Research On Photovoltaic MPPT Control Based On Improved Particle Swarm Optimization Neural Network

With the development of the whole society,fossil fuels are still the main energy source to meet the global energy demand.However,due to the limited supply and pollution feature of fossil fuels,such as solar energy and wind energy which can be substitute are more and more popular because of their free and non-pollution.Among them,solar energy is widely concerned and developed because of its rich reserves and no geographical restriction.However,in practical application,photoelectric conversion efficiency has always been one of the inevitable important defects,and the maximum power point tracking(MPPT)technology of photovoltaic power generation system has become an effective method to improve photoelectric conversion efficiency.Based on this,this article focuses on maximum power point tracking and photovoltaic cell simulation modeling,and uses the characteristics of high adaptability and strong fitting ability of artificial neural network to better realize the control of electrical characteristics and maximum power point tracking of photovoltaic system.The main research work of this paper is as follows:Firstly,a neural network photovoltaic MPPT algorithm based on improved particle swarm optimization is proposed.Through the nonlinear dynamic optimization of the inertia weight value in the particle swarm optimization algorithm,and using the improved particle swarm optimization algorithm to optimize the connection weight of the neural network,the improved neural network MPPT algorithm can effectively avoid falling into local optimization.The simulation results show that under the same experimental setting,compared with the existing traditional MPPT and artificial intelligence MPPT methods,the proposed control method increases the tracking accuracy of the maximum power point,makes the global tracking speed faster and improves the stability of tracking control.Secondly,facing the serious power loss caused by the sudden change of the illumination area received by the module under the actual photovoltaic array connection,through the research on the characteristics of the bypass diode in the photovoltaic cell,two connection modes,bridge connection and cross connection are proposed.Based on the actual module connection model,the input structure of neural network algorithm is improved.Through simulation modeling,the output parameters are verified by corresponding data indicators.The results show that under the same experimental conditions,the output efficiency of the improved neural network in the cross connection mode is more than 50% higher than that in the traditional serial connection mode.In order to verify the feasibility of the above theory in practice,a photovoltaic MPPT controller is designed independently.Through the hardware circuit construction and software design,the output power is compared with that under the non photovoltaic MPPT control mode.The results show that the designed photovoltaic MPPT controller can output higher power value.Compared with the non MPPT control mode,its output efficiency is increased by more than 25%.The photoelectric conversion efficiency of photovoltaic cells is effectively improved,and the practicability and feasibility of the algorithm of photovoltaic MPPT controller designed in this paper are verifiedTo sum up,the improved neural network algorithm proposed in this paper and the controller designed in this paper realize the advantage of tracking the maximum power under different environmental conditions.They have great significance in improving the photoelectric conversion efficiency in the photovoltaic field.