Research on vibration control of intelligent suspension for autonomous vehicles

To enhance ride comfort and safety in autonomous vehicles, a suspension control strategy based on multi-sensor information fusion is proposed. A multi-sensor processing system integrates light detection and ranging (LiDAR) and cameras, utilizing point cloud clustering, segmentation techniques, and the You Only Look Once version 7 (YOLOv7)-tiny algorithm to analyze the geometric characteristics and spatial positions of road obstacles. The chain code method is applied to calculate obstacle areas. A vibration model is established to describe the relationship between road roughness and vehicle vibrations, followed by the development of an optimal damping ratio model for intelligent suspension control. A fuzzy neural network algorithm is employed to pre-adjust suspension damping based on obstacle characteristics and vehicle speed, enabling proactive adaptation to road irregularities. If an obstacle exceeds a predefined threshold, the system automatically reduces vehicle speed to further enhance safety and ride comfort. Experimental validation demonstrated that for potholes of 0.5 m 2 with depths of 0.06, 0.08, and 0.10 m, vertical vibration acceleration was reduced by 11.4%, 12.8%, and 15.9%, respectively. These results confirm the effectiveness of the proposed approach in improving ride quality and vehicle stability.