Research Of Multi-objective Optimization Based On Improved Immune Algorithm

Many optimization problems in scientific research and engineering practice can be modeled as multi-objective optimization problems. Thus, multi-objective optimization problems (MOP) have a wide range of applications, and designing effective algorithms for them is of not only the great importance in scientific research, but also the great value in applications. Based on the theory of artificial immune system and multi-objective optimization, the paper focuses on the problem of solving complex multi-objective optimization algorithm which is based on immune theory. The main contents include:Firstly, the paper introduces the status quo in the artificial immune system and its multi-objective optimization algorithm. After comparing several operators characteristics of multi-objective optimization which are based on immune algorithm, and analysis of the existing multi-objective optimization algorithm in handling complex problems with high dimensional mode carefully, it is difficult to make the optimization results to approximation the best optimization of the Pareto while maintaining solution diversity issues, so we come up with a hybrid mutation strategy, and a hybrid multi-objective optimization algorithm of the clonal selection variation at the same time. In this algorithm, we evolve two antibody groups, different groups using different antibodies mutation operator, and by updating the external memory antibody population, to retain the optimal evolution of antibody, to avoid degradation in the late evolutionary algorithm. We select five ZDT multi-objective optimization problem to test, and compare the performance of the algorithm with the classical NSGA-Ⅱ, SPEA-Ⅱ, NNIA algorithm, the results show that the new algorithm has a better solve performance and robustness.Finally, we introduce high-dimensional multi-objective optimization problem whose target number is more than three. For the treatment of high dimensional objective optimization problem, we often face up with these problems that the environment of the best optimization front of Pareto is complex and the computing is large. So we improve the mixed mutation strategy Strategies and individual choice, and select the two objectives of scalable high-dimensional optimization problem DTLZ, simulation the improved algorithm, after comparing with better NNIA algorithm in the current treatment algorithm for high-dimensional mode, which shows that the algorithm performance has a better improvement.