Research Of MPPT Method Based On Artificial Neural Network Algorithm

Solar energy has attracted wide attention as a clean energy,but solar power generation is greatly influenced by environmental factors.Therefore,it is the main research topic of engineering to maintain the maximum power point output of PV system and enhance the efficiency of photoelectric conversion.Based on the Radial Basis Function Network(RBFN)in Artificial Neural Network(ANN),this paper is presented an improved Maximum Power Point Tracking(MPPT)for photovoltaic system.Along the research of MPPT algorithm based on artificial neural network,this paper developed two different improvement measures from the point of increasing the speed of model building: Improved RBFN model PSO-RBFN based on particle swarm optimization(PSO)algorithm and improved RBFN model LS-RBFN based on Least Square(LS).These two methods can hardly have a negative impact on the prediction performance of the original model while improving the speed of the model construction.The main research and solving problems of this paper include the following points:1)Based on the multi junction solar cell Matlab/Simulink photovoltaic system model,PSO-RBFN model is built.Using PSO algorithm to find the optimal connection weights of the network,and improve the speed of network construction from the perspective of reducing the number of training iterations.The speed and accuracy of the presented MPPT computation under photovoltaic and temperature conditions are simulated.The results show that the MPPT based on PSO-RBFN reduces the training time of the network,effectively improves the speed of the construction of the MPPT algorithm of the PV system,and does not have a negative impact on the prediction performance.2)LS-RBFN model of applying the PV system MPPT is studied which the connection weight value in RBFN has a linear relationship with the network output,therefore and the optimal connection weight of the network can be calculated directly by the LS algorithm.From the perspective of network training process,the speed of MPPT algorithm of PV system is further improved.The speed and accuracy of MPPT computation under photovoltaic and temperature conditions are simulated.The results shows that the MPPT based on LS-RBFN can effectively improve the speed of MPPT algorithm in the PV system at the cost of 0.07% prediction accuracy.3)In the case that the laboratory testing data equipment can not directly measure the real experimental data from the actual environment,the theoretical values are obtained by using the Taylor expansion and other methods according to the theoretical formula of the photoelectric conversion.Andhe data obtained is the current value of the system at the maximum power point corresponding to the specific illumination intensity R and the ambient temperature T.The advantages and disadvantages of the presented two MPPT algorithms are compared,which shows that the MPPT algorithm based on PSO-RBFN and LS-RBFN has a good application potential if the experimental data are not very large.