基于人工生命的优化方法及应用  

An Optimization Method Based on Artificial Life and Its Application

作　　者：张奇[1] 陈国初[2] 俞金寿[1] 

机构地区：[1]华东理工大学自动化研究所,上海200237 [2]上海电机学院电气学院,上海200240

出　　处：《华东理工大学学报（自然科学版）》2008年第2期273-277,共5页Journal of East China University of Science and Technology

摘　　要：将人工生命科学引入优化问题的处理中,通过对"生命游戏"这一特殊元胞机制的研究,并结合遗传算法(GA),建立了基于GA的元胞自动机优化模型。将该方法用于常用测试函数优化问题的求解,仿真结果表明所提出的新方法具有良好的性能。把基于GA的元胞自动机优化方法和BP神经网络相结合,用于丙烯腈收率软测量建模。结果表明:基于此方法的软测量模型精度高、性能较好,具有应用前景。This paper presents a cellular automata model based on ＂life-game＂,a special mechanism in artificial life, by applying the artificial life into resolving the optimization problems. The model proposed in this paper is applied to the optimization of several test functions. The simulation shows that this new method has a good performance in optimization. Furthermore, the experimental results imply that a soft sensing model based on this method （GA-CA） has satisfying performance ,remarkable measuring precision and potential utility when it is applied to measur the selectivity to Acrylonitrile.

关 键 词：人工生命 遗传算法 元胞自动机 丙烯腈收率 软测量 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]