Application Research Of Improved Differential Evolution Algorithm In Power System Economic Dispatch

Differential evolution algorithm is a new type of intelligent algorithm.With the advantages of simple principle and few operating parameters,it is widely used to solve various optimization problems.In recent years,some scholars have also applied differential evolution algorithm to the economic dispatch of power system.Coming,but with the continuous improvement of the economic dispatch model and the increase in the scale of the power system,the economic dispatch problem has become a nonlinear,non-convex,and high-dimensional problem,leading to the traditional differential evolution algorithm when solving such problems There will be situations where the precision of the optimization results is low,and it is easy to fall into the local optimum and the convergence speed is slow.Based on this,this paper proposes two improved differential evolution algorithms to solve the ED problem.First,the basic principles of the DE algorithm are studied,the four operation procedures of the DE algorithm are introduced,and the influence and advantages and disadvantages of the DE algorithm parameters are analyzed.The static economic dispatch model of the power system,which takes the minimum fuel cost of the system as the dispatching target,and comprehensively considers the the valve-point effects,transmission network losses,operation prohibited zones and climbing limit,makes the model more consistent the actual operating state of the power system.Secondly,a differential evolution algorithm based on multi-population(MPDE)is proposed.In order to solve the problem of poor accuracy of DE algorithm and easy to fall into local optimization,MPDE algorithm uses a variety of co-evolutionary methods.In multiple populations,each population has its own mutation strategy and optimization parameters to enhance search capabilities.Moreover,a learning communication strategy between populations is designed.Individuals of a single population can not only receive information from their own populations,but also learn from individuals in other populations.This strategy promotes information exchange between populations.In addition,the algorithm also introduces a normal distribution function to dynamically adjust the scaling factor and crossover rate to increase the diversity of parameters.The MPDE algorithm was applied to the ED problem and tested on four test systems of 13,40,80 and 140 units.Simulation results show that compared with other intelligent algorithms,the improved algorithm has higher accuracy and stability in solving economic scheduling problems.Finally,an adaptive accelerated convergence DE algorithm is proposed(AMPDE).The MPDE algorithm enhances the individual diversity of the algorithm,which solves the situation that the DE algorithm is prone to fall into local optimality,but it also affects the convergence of the algorithm and causes the algorithm to slow down.In response to this shortcoming,on the basis of MPDE,this paper introduces correlation analysis between individuals,and adopts an adaptive adjustment mechanism for individuals far away from the optimal body to make it absorb the information of the current optimal individual and evolve towards a better region.The blindness of mutation is reduced,and the convergence speed and local exploration ability of the MPDE algorithm are enhanced.In order to test the comprehensive performance of AMPDE,it was tested on six benchmark functions and four test systems of 13,40,80 and 140 units,and compared with the MPDE algorithm.The simulation results show that the improved algorithm is solving the benchmark function Compared with the economic scheduling problem,the MPDE algorithm has an efficient convergence speed.