基于遗传微粒群算法的工程项目工期优化研究  

Optimal Duration of Engineering Project Based on GA-PSO

作　　者：李冰洁[1] 牛东晓[1] 邱金鹏[1] 

机构地区：[1]华北电力大学经济与管理学院,北京102206

出　　处：《电力与能源》2014年第5期553-556,共4页Power & Energy

基　　金：输变电工程造价动态管控关键技术研究(国家电网公司总部科技项目;201401-201512)

摘　　要：企业的项目管理活动均会涉及到两个重要的绩效管理目标,即项目最小工期和最优资源配置。研究了在资源有限的情况下工期最优问题。运用遗传和微粒群相结合的优化算法,解决了标准粒子群算法容易陷入局部最优,从而出现所谓的"早熟"现象。通过对某工程项目工期优化的实例分析,用Matlab 7.0软件编程,验证了算法的有效性,即循环计算1万次,经四舍五入圆整处理,得到最优工期为35天,比初始值计算结果减少3天。项目工期优化分析认为:还需要拓展研究模型参数的科学取值和独立制作软件模块供各类企业方便使用等问题。The proj ect management activities in enterprises will always involve two important performance man-agement objectives,namely the minimum project duration and optimal allocation of resources.Studied the problem of optimal duration when resources are limited.With the application of GA （Genetic Algorithm）and PSO （Particle Swarm Optimization）,solved the defect of easily falling into local optimum which is the so-called “premature”phenomenon in the standard particle swarm optimization.By analyzing an actual project schedule optimization,the effectiveness of the method is demonstrated combining with programming through Matlab 7.0 software,which was calculated 10 000 times,after rounding processing to give optimal duration of 35 days,compared with the initial value calculated decrease 3 days.Analysis：some problems still need to study the model parameters also need to expand the scientific value like selecting scientific model parameters and indie software module for all kinds of enterprises.

关 键 词：资源有限-工期最短 遗传微粒群算法 轮盘赌选择法 工期优化 

分 类 号：F204[经济管理—国民经济]