Study And Application Based On Adaptive Immune Clonal Selection Algorithm

Artificial Immune Systems (AIS) to simulate the excellentfeatures of biological immune system, is a kind of diversity, immunememory, self-organization, self-learning, adaptive and robust strongartificial intelligence system. This paper is to explore the foundationof understanding the biological immune, and mainly studied theimmune evolutionary learning mechanism implied in, put forward akind of very effective artificial immune system algorithm, and use it tosolve NP hard knapsack problem encountered in actual.In this paper, we mainly study the immune clonal selectionalgorithm (CSA) improvement method. First using the adaptiveadjustment of basic clonal selection algorithm is improved, andproposes a adaptive clonal selection algorithm (ACSA); Second usedynamic allocation cloning antibody group strategy, adjusting themutation probability strategy, clonal mutation strategy to improve theadaptive clonal selection algorithm, proposed an improved adaptiveclonal selection algorithm (New-ACSA); Finally combined with thedynamic programming algorithm is proposed to solve knapsackproblem in NP hard combinatorial optimization problems, an intelligentalgorithm—adaptive clonal selection algorithm, dynamicProgramming (DP-ACSA). By convergence analysis of the adaptive clonal selectionalgorithm, and compare with basic clone selection algorithmsimulation results, the facts show that the optimal solution in solvingfunction question when performance is superior to the latter. While incontrast, the improved adaptive clonal selection algorithm in thepopulation for clonal selection has certain targeted, to speed up itsglobal search, simulation results show the feasibility andeffectiveness of this algorithm. Finally, the improved adaptive clonalselection algorithm is used to solve the knapsack problem, dynamicprogramming strategy is introduced, using the optimal decisionprinciple of dynamic programming, provide excellent information onimmature population modifying individual genes to improvepopulation quality, according to the knapsack problem simulationtest shows that the algorithm is better than general clonal selectionalgorithm to find the optimal solution faster, and better performance.