Research And Implementation Of MPPT Algorithm Based On Adaptive Neural Network Control

Photovoltaic power generation system mainly converts solar energy into electricity required by the load through a photovoltaic array.Under the standard external environment,the power voltage(P-V)output curve of the PV array presents a single peak phenomenon.When the photovoltaic array is partially shaded or part of the photovoltaic cell is damaged,the P-V curve will have multiple peaks.Using the traditional maximum power tracking(MPPT)algorithm,it is easy to fall into the local optimal solution.Therefore,aiming at the above problems,this paper proposes a hybrid algorithm based on load voltage feedback and adaptive neural network control(DANC),compared with particle swarm optimization,hybrid algorithm based on differential flatness control and hybrid algorithm based on feedback load voltage and variable step increment(INC),the advantages of the proposed algorithm are highlighted.Through Matlab simulation and hardware experiments,the feasibility of the algorithm is further proved.In this paper,the structure of photovoltaic system is analyzed,and the photovoltaic cell and Boost circuit are modeled and analyzed.Under the standard external environment,the P-V characteristic output curve of the photovoltaic cell was obtained.Second,in the case of single peak,simulation analysis of disturbance observation method(P&O),INC method,fuzzy control,improved fuzzy control and DANC algorithm,highlight the advantages of the DANC algorithm.Then,the modeling and Simulation of the SP model of 3?3 the photovoltaic array,The P-V curve of the two scenarios is analyzed,which shows that the single peak algorithm is easy to fall into the local optimal solution.The PSO algorithm,the differential flat control and the P&O method are combined for simulation analysis,and a hybrid algorithm combining differential flat control and fuzzy control is proposed.Finally,the advantages and disadvantages of the PSO algorithm and the differential smooth control based hybrid algorithm are analyzed,and a hybrid algorithm based on feedback load voltage and DANC is proposed and compared with the hybrid algorithm based on the feedback load voltage and variable step length INC,the superiority of the proposed algorithm in tracking efficiency is highlighted.Through hardware experiments,the feasibility of the proposed algorithm is further verified.