Improved mutual information maximizing input clustering algorithm is a kind of discrete estimation of distribution algorithm, which is convenient to solve permutation flow shop scheduling problem. In this paper, the e...
详细信息
Improved mutual information maximizing input clustering algorithm is a kind of discrete estimation of distribution algorithm, which is convenient to solve permutation flow shop scheduling problem. In this paper, the encoding mode and probability model are improved, new individual strategy is proposed, greedy algorithm is introduced at the initial phase of the probability matrix, and crossover operator, mutation operator, insert operator and swap operator are adopted during the process of evolution, dynamic adjusted method is employed to determine the population size. These improvements gurantee the population diversity even in small population. Experiment results show that the improved mutual information maximizing input clustering algorithm is effective and stable.
暂无评论