Abstract: To overcome the shortcomings of quantitative analysis of shockwave phenomenon in high density pedestrian flow at the bottleneck, micro-characteristic parameters such as pedestrian passage time, walking speed and position change were calibrated and extracted, on the basis of pedestrian controlled experiment. Dynamic time warping (DTW) algorithm was introduced to realize the identification and quantitative analysis of shockwave phenomenon in passenger flow. Results show that it is available to identify the shockwave phenomenon of passenger flow with the DTW algorithm, which can be more accurate to identify the variable speed point and can reduce the complexity of shockwave velocity calculation, compared with the hydrodynamic theory algorithm. The shockwave velocity at the bottleneck is a variable in the process of transmission. The shockwave velocity decreases (increases) as the convergent wave (dissipative wave) propagates, and the blocking degree of the former to followers decreases gradually with the propagation of shockwave. The study provides a theoretical basis for the stability evaluation of rail transit passenger flow and the validity verification of micro- simulation model.

Key words: traffic engineering, shockwaves, DTW algorithm, traffic bottleneck, pedestrian behavior

摘要： 针对密集客流在瓶颈处激波现象量化解析的不足，基于行人可控试验，标定并提取行人通过时间、行走速度、位置变化等微观特征参 数，引入动态时间规整(dynamic time warping, DTW)算法，实现客流激波现象的识别与量化分析。结果表明：DTW算法可识别客流激波现象 ，与流体动力学理论算法相比，该算法能够更精确地识别变速点，降低激波波速计算的复杂度；瓶颈处激波波速在传递过程中为变量，集结 波(消散波)传递时波速递减(递增)，前者对跟随者的阻挡程度随激波传递逐渐降低。研究为轨道交通客流的稳定性评估、微观仿真模型的有 效性验证提供理论依据。

关键词: 交通工程, 激波, DTW算法, 交通瓶颈, 客流行为