Research And Optimization Of The MPPT Control Algorithm On Photovoltaic Power Generation

Now,Solar photovoltaic power generation technology has two major problems: One is the high cost of photovoltaic cells;Two is the low conversion efficiency of photoelectric.The two aspects of the problem are greatly hindered the development of photovoltaic industry,One way to solve this problem is that to use the maximum power point tracking technology.At present,the common MPPT algorithm mainly included the constant voltage method,the conductance increment method,the perturbation and observation method,the particle swarm algorithm,the neural network method,the fuzzy control method and so on.With the improvement of intelligent control,fuzzy control algorithm and neural network algorithm has showed its advantages,but their shortcomings are also exposed,The former is lack of the ability of online self-learning,and the latter is unable to express by the natural language of human beings.In view of the above problems,this paper proposes an improved neuro fuzzy control algorithm,which combines fuzzy control and neural network method,which to overcome their own shortcomings and improve the work efficiency.In this paper,we use Matlab/Simulink software to construct the engineering model of photovoltaic cells,And the simulation experiment is done under different external conditions,The nonlinear characteristics of the photovoltaic cells are obtained,And easily influenced by the environment.The several kinds of DC/DC transform circuit is analyzed and compared,and the advantages of BOOST converter are obtained,So in this paper,the maximum power point tracking of PV system is realized by BOOST transform circuit.In this paper,the implementation principle of the traditional MPPT algorithms are detailed research.And their respective advantages and disadvantages are given.The design procedure of the improved neuro fuzzy control algorithm is introduced in detail,Set up the entire pv system simulation module(PV?MPPT ? DC/DC and PWM),And it was simulated by MATLAB/Simulink software.Also compared with curve of the other algorithms,The simulation results show that the proposed algorithm have advantages of feasible,fast and high accuracy,And the algorithm can avoid the problem of the tracking speed and the steady state accuracy of the traditional method.