【作者】 郑强；

【导师】 卢建刚；

【作者基本信息】 浙江大学 ， 控制理论与控制工程， 2006， 硕士

【摘要】 多目标优化问题在科学和工程等领域是一个热点问题,同时也是研究的一个难点。传统的多目标优化方法在某些复杂的多目标优化问题中存在不足,因此渐渐地被一些优越的多目标优化算法所替代。第一代非支配排序遗传算法NSGA在多目标领域中显示出比较大的优势,但是随着应用范围的不断拓宽,其缺点就不断暴露出来。为了更好地解决多目标优化问题,在NSGA的基础上,研究人员提出了带精英策略的非支配排序遗传算法NSGA-Ⅱ。 国外关于NSGA-Ⅱ的理论研究和应用研究都比较深入,而在国内目前还比较少。本文通过对该算法的研究和学习,将其应用于一些实际的问题,利用NSGA-Ⅱ在多目标优化领域强大的优势,很好地解决了这些实际的问题。 本文主要的工作为: (1) 介绍了多目标优化问题及其研究现状,并系统地介绍了遗传算法的相关理论;重点论述了非支配排序遗传算法NSGA尤其是其改进算法NSGA-Ⅱ的基本原理、算法基本流程等。 (2) 对电网谐波进行概述,简单介绍了谐波的概念、谐波的产生及其危害等,并将NSGA-Ⅱ应用在用于滤除电网谐波的无源滤波器的设计中。经过对某铝厂电网谐波的分析,将无源滤波器的优化模型的目标函数设置为无源滤波器初期总投资额最小和无源滤波器滤波率最大,并对初期投资成本和各次单谐波支路的滤波率进行约束。这是带约束条件的多目标优化问题,利用NSGA-Ⅱ在该问题上的优势,对其进行优化,从而找到一系列的全局最优解。文中选取其中一个最优解,通过仿真证明:该方法不但可以有效滤去大量的电网谐波,更多还原

【Abstract】 Multi-objective optimization is a research focus as well as a difficult problem in the fields of science and engineering. The conventional methods of multi-objective optimization have been used to solve the problems associated with the multi-objective optimization, but these methods show several shortcomings in some complicated multi-objective problems, and they are gradually replaced by other methods. Nondominated Sorting Genetic Algorithm(NSGA) shows great advantages in the problems of multi-objective optimization, but it has also aroused criticisms in several aspects after widely used. In order to solve the problems of multi-objective optimization more effectively, the Elitist Nondominated Sorting Genetic Algorithm (NSGA-II) is proposed on the basis of NSGA.The research of NSGA-II from both theoretical and practical perspectives is conducted extensively in other countries, however, the research associated with this aspect is rather limited domestically. In this paper, the basic theory of NSGA-II is studied, and more importantly, NSGA-II is resorted to solve some practical problems of multi-objective optimization. With the great advantages of NSGA-II, those practical problems are solved well.The main contents are as follows:(1) The problems of multi-objective optimization and the current state of the research on these problems are introduced. Meanwhile, the basic theory of genetic algorithm is also systematically presented. In addition, greatimportance is attached to the introduction of NSGA and especially the basic theory of NSGA-II.(2) The concept and the harm of harmonic as well as how the harmonic come into being are introduced, and in order to eliminate the harmonic, NSGA-II is used to design the passive filter. Through analyzing the harmonic from an aluminum factory, the minimum initial cost and the maximum filtering rate of passive filter are set as the objectives of the passive filter’s optimization model in this paper. After optimized by NSGA-II, a series of optimal solutions are obtained. And this method of optimization proves effective through simulation.(3) NSGA-II is adopted to identify the parameters of the kinetic model of methanol-to-hydrocarbons process and the dynamic model of catalytic cracking of diesel oil. These models can be denoted by differential equations, but the models are uncertain due to the unknown parameters. In order to get the exact models, the identification of the unknown parameters is a necessity. This paper proposed a new method of parameters identification using NSGA-II and one-step-integral Treanor algorithm, and this method is proved to be more effective than conventional ones.更多还原