Optimization and Dynamic Simulation Model of Aerodynamic Structure in High-Speed Train Tunnel Based on Fuzzy Bayesian Network

High speed trains passing through tunnels can easily induce aerodynamic disasters such as micro pressure waves at the tunnel entrance, aerodynamic resistance in the vehicle tunnel, and severe pressure oscillations inside the tunnel. Traditional static CFD optimization is difficult to adapt to parameter uncertainty, multi-objective coupled game theory, and the challenge of adapting to vehicle speed conditions and spans. This article clearly distinguishes fuzzy Bayesian networks (FBN) from conventional deterministic CFD single objective static optimization methods. Based on fuzzy mathematics to quantify the random uncertainty of aerodynamic boundaries and structural dimensions, a Bayesian network multi index aerodynamic performance decision-making framework is constructed to achieve optimal combination of vehicle tunnel structural parameters. Combined with FLUENT-SIMPACK joint dynamic simulation to complete the verification. The current simulation benchmark speed is 350km/h, and optimization adaptation is carried out for the iterative requirements of 400km/h high-speed train engineering. The simulation results show that the optimization scheme of streamlined front end combined with tunnel buffer structure can reduce the peak pressure of the tunnel by 27.3%, the aerodynamic resistance of the whole vehicle by 19.5%, and the micro pressure wave at the entrance of the tunnel by 31.2%. The smoothness of vehicle operation and passenger comfort are significantly improved. This model addresses the limitations of traditional CFD optimization with deterministic assumptions and the weakness of multi-objective solving, clarifies the innovation boundary of uncertain decision-making, and is suitable for the refined engineering design of high-speed vehicle tunnel structures.