多项目多任务选择计划模型及其免疫遗传算法  

MULTI-PROJECT MULTI-TASK SELECTION PLAN MODEL AND ITS IMMUNE GENETIC ALGORITHM

作　　者：夏林丽[1] 张著洪[1] 

机构地区：[1]贵州大学系统科学及信息技术研究所,贵州贵阳550025

出　　处：《计算机应用与软件》2011年第4期124-126,131,共4页Computer Applications and Software

基　　金：教育部自然科学重点项目(208125)

摘　　要：针对多资源受限下项目群选择计划及项目流程调度问题,依据项目的权重和承继约束,将项目群按权值从大到小进行排序,各项目内部的任务分别采用网络计划图进行拓扑排序,并以资源约束和合同期等因素为项目选择标准,建立多项目多任务选择计划的资源配置数学模型。进而,在遗传算法中引入免疫系统的记忆性和多样性功能,设计免疫遗传算法求解所获模型的最佳决策方案。比较性的数值实验结果说明了模型的合理性和算法的有效性。For the problem of project group selection plan and project flow scheduling in condition of multiple resources constraints,the project group is ranked from big to small according to weight value based on the weights and inheritance constraints between the projects.Internal tasks of each project are topologically sorted using network planning graph respectively,and a resource-allocation mathematical model of multi-project multi-task selection plan is established taking the factors of resource-constraints and project contracts as the selective criteria.Furthermore,an immune genetic algorithm is designed to find the best decision scheme of the obtained model by introducing memory and diversity functions in immune system into the genetic algorithm.Comparative numerical experiment results show the rationality of the model and the effectiveness of the proposed algorithm.

关 键 词：多项目多任务选择计划 资源均衡 承继约束 拓扑排序 免疫遗传算法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]