基于光伏并网逆变器的一种矢量角补偿法有限控制集模型预测控制研究

A Vector Angle Compensation Method Study of Finite Control Set Model Predictive Control Based on Photovoltaic(PV) Grid Inverter

有限控制集模型预测控制（finite control set model predictive control,FCS-MPC）提供了一种与众不同的能量处理方法,将功率变换器当做离散和非线性的执行器。它具有无须解耦运算和安装脉冲宽度调制（pulse width modulation,PWM）调制器等优点,被广泛应用于电力电子变换器。传统FCS-MPC在延时补偿问题上采用拉格朗日外推法对未来参考值进行修正,但这种方法在出现阶跃变化时峰值较大。考虑到三相系统变量的矢量表示法后,能够根据一次采样时间的矢量角变化来估算未来参考值。因此,该文基于光伏并网逆变器,运用矢量角补偿法对FCS-MPC延时补偿后带来的未来参考值偏离进行修正,研究光伏光照强度变化时并网的动态控制效果,这个简单的补偿方法允许预测控制模型包含延时,并能够避免受控变量出现较大的纹波,同时该文在代价函数中加入了开关频率约束项,减少开关损耗,同时不会造成输出电流畸变。最后通过Matlab/Simulink仿真和实验,验证这种模型预测控制器在光照波动时并网运行过程中具有良好的控制性能,并论证此方法的可行性。

Finite control set model predictive control（FCS-MPC）, a unique approach to deal with energy, views the power converter as discrete and nonlinear actuators. It has no decoupling arithmetic and pulse width modulation（PWM） modulator installation, and is widely used in power electronic converter. Traditional FCS-MPC uses Lagrange extrapolation method to modify the future reference on delay compensation problems. But the peak is bigger in the event of the step change. After considering the vector representation of the three-phase system variables, which can be based on the change a sampling time vector Angle to estimate future reference value. This paper presents the dynamic control effect of changing light intensity by using vector angle compensation method of FCS-MPC to compensate the external reference value deviation based on photovoltaic（PV） grid inverter. this simple compensation method allows model predictive control to contain time delay and can avoid larger ripple of the controlled variables. Meanwhile, the cost function with the switching frequency constraints can reduce the switching loss but not result in output current distortion. At the end of this paper, the model predictive controller is verified by Matlab/Simulink and experiments in which good control performance and feasibility are proved in the process of light change.