电流型LCL逆变器矢量模型预测

Vector Model Prediction of Current Source LCL Inverter

为弥补LCL并网逆变器传统无源控制存在的参数整定困难、功率损耗过高、控制不精确等缺陷,提出一种电流型LCL并网逆变器矢量模型预测控制方法.以LCL逆变器输出电流为控制目标,根据系统预测参考值构建代价函数,并根据代价函数最小原则选取1~3个最优电压矢量作用于1个采样周期.针对LCL型逆变器单矢量、双矢量以及三矢量模型预测控制进行分析及仿真试验.仿真结果显示:随着周期内作用电压矢量数量的增加,系统电流纹波减小,控制精度及动、静态性能得到提升.

Because the traditional passive control of LCL grid connected inverter has the defects of difficult parameter setting,high power loss and inaccurate control,a vector model predictive control method of current source LCL grid connected inverter is proposed.This method takes the output current of LCL inverter as the control target,constructs the cost function according to the system prediction reference value,and selects 1~3 optimal voltage vectors to act on a sampling period according to the principle of minimum cost function.The single vector,double vector and three vector model predictive control of LCL inverter are studied and simulated;The simulation and experimental results concluded that the current ripple of the system decreases with the increase of the number of voltage vectors in the cycle,and the control accuracy and dynamic and static performance are improved.