A Hybrid Method for Predicting Traffic Congestion during Peak Hours in the Subway System of Shenzhen

Abstract

Traffic congestion, especially during peak hours, has become a challenge for transportation systems in many metropolitan areas, and such congestion causes delays and negative effects for passengers. Many studies have examined the prediction of congestion; however, these studies focus mainly on road traffic, and subway transit, which is the main form of transportation in densely populated cities, such as Tokyo, Paris, and Beijing and Shenzhen in China, has seldom been examined. This study takes Shenzhen as a case study for predicting congestion in a subway system during peak hours and proposes a hybrid method that combines a static traffic assignment model with an agent-based dynamic traffic simulation model to estimate recurrent congestion in this subway system. The homes and work places of the residents in this city are collected and taken to represent the traffic demand for the subway system of Shenzhen. An origin-destination (OD) matrix derived from the data is used as an input in this method of predicting traffic, and the traffic congestion is presented in simulations. To evaluate the predictions, data on the congestion condition of subway segments that are released daily by the Shenzhen metro operation microblog are used as a reference, and a comparative analysis indicates the appropriateness of the proposed method. This study could be taken as an example for similar studies that model subway traffic in other cities.

A Deep Learning Model with Conv-LSTM Networks for Subway Passenger Congestion Delay Prediction

“…e effectiveness of the method is verified by the operation data of the Beijing Subway. Luo et al [9] proposed a hybrid method, which combines the static traffic assignment model with the agent-based dynamic traffic simulation model to estimate the frequent congestion in the subway system.…”

A Method for Exploring and Analyzing Spatiotemporal Patterns of Traffic Congestion in Expressway Networks Based on Origin–Destination Data

“…However, existing studies pay less attention to the spatiotemporal patterns of traffic congestion for regional expressway networks. Instead, they mainly focus on the traffic congestion of urban roads and have developed various methods to identify [10][11][12][13][14][15][16], predict [17][18][19][20][21], and analyze [22][23][24][25][26][27][28][29] urban traffic congestion. Compared with urban traffic congestion, there are fewer studies that focus on traffic congestion in expressways.…”

Hybrid Solution Combining Kalman Filtering with Takagi–Sugeno Fuzzy Inference System for Online Car-Following Model Calibration

“…Traffic modeling is one of the research domains receiving a higher interest from scientists. Traffic congestion [ 1 ] affects everyone and introduces a dependence on this issue even if we consider the movement of people or the delivery of products or services. In this context, a concept has appeared regarding smart mobility that tries to use ITS (intelligent transportation systems)-specific algorithms to optimize road traffic flow.…”