Research On Maximum Power Point Tracking And Energy Storage Monitor Of Partially Shaded Photovoltaic System

With the development of social productivity, the social demand for energy isgrowing, and the energy crisis is increasing with each passing day. In renewableenergy, solar energy with its ample storage, environmental protection and otherfeatures, it has been the most potential development energy. But the photovoltaic powergeneration system exists major problems, firstly, the PV curve showing a multi-peakscurve in partially shaded environments, the traditional method of maximum power pointtracking (Maximum Power Point Tracking-MPPT), easy to fall into local optimummistakes, tracking result was low in accuracy and slow in convergence, secondly,because of the independent photovoltaic power generation system in which theenvironment is very bad, so the storage part state influencing factors, such asillumination,temperature, they are easy to change, so the traditional storage partmonitoring methods is big in error and poor in real-time performance.For MPPT issues, the paper proposed an AFSA optimized by extended PSOalgorithm(EP-FSA), firstly, it extend the standard PSO by add a memorial part, so thePSO algorithm not only has self-learning ability, but also has a memorial capacity,secondly, it using the extended PSO to optimize AFSA,make the AFSA iterativeprocess to two steps, and the algorithm introduced the speed parameters, finally itimitated the PSO, the AFSA added two kinds of movement patterns, namely cognitivebehavioral and social behavior.For storage monitoring issues, this paper proposed an extended Kalman filtermethod, firstly it established the battery circuit model, secondly it made the nonlinearpart linearized, finally the method used Kalman filtering to monitor.In this paper, the results of simulation comparison experiments for different MPPTalgorithm by Matlab showed that the EP-FSA compared P&O and PSO,its trackingresult was higher in tracking accuracy, more stable, and the ability for excluding thelocal optimum mistakes stronger. In addition, this results of simulation comparisonexperiments for different monitoring methods showed that the proposed method washigher in tracking accuracy, faster in tracking speed smaller in error and more stable inwaveform by compared with the standard Kalman filtering method, it is reasonable andfeasile method.