Populations can be Essential in Dynamic Optimisation

摘要: Real-world optimisation problems are often dynamic. Previously good solutions must be updated or replaced due to changes in objectives and constraints. It is claimed that evolutionary algorithms particularly suitable for dynamic because a large population can contain different may useful the future. However, rigorous, theoretical demonstrations how populations essential sparse restricted special cases.This paper provides explanations of optimisation. The ability track optimal investigated by considering Hamming ball points moves randomly through search space. shown based on single individual likely unable optimum while non-elitist population-based able do so with overwhelmingly high probability. this holds range most commonly used selection mechanisms even without diversity enhancing mechanisms. Appropriate parameter settings achieve behaviour derived these