Fusion of Q-learning multi-mode control and air pressure regulation for composite suspension system

The air suspension system is essential for regulating ride comfort in intelligent vehicles, directly influencing driving quality. To enhance ride comfort across diverse conditions, this paper presents a novel composite suspension system architecture. First, the vibration mechanism of the vehicle’s composite suspension is analyzed, and a model is constructed. Secondly, an adaptive control strategy based on Q-learning is proposed to regulate multiple strategies and modes, aiming to meet the driving requirements under complex operating conditions. Additionally, a normal distribution-based average arithmetic filter is incorporated to regulate air pressure signals, integrated with the model predictive control (MPC) system to balance vehicle vertical acceleration and air spring frequency. Finally, the performance of this adaptive composite suspension system is evaluated under different conditions using ride comfort metrics to confirm its effectiveness. Results demonstrate that the proposed system effectively adjusts air spring pressure across varied conditions, reduces spring wear, and maintains high ride comfort, achieving an 9%–16% improvement.