Research On The Performance Of Differential Evolution Algorithm

Differential Evolution(DE)algorithm is a stochastic population search algorithm proposed by Storn and Price in 1995 to solve nonlinear,high and complex optimization problems.Since it was put forward,it has been widely recognized and successfully applied in many fields of science.However,for specific optimization problems,DE algorithm also shows some deficiencies,such as slow convergence rate,sensitive to the setting of control parameters and so on.On the basis of the existing research,this paper proposes a new improved algorithm to overcome the shortcomings of the differential evolution algorithm,and verifies the effectiveness of the algorithm.The main research work of this paper are as follows:1.This paper introduces the differential Evolutionary algorithm through the brief introduction of Evolutionary algorithm,then briefly describes its research background and significance,enumerates the advantages and disadvantages of DE algorithm,and summarizes the application fields of DE algorithm and the research status at home and abroad by sorting out the literature.2.The evolution process of DE algorithm is introduced in detail,and its improvement methods and application fields are classified and summarized.It is described in six ways:initialization technology,adjustment of control parameters,mutation strategy,crossover strategy,population structure and hybrid algorithm.3.In this paper,we improve the DE algorithm from evolution mode and mutation strategy,and then a differential evolutionary algorithm based on piecewise evolution is proposed.After the initial population is initialized,the population is divided into three evolutionary stages according to the maximum evolution algebra.In each stage,three mutation strategies "DE/rand/1","DE/assemblable-to-pbest/1",and "DE/best/1" are used.Among them,"DE/assemble-to-pbest/1" is a new mutation strategy.Its base vector is a linear combination of the best and the worst individuals in the current population.It can make full use of the information of the best and the worst individuals in the population to better guide evolution.Tests were carried out on a set of 30 continuous single-objective optimization problems in different dimensions.The results show that the proposed algorithm has strong competitiveness in solving single-objective optimization problems.4.The performance of the proposed algorithm is evaluated in terms of convergence.The new algorithm and several advanced improved differential evolution algorithms are tested on the different dimensions of 30 test functions of CEC2014,and the convergence curves of the proposed algorithm and the comparison algorithm are drawn on the test platform of MATLAB R2014b.The experimental results show that the new algorithm has obvious advantages in both the convergence speed and convergence accuracy.