Clonal Selection Algorithm with Search Space Expansion Scheme for Global Function Optimization

摘要

Unlike evaluation strategy (ES) and evaluation programming (EP), clonal selection algorithm (CSA) strongly depends on the given search space for the optimal solution problem. The interval of existing optimal solution is unknown in most practical problem, then the suitable search space can not be given and the performance of CSA are influence greatly. In this study, a self-adaptive search space expansion scheme and the clonal selection algorithm are integrated to form a new algorithm, Self Adaptive Clonal Selection Algorithm, termed as SACSA. It is proved that SACSA converges to global optimum with probability 1. Qualitative analyzes and experiments show that, compared with the standard genetic algorithm using the same search space expansion scheme, SACSA has a better performance in many aspects including the convergence speed, the solution precision and the stability. Then, we study more about the new algorithm on optimizing the time-variable function. SACSA has been confirmed that it is competent for solving global function optimization problems which the initial search space is unknown.