Improvement Of Differential Evolution Algorithm And Its Application

As the scale of various practical problems including the location optimization in solar power tower plant,the location of the wireless sensor,the optimal path optimizatio of the robot and the problem of the cloud computing task scheduling etc.,increases,artificial computing methods have been unable to meet the requirements of people to solve problems.In order to get rid of this dilemma,researchers began to explore the intelligent behavior of natural biological populations to find more effective methods.In this process,evolutionary computing algorithm arises at the historic moment.The principle of evolutionary algorithm is derived from the thought of "survival of the fittest,the best and the least" in the theory of evolution.It is an important part of the field of artificial intelligence and a heuristic global search intelligent algorithm.Differential evolution algorithm is an important branch of evolutionary computation.In this paper,DE algorithm is studied and improved.The improved DE algorithm is used to solve the Economic Dispatch problem in the electric field and reduce the power generation cost of the power plant.The main research contents of this design include the following four aspects:1.In view of the fact that the DE algorithm is easy to fall into the local optimum and has weak directivity in the optimization process,the Cauchy perturbation operator is proposed.The operator is executed once in each generation to make some solutions jump out of the local optimum and correct the individuals that have deviated from the optimal direction.At the same time,a new method of generating position parameters and an updated mutation strategy are provided.Experiments on CEC2014 and CEC2017 standard test suites show that the improved DE algorithm has better accuracy,and the performance of the function shows a better trend as the problem dimension increases.2.In view of the phenomenon that the population diversity of DE algorithm is reduced and the algorithm is prone to be premature in the later period of evolution,a new improvement strategy is proposed,which is called the Jrand decreasing mechanism,to enhance the diversity of the algorithm.At the same time,the Jrand number change was controlled by using the feedback guidance technique,and the fitness difference was incorporated into the position parameter of the improved algorithm.Experiments on the CEC2017 standard test suite show that the improved DE algorithm has better convergence and optimization capability.3.Another problem of DE algorithm is that it is difficult to balance the exploration and exploitation.A new strategy is proposed,which is called double mutation strategy based on inferior solution eliminating mechanism.Strategy 1 is used to improve the exploration of the algorithm,and strategy 2 is used to improve the exploitation of the algorithm.The inferior solution eliminating mechanism is proposed to further enhance the accuracy of the solution and mitigate the loss of evaluation rate due to the dual-strategy construction.Experiments on the CEC2017 standard test suite show that the improved DE algorithm has better robustness and generalization ability.4.The model of Economic Dispatch problem is established and optimized by using the improved differential evolution algorithm,which makes each unit system undertake a reasonable power load and reduces the power generation cost of the power plant.