The Research And Application Of An Improved DE Algorithm Based On New Crossover Strategy And Social Operation

Differential evolution algorithm (Differential Evolution Algorithm, DE) is a simple and effective and robust good optimization algorithm. It’s concerned by many scholars, and made many improvements for solving the problems that it’s easy to fall into local optimal solution and the diversity of the population in the late reduce fast. Based on the in-depth research about DE, this paper made some improvements as follows.As we know, the differential evolution algorithm crossover factor is fixed, so it’s clearly inappropriate to different problems. This paper proposes an improved DE algotithm based on new crossover and secondary mutation strategies (CMDE). As to the Crossover-factor is decided by the proportion of the variance and the evolution process in each generation, so it can follow the process of evolution and constantly change; the added operation of Second Mutation can improve the capacity of solving problem, which algorithm falls into the local solution easily.Since every algorithm has own advantages, combining DE and other optimization algorithm can improve the performance of differential evolution algorithm. This paper combines the DE algorithm and the Social Emotion Operation Algorithm, proposes an improvement DE algorithm based on social emotional operations (DSE). DSE algorithm set two kind differential strategies, and will choose one of them based on the ratio of the adjacent two generations. After some evolution generations, if the optimal solution is not changed, then change the mutation strategy, differential strategy turns to social operation or social operation to differential strategy. This way can improve the algorithm performance.This paper combines the CMDE algorithm and DSE algorithm, proposed an improved DE algorithm based on new crossover strategy and social operation (CSDE). After the test of nine test functions, the results prove that CMDE algorithm, DSE algorithm and CSDE algorithm are better than DE algorithm in the convergence speed and the optimal solution. And the CSDE algorithm has the better performance than CMDE and DSE algorithm in solving the problem of much local optimal solution. At last, this paper use the CSDE algorithm to revise the para in Activated sludge wastewater treatment model, and the results show that the rectification para makes the model simulation more close to the actual water value.