基于智能体理论的空气悬架车身高度智能控制系统研究

Research on Air Suspension Vehicle Height Intelligent Control System Based on Agent Theory

为了进一步发挥空气悬架车身高度调节系统的性能,在Belief-Desire-Intention(BDI)框架下构建了目标车身高度控制智能体,并采用汤普森抽样算法构建智能体学习行为。结合车身高度调节系统模型,建立空气悬架车身高度智能控制系统。单一工况下的仿真结果验证了智能体学习行为的可行性以及学习结果的适用性;混合工况下的仿真结果验证了空气悬架车身高度智能控制系统的可行性和有效性。结果表明:在车身高度智能控制系统的控制下,簧上质量质心位置处的加权加速度均方根值上升了0. 45%,侧倾因子降低了22. 82%,在不恶化行驶平顺性的同时,提高了操纵稳定性。

In order to further improve the performance of air suspension vehicle height adjustment system, the target height control agent was constructed under the framework of the Belief-Desire-Intention (BDI)model, and the agent learning behavior was constructed through Thompson sampling algorithm. The vehicle height intelligent control system was constructed by combining the traditional vehicle height adjustment system. The simulation results under a single working condition verified the feasibility and effect of reinforcement learning. The simulation results under mixed conditions verified the feasibility and effectiveness of the air suspension vehicle height intelligent control system. The results show that under the control of the intelligent control system of vehicle height, the root mean square value of the weighted acceleration at the position of mass center of spring rises by 0.45%, and the roll factor decreases by 22.82%. The control stability is improved without deteriorating the ride comfort.