Research And Application On Reactive Power Optimization Of Power System Based On Fruit Fly Optimization Algorithm

The distribution of reactive power in power system is reasonable or not determines the voltage quality and directly affects the safety and economic operation of the whole power grid. It is very important to study of reactive power optimization problem that is the effective means to improve the stability of power system and reduce the network loss and improve the voltage quality. Reactive power optimization of power system is a complex optimization problem with the characteristics of continuous and discrete variables mixed and multi constrained and nonlinear and high dimension. The method of general mathematics is very difficult to solve reactive power optimization of power system that need to select the appropriate optimization algorithm for solving. We will research the mathematical model of reactive power optimization and related algorithms in this thesis. Through the comprehensive analysis of the characteristics of the traditional optimization algorithms and artificial intelligence algorithms, we will choose the emerging fruit fly optimization algorithm as the research object.The basic theoretical knowledge of the fruit fly optimization algorithm was studied deeply in the thesis, according to the disadvantage of the scope of the search and mutation probability is low, the population diversity is poor, the optimal solution may miss, the convergence precision is not high and easy to fall into local optimum, a new multi-swarm fruit fly optimization algorithm is proposed to improve the algorithm convergence precision and the ability to jump out of the local optimal solution through strategies such as the use of a variety of groups, the revised evaluation function, the use of local search method and dynamic contraction of the search radius. By comparing and analyzing the simulation results of several classica l test functions of different optimization algorithm, show that the multi-swarm fruit fly optimization algorithm is effective and feasible. Then the improved multi-swarm fruit fly optimization algorithm is applied to solve the problem of reactive power optimization of power system. This thesis analyzes the basic idea and calculation steps and process of reactive power optimization problem in power system based on the improved multi-swarm fruit fly optimization algorithm, detailing some problems of solution space coding, parameter setting and the criterion of the end. Writing the reactive power optimization program of power system based on improved multi-swarm fruit fly optimization algorithm and combined with Newton-Ravson power flow calculation in the Matpower kit in Matlabsoftware platform and applied to the IEEE-30 node test system. The reactive power optimization results were compared with the standard particle swarm optimization algorithm and the adaptive differential algorithm show that the improved multi-swarm fruit fly optimization algorithm can significantly reduce the network loss, improve voltage quality, improve the convergence speed and the global searching capability of the algorithm, and the convergence precision is higher. Experiments prove that application of the multi-swarm fruit fly optimization algorithm to solve power reactive optimization of power system has good theoretical and practical value.Finally, this thesis will use MATLAB GUI to write a computing platform for reactive power optimization based on multi-swarm fruit fly optimization algorithm. It can quickly realize the reactive power optimization calculation of actual system and parameter configuration and data storage and other functions. Users are more convenient and intuitive at the time of reactive power optimization operation. It can provide important guarantee for the safe and stable operation of power system and will has a broad application prospect.