Research On The Cooperative Co-evolutionary Algorithm For Large-scale Optimization Problem And Multi-objective Optimization Problem

Problems occurred in the fields of science,engineering,business and management are complicated.They usually contain a large number of decision variables,constraints or optimization goals that difficult for traditional optimization algorithms to solve.Co-evolutionary algorithm is proved to be a fast and effective algorithm to solve these complex problems,and has been widely used.In this paper,we will introduce the principle and research status of co-evolutionary algorithm,and enhance its performance by mixing various strategies,improving the search performance and so on to solve the problems we mentioned before.The main contents are as follows:1)This paper introduces the definition and research status of the large-scale optimization and multi-objective optimization problems.At the same time,we analyzes the difficulties and challenges of the existing optimization algorithms to solve these two kinds of optimization problems.2)We do a research on the development and status of cooperative co-evolutionary(CCEA)to introduce the framework and principle as well as the summary of the main improvement on it.This work provides a theoretical basis and ideas for improving the algorithm.3)CCEA for large-scale optimization problems.A hybrid differential evolution algorithm is proposed to improve the global search and local search capability of CCEA.First use the differential evolution algorithm and simulated annealing algorithm together to optimize the population,then use the local search chain for accurate research.This optimization mechanism will balance the global and local search ability,also overcome the premature convergence effectively.In this paper,we carry out the experiments with a series of representative large-scale test functions to verify the necessity of the components in the improved algorithm.The result shows that the algorithm we proposed is very competitive in solving large-scale optimization problems,comparing with a number of commonly used optimization algorithms.4)CCEA for multi-objective optimization problem.To improve the performance of the algorithm,we use two evolutionary strategies together to optimize the target population,design new Pareto solution selecting algorithm,and also introduce in the concept of elite strategy and local search algorithm into the algorithm we proposed.Compared with the results produced by other classical multi-objective optimization algorithm,the Pareto solution set obtained in this paper has a better convergence and distribution performance.