随机需求下公交时刻表设计的鲁棒性优化

考虑在实际运营中乘客需求具有随机性,固定需求下优化的公交时刻表不适应运营的要求.随机需求下的期望值模型忽略了不利可能事件对运营的负面影响,针对此情况研究随机需求下公交时刻表设计的鲁棒性优化.模型综合考虑乘客成本与运营成本,采用鲁棒性优化权衡目标期望值与偏差期望值.结合随机模拟技术,选用遗传算法求解模型.给出了算例,验证了模型和算法的有效性.通过比较固定需求模型、随机需求期望值模型、随机需求鲁棒性模型,说明在鲁棒性优化下需要提供更多的交通供给以降低偏差期望值.最后,对鲁棒性模型中的偏差权重系数进行了灵敏度分析.

Robust optimization for transit timetable design under stochastic demands

Considering stochastic passenger demands in actual operation, transit timetable optimized under deterministic demands did not meet operation’s requirements. Expected value model under stochastic demands ignored negative impact, which made adverse possible events on operation. Aiming at this problem, robust optimization for transit timetable design under stochastic demands was studied. The model considered user cost and operation cost, and robust optimization was used to make a trade-off between expected value of objective and expected value of variation. With stochastic simulation techniques, genetic algorithm was provided to solve the model. A numerical example was given, which demonstrated the validity of the model and algorithm. By contrasting deterministic demands model, stochastic demands expected value model, stochastic demands robust model, the results show that under robust optimization more transit supply needs to be applied in order to reduce expected value of variation. At last, sensitivity analysis on variation weight coefficient in robust model is given.