Research On Differential Evolution-based Intelligent Optimization Algorithms

Single objective optimization,multi-objective optimization and constrained optimization problems are ubiquitous in mathematics and engineering and have become more and more complex.Evolutionary computation is an effective method to solve such problems.In recent years,Differential Evolution(DE)has attracted much attention in the field of evolutionary computation.Differential Evolution algorithm is a population-based stochastic optimization technique,which has the advantages of simple structure,easy implementation and strong robustness.DE algorithm has been widely used in many fields.This dissertation aims at the deficiency of DE for single objective optimization,multi-objective optimization and constrained optimization,and proposes three improved DE algorithms respectively.The main contributions are as follows:1.For single objective optimization,in order to solve the drawbacks of premature and slow convergence in multi-population ensemble DE(MPEDE),an improved multi-population ensemble DE(IMPEDE)is proposed in this dissertation.IMPEDE proposes a new mutation strategy “DE/pbad-to-pbest/1” instead of the mutation strategy “DE/rand/1” in MPEDE,and the new strategy utilizes not only the good solution information(pbest),but also the information of the bad solution(pbad)toward the good solution to balance exploration and exploitation.Furthermore,IMPEDE employs the improved parameter adaptation approach to avoid premature convergence of the “DE/current-to-pbest/1” strategy by adding the weighted Lehmer mean strategy.Experiments have been conducted with CEC2005 and CEC2017 benchmark functions,and the results have demonstrated that IMPEDE outperforms than that of MPEDE and the other four recent proposed DE methods in obtaining the global optimum and accelerating the convergence speed.2.For multi-objective optimization,the balance of convergence and diversity for the current multi-objective DE algorithm is hard to maintain.This dissertation presents a multiobjective differential evolution algorithm based on decomposition and multi-strategy mutation(MODE/DMSM).MODE/DMSM utilizes the improved Tchebycheff decomposition approach to decompose a multiobjective optimization problem into multiple single-objective optimization problems.Moreover,MODE/DMSM adopts the efficient non-dominated sorting approach to select solutions which have both good convergence and diversity to guide the differential evolutionary process.Eventually,MODE/DMSM employs the multi-strategy mutation approach to balance the convergence and diversity in the evolutionary process.The results of simulations on 10 test functions of ZDT and DTLZ show that MODE/DMSM outperforms than the other six representative multiobjective optimization algorithms in terms of the good convergence and diversity of the Pareto optimal set.3.For constrained optimization,the tradition DE algorithm is inefficient in constraint processing and the ability of searching the optimal solution.This dissertation proposes multi-strategy mutation constrained differential evolution algorithm based on replacement and restart mechanism(MCODE)which utilizes the multi-strategy mutation and the replacement and restart mechanism to improve the exploration and exploitation and achieves the goal of balancing constraint and objective function.The simulation experiment on CEC2010 test shows that the proposed algorithm has the merits in searching for feasible regions and finding the optimal solution.