Differential Evolution with Alternating Strategies : A Novel Algorithm for Numeric Function Optimization

The Differential Evolution (DE) is a prominent meta-heuristic algorithm that has been successfully employed to numerous complex and diverse problems from the fields of mathematics, science and engineering. DE belongs to the evolutionary family of algorithms which is based on the Darwinian theory of natural selection and evolution. DE maintains a population of candidate solutions and uses the vector differences between randomly picked candidate solution vectors to produce new, improved solutions to advance its evolutionary optimization process, generation by generation. This paper introduces a novel DE-variant — the DE with Alternating Strategies (DE-AS) and evaluates its performance using a number of benchmark problems on numeric function optimization. DE-AS effectively combines the exploitative and explorative characteristics of five different DE-variants by randomly alternating and executing these DE-variants in a single algorithm. The experimental results indicate that DE-AS can perform better than many other existing DE-variants on most of the benchmark functions, in terms of both final solution quality and convergence speed.