基于改进粒子群区间二型模糊神经网络的MPPT控制研究

MPPT CONTROL BASED ON IMPROVED PARTICLE SWARM INTERVAL

针对太阳能发电单元最大功率点控制(MPPT)在复杂工况条件下存在的振荡、跟踪耗时长、精度较低的问题,提出一种基于改进区间二型模糊神经网络的预测控制模型。首先将减法聚类与区间二型模糊均值聚类算法相结合,辨识模型前件模糊规则层结构,计算得到聚类中心;其次,基于自导式粒子群算法优化后件权重层权值参数,进而提升网络全局寻优能力;最后,通过与TS模糊神经网络模型、基于反向传播算法的区间二型模糊神经网络模型进行仿真对比,验证所提模型在不同工况下对最大功率点追踪的快速性与精确性。

In order to solve the problems of maximum power point control,propose a predictive control model based on interval type two fuzzy neural network,such as oscillation,long tracking time and low accuracy under complex operating conditions.Firstly,the Fuzzy rule layer structure of interval type two fuzzy neural network is identified and the cluster center is calculated by combining subtractive clustering and interval two type fuzzy mean clustering algorithm;Secondly,self guided particle swarm optimization is used to optimize the weight layer of the subsequent layer to improve the global optimization capability of the network.Finally,through simulation comparison with TS fuzzy neural network model and interval typeⅡfuzzy neural network model based on back propagation algorithm,the rapidity and accuracy of the proposed model for maximum power point tracking under different working conditions are verified.