Research On APF Control Strategy Of Micro-grid Based On Improved Artificial Bee Colony Optimization BP Neural Network

With the construction of "smart grid" project,micro-grid technology came into being.The traditional energy depletion,deterioration of ecological environment and promote the development of the construction of the micro grid,with a large number of micro network equipment put into operation,the harmonic current produced by the injection of public power,resulting in varying degrees of pollution to the public power grid building.This paper discusses the basic principle and structure of APF,the micro grid harmonic control needs,proposed micro-grid parallel APF topology based on the principle of p-q detection method and ip-iq detection method based on instantaneous reactive power theory.Because the ip-iq detection method is the improvement of p-q detection method,this paper adopts the ip-iq detection method with higher detection accuracy and stronger real-time.In view of the uncertainty of parameter selection in traditional PI control,the optimal values of parameters can not be obtained in real time and can not be changed after the selection of parameters.BP neural network is used to train and optimize the ratio and integral parameters,and then the feasible and effective values are obtained.According to the BP neural network in training,the network convergence speed is slow and easy to fall into local minimum,may lead to the completion of training and other issues,put forward the improved artificial bee colony optimization BP neural network control strategy,through the optimization of the BP neural network,parameter adjustment of the network constantly,in order to obtain higher response speed.In order to verify the feasibility of the proposed control method,an experimental platform is built,the structure and function of each part of the system hardware and software are described in detail,the data show that the optimized control method overcomes the defects of the original network slow convergence rate,and improve the control accuracy.