The Establishment Method of Power Device Loss Model of Optical Storage Converter Based on BP Neural Network

The Optical Storage Converter (OSC) serves as a pivotal control unit within the photo voltaic energy storage system, where its primary source of heat generation arises from power device losses. Accurate prediction of these losses becomes challenging when the requisite voltage and current data of the OSC are not readily available, complicating the thermal safety estimations of the converter under various operating conditions. In this study, we propose a power device loss model for the OSC, founded on the principles of the Back Propagation (BP) neural network. A low-power optical storage system experimental platform is subsequently constructed to obtain loss data samples under differing operational conditions through manual data collection methods. Upon completion of sample training, a test set is deployed to verify the model's accuracy. The results showcase an estimation error of less than 5% for the proposed model, satisfying the accuracy requirements for the loss calculations of the OSC.