研究决策者面对突发事件,应对单机调度中的应急管理问题。在完全没有突发事件发生时间和发生次数信息条件下,利用局内决策理论与方法构建确定情形下的应急策略,并利用“竞争比”说明该策略的有效性。在此基础上进一步研究随机模型下的平均竞争比。理论和数值分析表明指数分布的引入使得竞争分析的性能得到显著改善。
The paper explores the problem that how managers deal with the single-machine scheduling in the case of unexpected events. In the absence of information about the time and the number of occurrence of unexpected events, we construct the contingency schemes for the program using decision theory and methods, and use the "competitive ratio" to explain the effectiveness of the schemer on this basis, we further study the average-case competitive ratio for a random model with exponential distribution. Theoretical and numerical results show that the performance measure of competitive analysis can be dramatically improved.