Hybrid Data-Mechanism Modeling for Power Loss of a Boost Converter

Abstract

Parameter design is the premise to ensure the optimal performance of power electronic converters, which depends on the accuracy of the established model. However, there is still many issues for improvements in modeling methods for the parameters design of power electronic converters. Due to the lack of key mechanism and the uncertainty of parameters, the existing modeling method shows poor accuracy or efficiency. In order to simplify the process and improve the accuracy of the existing modeling method, this paper proposes a hybrid data-mechanism modeling (HDMM) method for power loss of a typical power electronic converter, Boost converter, which is with UC3842 control chip. Based on the main loss from inductor, switch tube, diode and sampling resistance, the mechanism power loss modeling of Boost converter is built. Besides, the artificial neural network (ANN) is used to learn from the sampled data and train the data-driven model. To improve the accuracy of the mechanism model, the HDMM for power loss of Boost converter is established by modification of data-driven model with limited data. Through the proposed HDMM method, the mechanism modeling method and data-driven method are well combined and the HDMM has the advantage of high accuracy and strong generalization. In the end of this paper, feasibility and accuracy of the proposed HDMM method have been verified by hardware experiments.