模拟进化优化方法及其应用——遗传算法

The Optimization Methods by Simulated Evolution and Their Application

在本世纪六十年代中期,美国、德国等国家的一些科学家开始研究用模仿生物和人类进化的方法求解复杂的优化问题的方法,这里我们统称之为模拟进化优化方法(optimi:ation method by simulatedevolution)。但这些方法在六十年代和七十年代并未受到普遍的重视,一是因为当时这些方法还不成熟,二是当时计算机发展水平低,容量小,计算速度慢,这些方法又需要较大的计算量,难以实际应用。但在这期间有一些科学家一直在进行不懈的努力.代表性人物之一为美国的著名科学家J.H.Holland,他和他的学生甲直在对他所提出的一种模拟进化优化方法—遗传算法(Genetic Algorithm,GA)进行理论研究并开拓其应用领域。八十年代初期,伴随着人工神经元网络理论和机器学习理论的发展以及计算机容量和计算速度的不断提高,遗传算法的研究也越来越受到重视而逐步成熟起来,并日益受到各学科研究人员的普遍重视。自八十年代中期开始,这种方法除了在人工智能领域。

The optimization methods by simulated evolution are search procedures whose mecha* nism are based on those of natural evolution,and are powerful in finding the global or near global optimal solutions of optimization problems. This kind of optimization methods have found applica-tions in many fields,such as machine learning,robatics,civil engineering and power systems,and have shown to be of great promise. This paper systematically introduces some fundamental aspects of the optimization methods,including their theories,methods,implementation techniques and some common problems related to their applications,together with a concise survey of their application status in several fields.