基于RBF神经网络的光伏动态重组系统

Reconfiguration approach based on RBF neural network for PV arrays system

光伏系统的模块重组对于改善系统的P-V特性有非常重要的作用。提出了一种在部分阴影遮蔽情况下基于人工神经网络算法的光伏模块重组方案。这个方案中,光伏模块分为固定部分和自由部分,并通过开关矩阵来连接。然后测量的每个自由模块以及每行固定模块的短路电流值,测量值通过人工神经网络算法得到的结果决定开关矩阵的连接方式。仿真实验显示所提出的方案的参数测量实时简易、重组策略高效,能有效地改善部分阴影状态下的光伏阵列的功率输出。最重要是所提出的控制策略能应用于较大规模光伏系统结构重组。

The reconfiguration process of PV system plays an important role in improving P-V characteristics of the system. In this paper, a new PV reconfiguration method based on artificial neural network algorithm was proposed to against the partial shading effect. In the scheme, PV module was divided into fixed and adaptive part, which were connected by switch matrix. And then, the short-circuit current of each module and each line of the fixed module were detected to determine the connection of switch matrix by using artificial neural network algorithm. It has been verified with simulation experimental results that the proposed approach have an efficient reconfiguration strategy,real-time applicability, and easy measurable parameters,thus improve the power output of the whole photovoltaic.And the most important is that the proposed control strategy can be applied to the reconfiguration of large scale photovoltaic system.