多项目选择计划及其两层决策免疫遗传算法  

Multi-project selective planning and its immune genetic algorithm for bi-level decision

作　　者：雷宏[1] 张著洪[1] 

机构地区：[1]贵州大学理学院系统科学及信息技术研究所,贵州贵阳550025

出　　处：《计算机工程与设计》2010年第9期1992-1994,2154,共4页Computer Engineering and Design

基　　金：教育部科学技术研究重点基金项目(208125);贵州大学研究生创新基金项目(校理工2009031)

摘　　要：探讨项目计划管理中,资源受限下含承继约束的多项目选择计划两级规划模型及其求解算法。依据项目的承继关系,将所有项目划分为若干子群,各子群内部的项目之间存在承继约束,子群之间仅有优先关系,进而获得了资源受限下资源配置平衡的两级规划模型。基于遗传算法的基本结构,引入免疫系统的多样性、记忆性等特征,获得了求解此模型的免疫遗传算法。数值实验结果表明了该模型设计的合理性及算法的有效性。This work investigates a bi-level programming model of resource allocation balance subjected to multi-project selection planning with constrained resource and inheritance in project planning management. In this model, all given terms are divided into multiple subpopulations with only the relation of superiority between them, among which each subpopulation includes multiple terms with the relation of inheritance. Further, a new immune genetic algorithm is proposed based on the structure of the basic genetic algorithm in relation to the characteristics of memory and diversity in the immune system. Experimental results and comparison show the rationality of the model and the effectiveness of the algorithm proposed.

关 键 词：多项目选择计划 资源平衡 拓扑排序 免疫遗传算法 网络计划图 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]