研究了工件含不同到达时间且差异尺寸的单机批调度问题,优化目标为极小化最大完工时间Cmax。首先建立了该问题的混合整数规划模型,并使用运筹学软件CPLEX求解。针对问题构建性的特性,提出了蚁群优化算法求解该问题。通过分析到达时间和工件尺寸等约束对优化目标的影响,提出了空闲空间的概念Cmax,并证明极小化等价于极小化批空闲空间,基于此设计了动态的启发式信息以更精确地指导蚂蚁的行为。同时,引入候选列表策略,有效地减少了蚂蚁的寻优空间,提高了算法的收敛速度。仿真实验采用了文献中的测试算例,通过与CPLEX软件以及文献中的启发式算法和遗传算法的对比分析,验证了算法的有效性。
A single batch scheduling problem with different release time and non-identical job sizes was studied.The optimization objective was minimize Cmax.A mixed integer programming model was established and solved by operational research software CPLEX.Based on the constructive characteristic of the problem,an Ant Colony Optimization(ACO) was proposed.By analyzing the influence of release time and job sizes' constraint on the optimization objective,the concept of Idle Space(IS) was proposed and proved that minimizing IS was equivalent to minimize Cmax.Based on the analysis,dynamic heuristic information was designed to guide ants' behavior more accurately.Meanwhile,a candidate list strategy was employed to reduce ants' search space and to speed up the convergence rate of algorithm.In the experimental section,test instances from literature were selected.The effectiveness of proposed ACO algorithm was validated by comparative analysis with CPLEX,heuristic,and Genetic Algorithm(GA) in literature.