基于PI与神经网络准PR双闭环控制的光伏并网控制研究

Photovoltaic Grid-connected Control Research Based on PI and Neural Network Quasi-PR Double Closed-loop Control

传统的单相并网逆变器采用单一的比例积分(PI)控制器进行控制,虽然结构简单,但是以不变的模式及参数无法实现无静差跟踪,且并网容易产生波动,谐振抑制能力不足。基于此,提出了一种基于PI与神经网络准PR双闭环控制的单相并网逆变器系统控制策略。设计了PI与神经网络准PR双闭环控制的控制策略及控制框图,实现入网电流无静差跟踪,与PI与准PR双闭环控制策略进行比较,具有更好的并网电流波形,电流总谐波畸变率有明显下降。最后,通过MATLAB/Simulink平台搭建了两种并网控制策略的仿真模型,验证了本文提到的控制策略效果更佳。

The traditional single-phase grid-connected inverter adopts a single proportional integration (PI) controller for control,the structure is simple,but the unchanged mode and parameters can not achieve tracking without static error,and the grid connection is easy to produce fluctuations,resonance suppression ability is insufficient.Based on this,a single-phase grid-connected inverter system control strategy based on PI and neural network quasi-PR double closed-loop control is proposed.The control strategy and control block diagram for the PI and neural network quasi-PR double closed-loop control are designed to achieve tracking without static error of incoming current,compared with PI and quasi-PR double closed-loop control strategy,which has a better grid-connected current waveform,and the total harmonic distortion rate of the current is significantly reduced.Finally,a simulation model of two grid-connected control strategies is built through the MATLAB/Simulink platform,which verifies that the control strategies mentioned in this paper have better effect.