Regional Traffic Flow Prediction Model Based on Mobile Signaling Data- A Case Study of Chongqing City

With the rapid development of the Internet and the Internet of Things, we can obtain the real-time and massive transportation data. The transportation data can be used to predict the traffic flow trend to reduce traffic congestion, both from the perspective of road users and policy makers. Combined the real-time and massive transportation data with the real-time and massive mobile phone signaling data, the traffic flow trend can be more accurate and practical. The dataset used in this paper consists of two days mobile signaling data from all regions in Chongqing, which was utilized to generate and evaluate the models. In this study, a neural network model is proposed to predict spatio-temporal traffic flow based on mobile phone signaling data in Chongqing, and the model is evaluated by both predicted data and observational data. The proposed neural network model can help road-users obtaining the probability of congestion in a certain area at the next moment in real time, alleviating the generation of congestion at source, and helping traffic management departments to formulate macro policies in a more scientific manner.