基于缓冲时间的共享车位分配鲁棒方法

Robust method of shared parking allocation based on buffer time

共享停车平台根据供给车位和停车需求时间窗确定车位分配方案.为应对停车时发生提前到达或延迟离开等不确定情况,通过预留适当缓冲时间来增强车位分配方案的鲁棒性,进而构建了最大化平台收益和鲁棒性的车位分配问题双目标优化模型.利用加权和方法将双目标转化为单目标函数,加入辅助决策变量将非线性转化为线性模型.通过数值算例验证本方法的有效性,并对关键参数进行灵敏度分析.结果表明,在合理时间内,本方法能够求得算例的最优解及各项指标值.随着双目标调节系数的增大,平台收益分段增加,而鲁棒性分段下降.从不同调节系数得到的最优解中筛选出非支配解,并拟合得到近似帕累托前沿曲线,在供给数为5,需求数为25的测试算例中,单位平台收益增加所带来鲁棒性的减少量变化范围为[0.02,6.75].

Shared parking platform determines the parking allocation schemes according to the time windows of parking supplies and parking demands. In order to deal with uncertain situations such as early arrival or delayed departure during parking, the robustness of the parking allocation scheme is enhanced by reserving appropriate buffer time. This paper builds a bi-objective optimization model of parking space allocation problem that maximizes the platform revenue and the robustness. The bi-objective function is transformed into a single objective function by the weighted sum method, and the nonlinear model is transformed into a linear model by adding auxiliary decision variables. Finally, the effectiveness of the method is verified by a numerical example, and the sensitivity analysis of key parameters is carried out. The results show that within a reasonable time, the method can obtain the optimal solution and various index values of the example. With the increase of the dual-objective adjustment coefficient, the platform revenue segment increases, while the robustness segment decreases. The non-dominated solutions are screened out from the optimal solutions obtained by different adjustment coefficients, and the approximate Pareto frontier curve is obtained by fitting. The robustness reduction varies in the range of [0.02, 6.75].