Adaptive clonal selection with elitism-guided crossover for function optimization

Based on clonal selection principle, a novel evolutionary algorithm encoded in floating-point-number is proposed to solve function optimization problems. A micro-mutation operator and an elitism-guided crossover operator are defined respectively for the best and medium antibodies. The main features of the algorithm are combination of meticulous local with double-quick global search, and automatic adjustment of run-time parameters (adaptive extension or shrink of search space). The algorithm is empirically compared with similar approaches from the literature. The results demonstrate that the proposed algorithm can promptly and accurately locate the global optimum of complex function and has good stabilization.