Lane-Changing Risk Analysis in Undersea Tunnels Based on Fuzzy Inference

Abstract

Lane-changing in undersea tunnels has a negative impact on the normal traffic flow, and even lays hidden dangers for the occurrence of traffic accidents. Lane-changing behavior in undersea tunnels was divided into free, compulsory, and collaborative lane-changing types according to the characteristics of traffic flow to explore lane-changing risk in undersea tunnels. A fuzzy inference analysis on the three lane-changing behaviors was conducted on the basis of the behavior characteristics of fuzzy uncertainty of drivers. The most representative influencing variables, including speed difference, initial space of vehicles, traffic density, and distance for minimum lane-changing, were selected as fuzzy input variables, and lane-changing risk was used as an output variable to construct fuzzy rules for different lane-changing behavior. Risks of the three lane-changing behaviors were simulated by MATLAB/Simulink. Results demonstrated that the compulsory lane-changing in undersea tunnel was the riskiest, followed by collaborative and free lane-changing. Slope considerably influenced lane-changing risk. Specifically, the lane-changing risk at the downhill section was the highest, and the lane-changing risk at the uphill section was the lowest. The lane-changing risk at the flat section was between them.

Keywords

• Fuzzy inference
• risk analysis
• undersea tunnel
• vehicle lane-changing