Predicting the impact depolarization behavior of PZT-5H based on machine learning

In order to establish the relationship between the load and depolarized charge under stress waves, we established the stress depolarization experimental platform and obtained the time history curve of charge variation with the load. Then, we built a theoretical model of PZT-5H impact depolarization and physically described the model based on the characteristics of the experimental data and domain switching. Meanwhile, we obtained the uncertain parameters in the model by using the CBP neural network and the genetic algorithm. Finally, the effectiveness of the impact depolarization theory model was verified by comparing it with experimental values. The results indicated that the established PZT impact depolarization model can reflect the depolarization law of materials in the range of 0 similar to 250 MPa stress waves. It provided a predictive method for the application of piezoelectric materials in extreme engineering such as contact fuses, pulse power supplies, and impact sensors.