Artificial immune system application for solving dynamic optimization problems

Zhijie Li, Yuanxiang Li, Li Kuang, Fei Yu

Abstract: For the purpose of adaptation to a changing environment, immune mutation and memory mechanism in the immune system are introduced in thermodynamic genetic algorithm, which helps to prevent the diversity loss and rapidly track the optimum in dynamic environments. Experimental results on 0/1 dynamic knapsack problems demonstrate the merits of the proposed immune thermodynamic genetic algorithm (ITDGA). Compared with the existing classical primal-dual genetic algorithm (PDGA), this algorithm can maintain better diversity and be more suitable to solve 0-1 dynamic problems.