基于模糊聚类和自适应神经模糊推理的MPPT研究

摘要

For the deficiency of conductance increment and fuzzy control methods for the maximum power point tracking (MPPT) of photovoltaic system,a novel MPPT algorithm is proposed in combination with fuzzy clustering and adaptive neuro⁃fuzzy inference system (ANFIS). The algorithm extracts fuzzy control rules and confirms membership function by using fuzzy clustering,and then a fuzzy inference system is generated by means of train of neural network. The simulation results show that the algorithm can realize fast MPPT,and has strong anti⁃jamming capability against environmental changes,good control effect and a certain practical value.%针对光伏系统最大功率点跟踪过程中电导增量法和模糊控制法存在的不足，结合模糊聚类和自适应神经模糊推理系统提出一种新型MPPT算法，对实测数据模糊聚类后，提取模糊控制规则，确定隶属函数，通过神经网络训练生成模糊推理系统。仿真结果表明，该算法能够实现对最大功率点的快速跟踪，对环境变化的抗干扰能力强，控制效果良好，具有一定的应用价值。