Semi-active and active suspension control with road preview

This thesis presents the four-year work on improving vehicle ride and handling with vertical vehicle control between 2019 and 2023 as part of Groupe Renault's Vehicle Motion Control Project. The comfort expectation and competition in today's vehicles are increasing with the involvement of many automobile manufacturers in the market. Passive suspension systems have limited performance, since there is only one actuator that dampen the body and wheels, thus, the damper has to be calibrated with a constant damping coefficient considering the balance between the body and the wheels. Electronically controlled dampers with variable damping ratio have been developed to improve damper performance. First, two–stage, then an improved continuously variable dampers were developed, and this system made possible to change the vehicle characteristics between comfort and sport. Active suspension systems are generally developed by manufacturers for high-end vehicles. On the other hand, semi-active suspension systems are expected to be integrated into lower class vehicles because of its lower cost compared to active systems and depending on market demand. Thus, the importance of these systems in terms of competition is increasing. Studies on semi-active dampers have long been of interest not only to the automobile industry but also to researchers in academia, and plenty of studies have been presented in the field. Ever-increasing electronic components in vehicles nowadays increase the number of sensors in vehicles, and the interaction of steering, braking, four-wheel control, conventional or electric engines' traction systems with semi-active dampers increases the importance of this field. Advanced driving assistance systems made it possible to detect the environment of the vehicle as well as the road profile in front of the vehicle and to control the suspension systems with this preview information. At the beginning of this study, feedback control methods were examined and compared with each other in terms of their performance and applicability to the vehicle. Later in the tests, road data was acquired with a front camera. For efficient evaluation of raw data that produced by the camera, the DBSCAN method were proposed and compared with the existing method. Model Predictive Control (MPC) method with semi-active dampers were applied to the full vehicle model and compared with Skyhook control methods. Then, a 7-DoF virtual vehicle model was moved at the same speed as the vehicle on the camera-detected road profile. Thus, the outputs of virtual vehicle were examined to be used in prioritizing the vehicle motions such as bounce, pitch and roll. In the last section, SH-like derivative MPC controller was introduced to control the vehicle body by controlling global forces and moments instead of forces on the actuators. Calculated forces and moments around the CoG were distributed to the dampers by the IDT method. Semi-active and active control methods were co-simulated using high fidelity IPG® Carmaker® models and Matlab®. The simulation results showed that the preview and feedback model performances were similar. The results of the MPC were promising, and it was deduced that a more accurate vehicle model, tyre model and road profile information may be needed to improve its performance.