, using an operation performance model, a set of frames of the video stream that indicates a state of a performance of the operation by the user. The operation management system may determine, based on the state of the performance by the user, a next task of the operation. The operation management system may configure display data that is associated with a physical object that is associated with the next task. The display data may be associated with an indicator that identifies the physical object and/or task information associated with performing the next task. The operation management system may provide the display data to the wearable device.

A method operates a fault indicator, in particular a fault current indicator, which can detect a fault in an electrical energy transmission line, in particular a fault current in the energy transmission line. In the event of a detected fault, the fault indicator transmits a fault signal to a superordinate control center monitoring the energy transmission line. The fault indicator regularly or irregularly transmits at least one item of local weather information, which relates to the weather in the environment of the fault indicator, to the superordinate control center, directly to a central device other than the control center or indirectly to the other central device via the control center.

The invention relates to a method for generating a control system for a target system, wherein: operational data is received; a first neural model component is trained with the received operational data for generating a prediction on a state of the target system based on the received operational data; a second neural model component is trained with the operational data for generating a regularizer for use in inverting the first neural model component; and the control system is generated by inverting the first neural model component by optimization and arranging to apply the regularizer generated with the second neural model component in the optimization. The invention relates also to a system and a computer program product.

A mechanical ventilation system comprises a mechanical ventilator configured to deliver ventilation to a patient. An electronic controller is programmed to control the mechanical ventilator to perform a mechanical insufflation-exsufflation (MI-E) therapy method including performing a MI-E cycle including: (i) during an insufflation cycle, delivering pressure to the patient at a positive insufflation gauge pressure; (ii) during an exsufflation cycle following step (i), delivering pressure to the patient at a negative exsufflation gauge pressure and detecting whether an upper airway collapse occurs; and (iii) reducing a magnitude of the negative exsufflation gauge pressure if an upper airway collapse is detected in step (ii).

Systems and methods are provided for maintaining an air conditioning system. A system can include one or more sensors positioned inside of the air conditioning system configured to transmit current sensor data to a remote location. A data repository contains historic sensor data and corresponding air conditioning system status data. A neural network is trained using the historic sensor data and the corresponding air conditioning system status data to predict a future air conditioning system status based on the transmitted current sensor data. A server computer system is configured to predict the future air conditioning system status based on the current sensor data using the neural network, and a graphical user interface is configured to display the predicted future air conditioning system status to a remote client. The current sensor data is stored in the data repository and the neural network is further trained based on the current sensor data.

Provided is a machine learning data generation device including: a virtual sensor input generator configured to generate a virtual sensor input, which is obtained by virtually generating a sensor input obtained as a result of performing sensing, by a sensor of a work machine, on a plurality of randomly piled subjects to be subjected to physical work by an operating machine of the work machine; a virtual operation command generator configured to generate a virtual operation command, which is obtained by virtually generating an operation command for the operating machine of the work machine; a virtual operation outcome evaluator configured to evaluate an outcome of the physical work in response to the virtual operation command in a virtual space; and a machine learning data generator configured to generate machine learning data based on the virtual sensor input, the virtual operation command, and the evaluation of the virtual operation outcome evaluator.

An operation management system is disclosed. The operation management system may receive a video stream from a wearable device of a user that is performing an operation in a physical environment. The operation management system may process, using an operation performance model, a set of frames of the video stream that indicates a state of a performance of the operation by the user. The operation management system may determine, based on the state of the performance by the user, a next task of the operation. The operation management system may configure display data that is associated with a physical object that is associated with the next task. The display data may be associated with an indicator that identifies the physical object and/or task information associated with performing the next task. The operation management system may provide the display data to the wearable device.

The system generally includes a crosspoint switch in the local data collection system having multiple inputs and multiple outputs including a first input connected to the first sensor and a second input connected to the second sensor. The multiple outputs include a first output and a second output configured to be switchable between a condition in which the first output is configured to switch between delivery of the first sensor signal and the second sensor signal and a condition in which there is simultaneous delivery of the first sensor signal from the first output and the second sensor signal from the second output. Each of multiple inputs is configured to be individually assigned to any of the multiple outputs. Unassigned outputs are configured to be switched off producing a high-impedance state. The local data collection system includes multiple data acquisition units each having an onboard card set configured to store calibration information and maintenance history of a data acquisition unit in which the onboard card set is located. The local data collection system is configured to manage data collection bands.