Welded portion characteristic prediction method, prediction model generation method, and prediction model

A prediction model generation method of generating a model to predict characteristics of welded portions obtained by resistance spot welding to pass a current between two electrodes clamping three or more overlapping metal sheets including at least a first metal sheet, a second metal sheet, and a third metal sheet, the welded portions including a first welded portion between the first metal sheet and the second metal sheet and a second welded portion between the second metal sheet and the third metal sheet, the method comprising:

• acquiring electrical information outputted during resistance spot welding, the electrical information including at least one of the current flowing between the two electrodes, a voltage between the two electrodes, a resistance between the two electrodes, and an amount of heat generated between the two electrodes during the resistance spot welding;
• acquiring an analysis result of characteristics of the welded portions when the resistance spot welding is performed;
• generating a first prediction model to predict characteristics of the first welded portion and a second prediction model to predict characteristics of the second welded portion by performing machine learning based on learning data associating the electrical information with the analysis result of the characteristics of the welded portions when the resistance spot welding is performed;
• and
• generating a prediction model including the first prediction model and the second prediction model.