基于神经网络PSD算法的GaN器件变流器控制系统设计

Design of GaN Device Converter Control System Based on Neural Network PSD Algorithm

伴随电力电子技术在环保节能、智能电网中的应用,高压大功率的电力电子设备得到大力推广。GaN器件变流器因其特殊的控制性能变成大功率、高压化控制的核心。设计了一种基于神经网络PSD算法的GaN器件变流器控制系统。当GaN器件变流器出现异常后,系统中故障信息采集模块能够高效采集GaN器件变流器异常信息并传输至自适应控制模块,自适应控制模块采用神经网络PSD算法,实现GaN器件变流器自适应控制。仿真结果表明:所设计系统对GaN器件变流器的功率、直流电流、环流以及电容电压可实现高精度控制,且该系统控制下交流器输出电压质量佳,相间短路、两相接地短路故障率得以降低,运行效率高。

With the application of the power electronics technology in environmental protection,energy saving and smart grid,the high-voltage and high-power power electronics equipment has been vigorously promoted.The GaN device converter becomes the core of high-power and high-voltage control owing to its own special control performance.In this paper,a kind of control system of GaN device converter based on the neural network PSD algorithm is designed.When the GaN device converter is abnormal,the fault information acquisition module in the system can efficiently collect the abnormal information of the converter and transmit it to the adaptive control module.The adaptive control module uses the neural network PSD algorithm to realize the adaptive control of the GaN device converter.Simulation results show that the power,DC current,circulating current and capacitor voltage of the GaN device converter can be controlled with a high precision.Under the control of the system,the quality of output voltage from the converter is satisfying,the fault rates of phase-to-phase short circuit and two-phase grounding short circuit are reduced,and the operation efficiency is high.