Scenario-driven cooperative artificial bee colony algorithm for energy-efficient robust machine scheduling
针对不确定加工时间下的节能稳健机器调度问题，提出一种场景驱动合作人工蜂群算法，以最小化最大完工时间和平均总能耗，实验表明该算法优于四种现有算法。
This paper proposes an energy-efficient robust machine scheduling problem with the makespan. Uncertain processing times are modelled by the multi-scenario approach. The efficiency objective is to minimise the min–max criterion, and the energy objective is to minimise the average total energy consumption across scenarios. Specifically, the energy-efficient robust scheduling problem is discussed in the context of unrelated parallel machines. The property of the proposed problem is discussed with respect to the energy objective, which is addressed using the division strategy. A scenario-driven cooperative artificial bee colony algorithm is developed. Scenario neighbourhood structures are constructed under dynamically selected target scenarios based on problem-specific knowledges. In the employed bee phase, bi-subswarm cooperative local search is performed based on two scenario neighbourhood structures guided respectively by two objectives. In the onlooker bee phase, the local search is performed based on the third scenario neighbourhood structure to improve Pareto-solution archive. An extensive experiment was conducted. The computational results show that the developed algorithm is obviously advantageous to four state-of-the-art alternative algorithms for the proposed problem. The Pareto solution set obtained by the proposed problem could provide various selections of robust schedules with different trade-offs between solution conservativeness and energy reduction for the decision maker.