Evolutionary Computation Research Based On Immune Theory And Mind Evolution

In this paper, a new hybrid evolutionary model is proposed by combining Immune Theory with Mind Evolutionary Computation. And on its basis, a new Immune Algorithm Based on Multi-population (IABM) is presented.Based on Multi-population, IABM is mainly put forward to deal with optimization problems. It defines five basis operators such as selection, memory, clone, hyper-mutation and restraint. A Memory Operator is divided into a global memory-cell and local memory-sections. It completely utilizes parents' good gene information to guide the production of kids and accelerates the convergence speed. Adaptive mutation is used to enhance the effect of mutation by the way that individuals confirm the search field according to their quality and generation numbers of evolution. Restraint Operator can ensure the diversity of the global population. Experimental results show that the presented algorithm has efficient convergence speed and the global optimal value can be converged. The convergence speed of IABM is higher than MGA (Multi-population Genetic Algorithm) and MEC (Mind Evolutionary Computation). Additionally, its iteration number of convergence is less than them.The research work of this paper is following:1. Multi-population is introduced into Immune Evolutionary Algorithm. It can enhance the variety of populations in IABM and strengthen the dynamic control over them in the process of evolution in the other way.2. Mind Evolutionary Algorithm is organically combined with Immune Algorithm. The former can accelerate the local convergence speed. And the Restrain Operator of Immune Algorithm can ensure the diversity of population evolution. Therefore, the characteristics of better global convergence can be shown in IABM algorithm.3. Self-adaptive strategy is adopted. For example, the range and probability of mutation can be flexibly controlled by its application.