CIESC Journal ›› 2014, Vol. 65 ›› Issue (12): 4857-4865.DOI: 10.3969/j.issn.0438-1157.2014.12.029

Previous Articles     Next Articles

AEA combined with differential evolution and its application on parameter estimation

HE Pengfei, LI Shaojun   

  1. Key Laboratory of Advanced Control and Optimization for Chemical Processes, Ministry of Education, East China University of Science and Technology, Shanghai 200237, China
  • Received:2014-09-18 Revised:2014-09-25 Online:2014-12-05 Published:2014-12-05
  • Supported by:

    supported by the National Natural Science Foundation of China (21176072).

融合差分进化算法的AEA算法及其在参数估计中的应用

何鹏飞, 李绍军   

  1. 华东理工大学化工过程先进控制和优化技术教育部重点实验室, 上海 200237
  • 通讯作者: 李绍军
  • 基金资助:

    国家自然科学基金项目(21176072).

Abstract: Focusing on the demerits of AEA (Alopex-based evolutionary algorithm), this paper proposed a modified AEA algorithm (MAEA), which was fused AEA with differential evolution (DE). In order to enhance the performance of the algorithm, the generation method of population in AEA was improved by applying an improved DE operation to AEA. The modified algorithm not only takes advantage of heuristic search and deterministic search of AEA, but also increases the population diversity and is adapted to global search and local search. Then the MAEA algorithm was tested by 21 benchmark functions,the results show that MAEA outperforms DE, MDE and AEA. Further comparison results between MAEA algorithm and a representative state-of-the-art algorithm(ISDEMS) indicate that, the performance of the modified algorithm is significantly improved, in both accuracy and stability. Furthermore, the algorithm was applied to the parameter estimation of the models of fermentation dynamics, and the satisfactory results were obtained.

Key words: Alopex, AEA, differential evolution, optimization, parameter estimation

摘要: 着眼于AEA(Alopex-based evolutionary algorithm)算法本身的不足,构造出一种融合了差分进化算法和AEA的改进型算法——MAEA(modified AEA).MAEA算法将改进后的差分进化算法嵌入到AEA中,改进AEA算法中种群的生成方式,提高算法的寻优能力.改进的算法不仅拥有启发搜索和确定性搜索的优点,同时还增加了种群的多样性,使算法能够更好地进行全局和局部搜索.通过21个标准函数的测试结果表明,该算法较标准AEA算法、差分进化算法的性能有较大提升.进一步和当前具有代表性的先进算法(ISDEMS)的比较结果表明,MAEA算法有较高的精确度和稳定性.将算法用于发酵动力学模型参数的估计,通过优化得到了较好的结果,验证了本文提出的算法的可行性和有效性.

关键词: Alopex, AEA, 差分进化算法, 优化, 参数估计

CLC Number: