Simulated Mutation in Differential Evolution

Population-based evolutionary algorithms (EAs) have been widely applied to solve various real-world optimization problems. In essence, they are a type of optimization techniques and traverse the population landscape in whatever direction that can lead to a peak or an optimal solution through climbing hills. The differences between different EAs are operators and combinations. Each EA has its own particular operators and its own particular combination of these operators. Typically, in EAs, there are three operators (i.e., crossover and/or mutation, and selection). Each EA utilizes its crossover operator and/or mutation operator to exploit and/or explore the search space of the problem, and utilizes its selection operator to guide the search towards the optimal region. Differential evolution (DE) and particle swarm optimization (PSO) are two relatively recent branches of EAs, have been successful in solving many real-world optimization problems, and have increasingly attracted researchers' attention. Through comparing and contrasting similarities and dissimilarities between DE and PSO, in this study, we make an attempt to develop some variants of mutation inspired by PSO in DE. Numerical experiments are conducted on a test set of global optimization problems.