Abstract

Info

Title:: Prediction model of passenger train running speed onexpressway tunnel section

Author(s):: MA Cong; ZHANG Shengrui; WANG Yaqun; （1. School of Highway, Changan University, Xian 710064, Shaanxi, China; 2. Yunnan TransportationResearch Institute, Kunming 650011, Yunnan, China; 3. Urumqi City Comprehensive TransportationProject Research Center, Urumqi 830063, Xinjiang, China）

Keywords:: traffic engineering; tunnel section; change characteristics of speed; prediction model of speed; model calibration

Abstract:: To improve the safety of tunnel entrances and exits, a prediction model of the speed of a car in a tunnel section was studied. Based on analysis of the continuous change characteristics of the speed of the car in short, medium, and long tunnel sections, the factors that affected the speed were determined. A singlefactor analysis was performed to select the influence speed. The significant factors, such as curvature change rate, curvature, bending slope combination, and circular curve radius were obtained using multiple regressions. The prediction model of the tunnel entrance and tunnel running speed were established, and the model was tested on real vehicle data. The results show that the speed model for the car in the highway tunnel section can accurately predict the optimum limit of speed for each section of the tunnel. The average residual value of the prediction results compared to the actual results is 4.06 km/h. The test residuals conform to a positive distribution, which indicates that the model is effective, and the speed prediction model is thus formulated. This model can be used in highway tunnel sections to provide references for safety reviews and evaluation of speed limits or predictions for special cases for expressways in mountain areas. 7 tabs, 7 figs, 16 refs.