截断随机出行时间下可靠网络均衡模型

Reliability-based network equilibrium model with truncated stochastic travel time

针对现有随机交通网络均衡模型未考虑路径出行时间的有界性和准时到达概率对出行者路径选择行为影响的问题,基于截断随机出行时间,提出了克服其局限性的可靠网络均衡条件,该均衡条件下没有出行者可以通过单方面改变出行路径来提高准时到达概率.构建了该均衡条件的等价变分不等式(VI)模型,并证明了其等价性和解的存在性.设计了基于路径的相继平均(MSA)算法对模型求解.采用Nguyen-Dupuis网络对可靠网络均衡模型和MSA算法的有效性进行了测试.研究结果表明:该算法能够快速收敛到较高精度;与不考虑随机出行时间有界性的模型相比,网络均衡状态下的准时到达概率和流量分布均存在差异,最大路段流量相对变化值达到38.5%;增加出行时间预算和降低出行时间上界均可以有效提高起讫点间的准时到达概率.

Aiming at the problem that the existing stochastic transportation network equilibrium models did not consider the effect of the boundedness of path travel time and on-time arrival probability(OTAP) on traveler’s route choice behavior, the reliability-based network equilibrium condition was proposed based on the truncated stochastic travel time to overcome the limitation. In this equilibrium condition, no travelers could improve the OTAP by changing the path unilaterally. The variational inequality(VI)model equivalent to the equilibrium condition was established, and the equivalence and the existence of the solution were proved. The path-based MSA(method of successive average) algorithm was designed to solve the model. The Nguyen-Dupuis network was used to test the validity of the reliability-based network equilibrium model and the MSA algorithm. The results show that the algorithm can rapidly converge to a high accuracy. Compared with the model without considering the boundedness of stochastic travel time, the OTAP and the flow pattern are quite different in the network equilibrium state, and the maximum relative change of the link flow reaches 38.5%. Increasing the travel time budget and lowering the travel time upper bound can both effectively improve the OTAP between the origin and the destination.