2016 IEEE Congress on Evolutionary Computation, CEC 2016

Over the last two decades, many differential evolution algorithms have been introduced to solve constrained optimization problems. Due to the variability of characteristics of such problems, no single algorithm performs consistently well over all of them. In this paper, for a better coverage of the problem characteristics, we introduce an enhanced multi-operator differential evolution algorithm, which utilizes the strengths of multiple search operators at each generation, and places more emphasis on the best-performing ones during the optimization process based on three measures: (1) the quality of solutions; (2) the feasibility rate; and (3) diversity. In addition, an improved self-adaptive mechanism for automatically controlling the scaling factor and crossover rate is proposed. The performance of the algorithm is assessed using a well-known set of constrained problems, with the experimental results demonstrating that it is superior to state-of-the-art algorithms