基于混合蚁群算法的产品开发过程优化方法  

Optimized method of iterated development process based on a hybrid ant colony algorithm

作　　者：梁亮[1] 郭波[2] 

机构地区：[1]长沙理工大学汽车与机械工程学院,长沙410014 [2]国防科技大学信息系统与管理学院,长沙410073

出　　处：《系统工程理论与实践》2009年第10期118-128,共11页Systems Engineering-Theory & Practice

摘　　要：通过对迭代产品开发过程的分析,提出了将产品开发过程中设计活动被首次访问视为TSP问题中蚂蚁访问城市的思想,将Markov过程建模方法与基本蚁群算法相结合,建立了混合蚁群算法对产品开发过程进行优化求解.示例表明该方法成功地将蚁群算法扩展到复杂产品开发过程优化问题,在考虑设计迭代以及设计活动完成时间服从任意分布的情况下,建立了产品开发过程优化模型,为该类问题的求解提供了一个新的思路和方法.Design iteration and time uncertainty are the two basic characteristics in the development process. That makes the development process optimization problem unlike the Traveling Salesman Problem （TSP） and can not be solved directly by using the general ant colony algorithm. Through the analysis of the iterated development process, a new idea was proposed that the design activity firstly being visited in the development process would be treated as the city visited by ants in the traveling salesman problem. Then a hybrid ant colony algorithm which integrates the Markov process methods into the general ant colony algorithm was built for solving the development process optimization problem. Case studies demonstrate that the method successfully extended the general ant colony algorithm to the more complicated development process optimization problems. With consideration that the completed times of design activities follow arbitrary distribution and design iteration dose exist, an optimized model has been established which provides a new solution method for the development process optimization problems.

关 键 词：产品开发过程 设计迭代 混合蚁群算法 PH分布 MARKOV过程 优化 

分 类 号：TH165[机械工程—机械制造及自动化]