基于PSO优化BPNN估计光伏阵列MPPT控制系统研究

MPPT control system of photovoltaic array based on PSO-BPNN estimation

为了充分利用光伏阵列转换能量,提高光伏阵列的发电效率,在分析光伏阵列的伏安特性及最大功率点跟踪(MPPT)原理的基础上,提出了一种基于粒子群算法优化BP神经网络(PSO-BPNN)的建模方法,并用这种改进的神经网络构建了光伏阵列的动态模型。通过PSO-BPNN模型拟合光伏阵列输出功率与输出电压的非线性关系,实现了对光伏阵列的最大功率点跟踪。Matlab/Simulink仿真及在线测试结果表明:基于PSO-BPNN估计的光伏阵列MPPT控制系统能快速、精确地跟踪光伏阵列的最大功率点,改善了BP神经网络收敛速度慢,易陷入局部极值,建模精度不高的缺点,提高了系统的稳定性和能量转换效率,是研究光伏发电这个复杂非线性系统的一个可行办法。

In order to make full use of PV to transform energy and improve the generating efficiency of photovoltaic array, a modeling method based on PSO-BPNN was raised based on the analysis of photovoltaic array＇s volt-ampere characteristics and maximum power point tracking （MPPT）. Consequently, the dynamic model of the photovoltaic array was constructed with the improved neural network and the maximum power point tracking for the photovoltaic array was realized through PSO-BPNN model fitting the nonlinear relation between output power and output voltage of PV array. Matlab/Simulink simulation and on-line test results indicate that the photovoltaic array MPPT control system based PSO-BPNN can track photovoltaic array＇s maximum power point rapidly and accurately, which makes BP neural network＇s convergence speed quicker and fall into the local extremum easily. Above all, the modeling precision, system stability and energy conversion efficiency can be improved. So, MPPT control system of the photovoltaic array based PSO-BPNN is a feasible way of studying photovoltaic power generation.