客户竞争驱动的任务调度非合作博弈

NON-COOPERATION GAME FOR CUSTOMER'S COMPETITION DRIVEN JOB SCHEDULE

从客户竞争需求的角度出发,在分析其提交制造任务的各自利润最大化(重点考虑任务的完成时间)调度目标的基础上,针对传统的任务调度策略、方法和模型的不足,采用博弈论,提出并构建了一种具备完全信息的非合作博弈的任务调度模型。在该任务调度模型中,来源于不同客户的制造任务被映射为博弈模型中的局中人,与各制造任务所包含的工序集对应的可选加工设备被映射为各制造任务的可行方案集,各制造任务加工完成时间的倒数被映射为其收益函数。据此,对上述任务调度目标的求解就等价为寻求该任务调度模型的Nash均衡点,对Nash均衡点的具体求解,采用遗传算法实现。最后给出具体的实例来验证其可行性。

From the perspective of viewing customer＇s competition requirements and on the basis of analyzing the goal of each manufacturing task submitted by correspondent customer maximizing its own payoffs （mainly focusing on the finishing time of each manufacturing task）, together with the deficiency of traditional strategies, methods and models of task schedule, a non-cooperation game model with complete information for task schedule is presented and established. In this task scheduling model, the players correspond to the jobs submitted respectively by related customers, the strategies of each job correspond to the selectable geographical distributive machines related to the processes of this job, and the payoff of each job is defined as the reciprocal of its finishing time. As such, obtaining the optimal scheduling results is determined by the Nash equilibrium （NE） point of this game. In order to find the NE point, genetic algorithm is introduced. A numerical case study is given to demonstrate the feasibility of above methods.