Title :
An improved adaptive differential evolution based on double populations for constrained optimization problems
Author :
Li, Meiyi ; Qiu, Qianqian ; He, Cheng
Author_Institution :
Coll. of Inf. Eng., Xiangtan Univ., Xiangtan, China
Abstract :
This paper presents an improved adaptive differential evolution algorithm based on double populations (IADE) employing control parameter CR of the differential evolution algorithm and infeasible solutions of the population to solve constrained optimization problems. The proposed algorithm can dynamically adjust CR by the individual fitness value of the population during evolution process. Using information of infeasible solutions to reduce solution space it is effective to avoid falling into local optimum and find the optimal solution quickly. It adopted searching mechanism based on double populations which have the advantage of avoiding constructing penalty function and deleting infeasible solutions directly. The algorithm shows outstanding performance on widely used eight Benchmark problems.
Keywords :
evolutionary computation; search problems; adaptive differential evolution algorithm; benchmark problem; constrained optimization problem; control parameter; double population; evolution process; searching mechanism; Benchmark testing; Educational institutions; Evolutionary computation; Heuristic algorithms; Optimization; Standards; Vectors; adaptive; constrained optimization; differential evolution; local optimum;
Conference_Titel :
Natural Computation (ICNC), 2012 Eighth International Conference on
Conference_Location :
Chongqing
Print_ISBN :
978-1-4577-2130-4
DOI :
10.1109/ICNC.2012.6234644