Abstract: To address the insufficient lateral stability of single-axle straddle-type monorail vehicles during operation, a specific type of single-axle straddle-type monorail vehicle was taken as the research object. Based on the Adams multibody dynamics vehicle model, an Adams/Simulink co-simulation environment was established. The SAC algorithm was integrated with the dynamics model of the single-axle straddle-type monorail vehicle to construct an intelligent active control framework, and the corresponding reward function and constraints were designed. Random road excitation generated from the Japanese track spectrum and finger-shaped road excitation were used as inputs to conduct control strategy training and simulation verification under different vehicle speed conditions. The simulation results show that, compared with the passive state and Skyhook control, SAC control can reduce key evaluation indicators such as vehicle body lateral acceleration, yaw angular acceleration and roll angular acceleration, thereby improving the lateral stability of the vehicle body. Meanwhile, SAC control exhibits good stability and control performance under different vehicle-speed conditions.

Key words: vehicle engineering; single-axle straddle-type monorail vehicle; active control; SAC algorithm

摘要： 针对单轴跨座式单轨车辆运行过程中横向稳定性不足的问题，以某型号单轴跨座式单轨车辆为研究对象，基于Adams多体动力学车辆模型，搭建Adams/Simulink联合仿真环境。将SAC算法与单轴跨座式单轨车辆动力学模型相结合，构建车辆智能主动控制框架，并设计相应的奖励函数与约束条件。采用基于日本轨道谱生成的随机路面激励和指形板路面激励作为输入，开展不同车速工况下的控制策略训练与仿真验证。仿真结果表明：与被动状态和Skyhook控制相比，SAC控制能有效降低车体横向加速度、车体横摆角加速度和车体侧倾角加速度等关键评价指标，提高车体横向稳定性。同时，在不同车速工况下，SAC控制均表现出较好的稳定性和控制效果。

关键词: 车辆工程；单轴跨座式单轨车辆；主动控制；SAC算法