A Self-adaptive Multipeak Artificial Immune Genetic Algorithm

Genetic algorithm is a global probability search algorithm developed by simulating the biological natural selection and genetic evolution mechanism and it has excellent global search ability, however, in practical applications, premature convergence occurs easily in the genetic algorithm. This paper proposes an self-adaptive multi-peak immune genetic algorithm (SMIGA) and this algorithm integrates immunity thought in the biology immune system into the evolutionary process of genetic algorithm, uses selfadaptive dynamic vaccination and provides a downtime criterion, the selection strategy of immune vaccine and the construction method of immune operators so as to promote the population develop towards the optimization trend and suppress the degeneracy phenomenon in the optimization by using the feature information in a selective and purposive manner. The simulation experiment shows that the method of this paper can better solve the optimization problem of multi-peak functions, realize global optimum search, overcome the prematurity problem of the antibody population and improve the effectiveness and robustness of optimization.