Multi-objective optimization of a steering mechanism for a virtual rail train based on spatial kinematics and particle swarm optimization

Virtual rail trains (VRTs) are multi-module, multi-axle, all-wheel-steering vehicles. During turning maneuvers, the ideal steering angle relationships are highly dependent on steering control strategies and the selection of trajectory reference points. However, existing VRT steering mechanism designs fail to fully integrate vehicle steering control strategies, leading to significant deviations between the actual and ideal steering angles during turns, which exacerbates abnormal tire wear. This study focuses on a three-module eight-axle VRT. A spatial mathematical model of the steering mechanism is constructed to derive the actual steering angle relationships of the wheels. Simultaneously, the ideal steering angle relationships are calculated based on the Ackermann steering principle and vehicle steering control strategies. A particle swarm optimization (PSO) algorithm is employed to perform multi-objective optimization of the steering mechanism, aiming to minimize the error between the actual and ideal steering angles. The optimization results demonstrate that the redesigned steering mechanism significantly reduces steering angle deviations, bringing the wheels closer to a pure rolling state. This study provides an effective methodological reference for the optimization of VRT steering mechanism designs.