多目标优化问题与算法的研究方案.docVIP

多目标优化问题及其算法的研究 摘要：多目标优化问题(MOP)由于目标函数有两个或两个以上，其解通常是一组Pareto最优解。传统的优化算法在处理多目标优化问题时不能满足工业实践应用的需要。随着计算机科学与生命信息科学的发展，智能优化算法在处理多目标优化问题时更加满足工程实践的需要。本文首先研究了典型多目标优化问题的数学描述，并且分析了多目标优化问题的Pareto最优解以及解的评价体系。简要介绍了传统优化算法中的加权法、约束法以及线性规划法。并且研究了智能优化算法中进化算法(EA)、粒子群算法(PSO)和蚁群优化算法(ACO)。 关键词：多目标优化问题；传统优化算法；进化算法；粒子群算法；蚁群优化算法 中图分类号：TP391 文献标识码：A Research of Multi-objective Optimization Problem and Algorithm Abstract: The objective function of Multi-objective Optimization Problem is more than two, so the solutions are made of a term called best Pareto result. Traditional Optimization Algorithm cannot meet the need of advancing in the actual industry in the field of the Multi-objective Optimization Problem. With the development in computer technology and life sciences, Intelligent Optimization Algorithm is used to solve the Multi-objective Optimization Problem in the industry. Firstly, the typical mathematic form of the Multi-objective Optimization Problem, and the best Pareto result of Multi-objective Optimization Problem with it’s evaluate system were showed in this paper. It’s take a brief reveal of Traditional Optimization Algorithm, such as weighting method, constraint and linear programming. Intelligent Optimization Algorithm, including Evolutionary Algorithm, Particle Swarm Optimization and Ant Colony Optimization, is researched too. Keyword: Multi-objective Optimization Problem; Traditional Optimization Algorithm; Evolutionary Algorithm; Particle Swarm Optimization; Ant Colony Optimization. 引言 所谓的目标优化问题一般地就是指通过一定的优化算法获得目标函数的最优化解。当优化的目标函数为一个时称之为单目标优化(Single-objective Optimization Problem, SOP)。当优化的目标函数有两个或两个以上时称为多目标优化(Multi-objective Optimization Problem, MOP)。不同于单目标优化的解为有限解，多目标优化的解通常是一组均衡解。显而易见，多目标优化问题比单目标优化问题更接近工程实践，同时更加复杂。很多工程实践中的优化问题最后都可以转化为多目标优化问题。因此，对多目标优化问题的深入研究对于实践应用更具价值。 通常，多目标优化问题都是通过一定的算法实现求解的。对多目标优化问题的研究也更多地集中于对各种算法的研究。目前多目标优化算法归结起来有传统优化算法和智能优化算法两大类。传统优化算法包括加权法、约束法和线性规划法等。智能优化算法包括进化算法（Evolutionary Algorithm, 简称EA）、粒子群算法（Particle Swarm Optimization, PSO）、人工免