Research On Real-Time Simulation Modeling And Optimization Methods Of Power Electronic Converter Based On FPGA

The real-time simulator based on Field Programmable Gate Array（FPGA）has the characteristics of high reliability,small simulation step size and strong parallel computing ability,and has been widely used in the development and design of power electronic converters.In recent years,with the rapid development of the power field,power electronic converters have shown an application trend of modular and high-frequency.A large number of switching devices in the system and excessive switching frequencies increase the difficulty of real-time simulation applications.When establishing a system model,the traditional real-time simulation modeling method cannot meet the simulation accuracy requirements,there are large errors,or occupy too many FPGA hardware resources,which affects the model expansion,so it is difficult to take into account the simulation accuracy and simulation efficiency at the same time.Therefore,how to accurately and efficiently establish the mathematical model of the power electronic converter is a key issue for real-time simulation applications.Aiming at the traditional real-time simulation modeling methods of power electronic converters—switching function method and R_ON/R_OFF modeling method,this paper analyzes the shortcomings of the two modeling methods,and proposes corresponding algorithm optimization measures to make them more satisfactory to real-time simulation application requirements of real-time simulation of power electronic converters,the main work completed in this paper is as follows::（1）Aiming at the problem of high-frequency voltage oscillation when the converter pulse is turned off in the traditional switching function modeling method,combined with the simulation waveform to analyze the source of the oscillation,a zero-crossing detection optimization algorithm is proposed to be added to the traditional method,and the improved switching function model is applied to the three-phase inverter and the Neutral Point Clamped（NPC）inverter,the simulation results demonstrate the generality and accuracy of the proposed optimization method.（2）Aiming at the problem of resource consumption caused by iterative operation in the traditional R_ON/R_OFF modeling method,this paper first analyzes the influence of iterative calculation on the system model.The more iterations of the model,the more accurate the waveform,but the greater the resource consumption.This paper judges the state of the switching device based on the switching function model and the zero-crossing detection algorithm,and cancels the iterative operation in the R_ON/R_OFF modeling method.The simulation results prove the correctness of the method used in this paper.When the model is implemented in FPGA,the resource analysis results prove that this method can effectively improve resource consumption.（3）The paper completes the real-time simulation experiment verification of the permanent magnet synchronous motor traction drive system.Firstly,the working principle and the stability of the semi-implicit delay decoupling method is introduced.Then,the semi-implicit delay decoupling method is used to split the traction drive system into five subsystems,and the mathematics of each subsystem is established respectively.The model is compared with the reference waveform,which proves the accuracy of the model.Then,the components of the dSPACE real-time simulation system are introduced,and the experimental process of the real-time simulation is described;the real-time simulation experiment of the permanent magnet synchronous motor traction drive system is completed,and the experimental waveform and resource analysis results prove that the optimization model described in this paper can reduce the FPGA resource consumption and has good engineering application prospects.