Operation Performance Assessment Of Urban Rail Transit Based On Travel Time Delay

With the rapid development of the urban rail transit network operation, passenger flow scale grows rapidly as well as metro network and station structure become complicated gradually,, leading to the frequent phenomenon that passengers congest and get lost are more and more, which is an important reason for travel-delay and increase of travel costs. The passenger travel time delay (DLT) reflects the operation quality of urban rail transit. To gain the history or future operation states of the urban rail transit efficiently, an estimation algorithm for the DLT based on a large amount of AFC data is developed in this paper, and the operation state of the urban rail transit can be evaluated by delay indexes.Delay referred in this paper is relative to the standard of the travel time. The paper proposes two delay evaluation strategies based on differentiated understandings for travel time standard, including the delay for enterprises and for passengers. By analyzing the characteristics of two kinds of delay, an evaluation strategy based on different delay characteristics of frequent passengers and random passengers during different period. To reflect the operation state objectively, different weight values are given to the enterprise and passenger respectively constituting comprehensive delay value. The paper regards the total delay, average delay, number of delay passengers and delay rate as delay indexes, and study the DLT evaluation methods on the levels of subway OD pairs, lines and the network. The method which can calculate the DLT based on OD pairs is the main focus in this paper. With the comprehensive consideration for delay degrees and delay probabilities, the paper defines expect delay time. According to the expect delay time, the clustering methodology is used in this paper to define grading standards for operation state evaluations among subway OD pairs, stations, lines and the network. Taking Beijing Subway as an example, with the consideration of no passenger transfers, the paper analyzes operation states between stations and lines during morning peak. The evaluation results is verified by comparing current limiting stations which are in the same periods, and the relatively high matching degree shows that the method developed in this paper is effective and reliable.