冷却时间约束多对多任务分配及其优化  

Task Allocation under Cool Down Time Constraints via the Many to Many Assignment

作　　者：刘冬宁[1] 郑楚楚 Liu Dong-ning;Zheng Chu-chu(School of Computers,Guangdong University of Technology,Guangzhou 510006,China)

机构地区：[1]广东工业大学计算机学院,广东省广州市510006

出　　处：《广东工业大学学报》2021年第5期10-15,共6页Journal of Guangdong University of Technology

基　　金：国家自然科学基金资助面上项目(62072120)。

摘　　要：由于人员分配或资源调度过程中存在冷却时间,导致人机协同多对多任务分配困难并且难以优化,协同效应急剧下降,本文旨在解决冷却时间约束多对多任务分配问题。主要使用角色协同理论(Role-Based Collaboration)及其通用模型E-CARGO(Environments-Classes,Agents,Roles,Groups,and Objects)的子模型群组角色多对多指派(Group Multirole Assignment,GMRA),以医院医生出诊场景为例,对冷却时间约束多对多指派形式化建模;继而采用整数规划方法对冷却时间约束进行解耦与消解,利用IBM CPLEX优化包对团队执行力的最大化做最优求解。进而在大规模仿真实验中,采用必要条件对解空间进行了快速归约,实现了秒级精确求解,进一步论证了模型与方法的一般性、高效性和可靠性。Task allocation is a common and important problem in personnel management.In the process of allocation,it is often affected by many different constraints.One of them is the cool down time constraint.Due to the cool down time in personnel allocation or resource scheduling,it is difficult to allocate and optimize the many to many tasks in human-computer cooperation,and the collaboration effect drops sharply.The group multirole assignment(GMRA)is used to formalize the problem,and the integer programming used to decouple and eliminate the cool down time constraints.After that,IBM ILOG CPLEX optimization package(CPLEX)is used to optimize the team execution.In addition,in the large-scale simulation experiment,the necessary conditions are used to reduce the solution space quickly,and the second level accurate solution is achieved.The generality,efficiency and reliability of the model and method are further demonstrated.

关 键 词：人机协同 任务分配 多对多指派 冷却时间约束 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]