Traffic demand resilience analysis in highway tunnels based on equilibrium allocation model

• This paper proposes a highway tunnel traffic resilience method with equilibrium allocation and stochastic utility theory. • Builds tunnel traffic resilience model with efficiency and risk functions to solve residual capacity of congested sections. • Validated by expressway toll data, the method has tunnel residual capacity error <10%, queue accuracy 83.05% and 93.62%. • Maintains high analysis precision under abnormal conditions to support tunnel traffic regulation, resource allocation and stable operation. Due to the uncertainty of travelers’ choices and the complex factors affecting the traffic conditions in highway tunnels, it is difficult to accurately grasp the changing patterns of traffic demand, which in turn affects the accuracy of traffic demand resilience analysis. Therefore, this article proposes a tunnel traffic demand resilience analysis method based on the equilibrium allocation model to improve the operational stability of the tunnel traffic system in emergency situations. This method comprehensively considers key congestion influencing parameters such as traffic flow, vehicle speed, vehicle type composition, tunnel length and slope, and combines stochastic utility theory to construct a stochastic equilibrium allocation model to characterize the random selection behavior of travelers and achieve tunnel traffic allocation under elastic demand. On this basis, a tunnel traffic demand resilience analysis model was established by constructing a traffic efficiency function and an operational risk function, which can accurately solve the remaining capacity of congested road sections. To verify the effectiveness of the proposed method, actual toll data from the highway network in J province were selected for case analysis. The research results show that this method has an error of less than 10% in the analysis of remaining traffic capacity in congested tunnel sections, with an overall queue length accuracy of 83.05% and a maximum queue length accuracy of 93.62%. In addition, this method can maintain high accuracy in flow demand analysis under abnormal working conditions, providing a reliable basis for dynamically adjusting traffic flow distribution, optimizing traffic resource allocation, and ensuring stable operation of tunnel traffic systems. It has important theoretical and practical value.