DynDE: a Differential Evolution for Dynamic Optimization Problems

Abstract

This paper presents an approach of using Differential Evolution (DE) to solve dynamic optimization problems. Careful setting of parameters is necessary for DE algorithms to successfully solve optimization problems. This paper describes DynDE, a multi-population DE algorithm developed specifically to solve dynamic optimization problems that doesn't need any parameter control strategy for the F or CR parameters. Experimental evidence has been gathered to show that this new algorithm is capable of efficiently solving the moving peaks benchmark.

Self-Adaptive Differential Evolution for Dynamic Environments with Fluctuating Numbers of Optima

“…An effective method of tracking all optima is to maintain several independent sub-populations of DE individuals, one sub-population on each optimum. In their most successful experiments Mendes and Mohais [19] used 10 subpopulations, each containing 6 individuals.…”

“…Use of Brownian individuals involves the creation of individuals close to the best individual by adding a small random value, sampled from a normal distribution, to each component of the best individual. Mendes and Mohais [19] adapted the ideas from Blackwell and Branke [3] to create their multi-population algorithm, DynDE, which uses exclusion to prevent populations from converging to the same peak. Mendes and Mohais [19] showed that DynDE was at least as effective as its PSO based counterparts.…”

“…Mendes and Mohais [19] adapted the ideas from Blackwell and Branke [3] to create their multi-population algorithm, DynDE, which uses exclusion to prevent populations from converging to the same peak. Mendes and Mohais [19] showed that DynDE was at least as effective as its PSO based counterparts. DynDE will be discussed in detail in Section 1.5.…”

“…An approach similar to quantum individuals, called Brownian individuals, is utilized by DynDE, a DE based algorithm for dynamic environments [19]. Use of Brownian individuals involves the creation of individuals close to the best individual by adding a small random value, sampled from a normal distribution, to each component of the best individual.…”

“…DynDE is a differential evolution algorithm developed by Mendes and Mohais [19] to solve dynamic optimization problems. Sections 1.5.1 to 1.5.5 provides an overview of the components inherited by DynPopDE from DynDE.…”