The AEA Combined With Clonal Selection Algorithm (AEA-C) And Its Application On Constrained Optimization Problems

AEA (Alopex-based Evolutionary Algorithm) is a kind optimization Algorithm which combines the heuristic way of Alopex and the swarm intelligence of Evolutionary Algorithm. In this paper, an improved AEA algorithm (AEA-C) which is fused AEA with Clonal Selection Algorithm is proposed. In addition, the AEA-C is applied to solve the difficult constrained optimization problems.The optimization process of AEA algorithm mainly depends on the Alopex operation between individuals in population, so the performance of the algorithm directly depends on the quality of the population. This paper puts forward using clonal selection to excavate the potential information of excellent individuals. The diversity of evolution population is promoted by the operations of cloning and mutation. In the mutation, the step length of AEA is used for reference to strengthen global search and local search at various search stages. The performance of AEA-C is studied by using22benchmark functions and shows that the AEA-C clearly outperforms the other algorithms for almost all the benchmark functions. Furthermore, AEA-C is applied to parameter estimation for fermentation dynamics models, and the satisfactory results are obtained.In addition, a new adaptive penalty function method based on AEA algorithm is proposed to solve constrained optimization problems. This method adds different penalties to individuals according to their performance in the whole population. To explore the cryptic and deep information of infeasible individuals, the proposed method tries to maintain a given proportion of infeasible individuals in the population using an adaptive penalty mechanism during the optimizing process. The performance of the proposed method is tested by11benchmark functions and compared with five well-known algorithms at the same conditions. The results show that the adaptive penalty function method is efficient in solving constrained optimization problems. In addition, this method is used to optimize a butene alkylation process and performs excellent at main characters. The statistical experiment results demonstrate that the new method is effective to solve the practical problem.