Meta-heuristic and Evolutionary Algorithms for Engineering Optimization

9 Differential Evolution

Summary

This chapter describes differential evolution (DE), which is a parallel direct search method that takes advantage of some features of evolutionary algorithms (EAs). The DE is a simple yet powerful meta‐heuristic method. This chapter begins with a brief literature review about the DE and its applications, followed by a presentation of the DE’s fundamentals and a pseudocode.

9.1 Introduction

Differential evolution (DE) was developed by Storn and Price (1997). The DE was designed primarily for continuous optimization problems. Lampinen and Zelinka (1999) presented a modified DE for discrete optimization. Vesterstrom and Thomsen (2004) demonstrated that DE had a better performance in comparison with other optimization techniques such as the genetic algorithm (GA) and particle swarm optimization (PSO). The DE algorithm has been successfully applied to solve a wide range of optimization problems such as clustering, pattern recognition, and neural network training (Price et al., 2005). Tang et al. (2008) applied the DE to structural system identification. Lakshminarasimman and Subramanian (2008) implemented the DE for optimization of power systems. Qing (2009) demonstrated different applications of the DE in electrical engineering. Wang et al. (2009) applied the DE for optimum design of truss structures. Gong et al. (2009) applied the DE to optimal engineering design. Xu et al. (2012) implemented the DE to estimate parameter of a nonlinear Muskingum ...