The present application relates to a method, device, and system for detecting welding spot quality abnormalities based on deep learning. The method includes: acquiring a dynamic welding parameter in a welding process corresponding to any target welding spot; inputting the dynamic welding parameter into a pre-trained dynamic welding parameter simulation model for simulation, and acquiring a welding simulation parameter output by the dynamic welding parameter simulation model; determining a deviation of the dynamic welding parameter from the welding simulation parameter, and determining that the target welding spot is an abnormal welding spot when the deviation is greater than a preset threshold. The solution of the present application can reduce the frequency of manual tearing down and batches for abnormality detection, which has a faster abnormality detection speed and may cover all welding spots.

A welding apparatus includes a welding rod; a temperature measuring jig to measure a temperature of the welding rod; and a welding rod cooler to cool the welding rod.

A method for determining the cooling rate of a weld, comprising locating at least one non-consumable thermal sensor at a predetermined location from a weld pool generated during an active welding process, wherein the thermal sensor is located within the same plane of travel as a welding device creating the weld pool; determining the travel speed of the welding device; using the at least one non-consumable thermal sensor to gather temperature data from the weld pool; and determining the time interval between when the weld pool has a first measured temperature and when the weld pool has a second measured temperature that is less than the first measured temperature, wherein the determined time interval represents the cooling rate of the weld.

A method for determining the cooling rate of a weld, comprising locating at least one non-consumable thermal sensor at a predetermined location from a weld pool generated during an active welding process, wherein the thermal sensor is located within the same plane of travel as a welding device creating the weld pool; determining the travel speed of the welding device; using the at least one non-consumable thermal sensor to gather temperature data from the weld pool; and determining the time interval between when the weld pool has a first measured temperature and when the weld pool has a second measured temperature that is less than the first measured temperature, wherein the determined time interval represents the cooling rate of the weld.

An intrinsic process signal-based online expulsion detection method for resistance spot welding process, which comprises: acquiring the intrinsic process signal and current signal output by sensors installed at two electrodes in real-time during the welding process and establishing a relationship graph; performing expulsion judgement based on the relationship graph to obtain expulsion frequency, single intrinsic process signal feature and an accumulated feature; calculating expulsion metal volume according to the accumulated feature and electrode profile features to obtain a prediction expulsion metal amount. The method performs online prediction of the expulsion metal amount according to the intrinsic process signal for resistance spot welding process, thereby achieving online quantitative estimation of the expulsion intensity, overcoming the defect of the traditional technology which relies on manual detection, and improving detection efficiency and accuracy.

An intrinsic process signal-based online expulsion detection method for resistance spot welding process, which comprises: acquiring the intrinsic process signal and current signal output by sensors installed at two electrodes in real-time during the welding process and establishing a relationship graph; performing expulsion judgement based on the relationship graph to obtain expulsion frequency, single intrinsic process signal feature and an accumulated feature; calculating expulsion metal volume according to the accumulated feature and electrode profile features to obtain a prediction expulsion metal amount. The method performs online prediction of the expulsion metal amount according to the intrinsic process signal for resistance spot welding process, thereby achieving online quantitative estimation of the expulsion intensity, overcoming the defect of the traditional technology which relies on manual detection, and improving detection efficiency and accuracy.

A joining device includes: an electrode for transmitting a current to an object to be joined; a current supply section for generating the current supplied to the object due to a voltage being applied thereto; a current measurement section for measuring the current supplied to the object; and a first control unit for correcting the voltage applied to the current supply section based on a current value measured before a last time by the current measurement section such that a current value measured by the current measurement section becomes a target current value. A method for manufacturing a joined object includes: supplying a current, which is generated by applying a voltage to a current supply section, to the object to be joined; measuring the current supplied to the object; and correcting the voltage applied to the current supply section based on a ratio between the measured current value and a desired current value.

Provided is a resistance spot welding method wherein main current passage includes two or more electrode force application steps including a first electrode force application step and a second electrode force application step following the first electrode force application step, an electrode force F.sub.1 in the first electrode force application step and an electrode force F.sub.2 in the second electrode force application step in the main current passage satisfy a relationship F.sub.1<F.sub.2, and an electrode force switching point T.sub.f from the first electrode force application step to the second electrode force application step in the main current passage is set to satisfy predetermined relational formulas.

Provided is a resistance spot welding method wherein main current passage includes two or more electrode force application steps including a first electrode force application step and a second electrode force application step following the first electrode force application step, an electrode force F.sub.1 in the first electrode force application step and an electrode force F.sub.2 in the second electrode force application step in the main current passage satisfy a relationship F.sub.1<F.sub.2, and an electrode force switching point T.sub.f from the first electrode force application step to the second electrode force application step in the main current passage is set to satisfy predetermined relational formulas.

Provided is a resistance spot welding method wherein main current passage includes two or more electrode force application steps including a first electrode force application step and a second electrode force application step following the first electrode force application step, an electrode force F.sub.1 in the first electrode force application step and an electrode force F.sub.2 in the second electrode force application step in the main current passage satisfy a relationship F.sub.1>F.sub.2, and an electrode force switching point T.sub.f from the first electrode force application step to the second electrode force application step in the main current passage is set to satisfy predetermined relational formulas.