Estimation of distribution algorithm with scatter search for dynamic optimisation problems

Aiming at the trouble to track the optima in dynamic environments with estimation of distribution algorithms EDAs. An estimation of distribution algorithm with scatter search EDASS is proposed in this paper. Its basic idea is to employ a scatter search to increase the diversity in a guided fashion and an adaptive leader clustering method to locate multiple local optima. Both the information of current population and the part history information were referred for building probability model. The experimental results show that the EDASS is effective for dynamic optimisation problems.