Adaptive Tire Model For Dynamic Tire-Road Friction Force Estimation

• Main Author: Spike, Jonathan
• Language: en
• Published: 2013
• Subjects: Vehicle Dynamics; Vehicle Tire Estimation; Vehicle Tire Forces; Tire Model
• Online Access: http://hdl.handle.net/10012/7983

As vehicle dynamics research delves deeper into better insights in performance, modeling, and vehicle controls, one area remains of utmost importance: tire and road friction forces. The vehicle???s interaction with the road remains the dominant mean of vehicle control. Ultimately, the tire-road interaction will determine the majority of the vehicle???s capabilities and as the understanding of the interface improves, so too can the performance. With more computationally intensive systems being instrumented into modern vehicle systems, one is able to observe a great deal of important vehicle states directly for the remaining vehicle information; excellent estimation techniques are providing the rest of the insights. This study looks at the possible improvements that can be observed by implementing an adaptive dynamic tire model that is physical and flexible enough to permit time varying tire performance. The tire model selected is the Average Lumped LuGre Friction Tire Model, which was originally developed from physical properties of friction and tire systems. The material presented here examines the possibility of an adaptive tire model, which can be implemented on a real-time vehicle platform. The adaptive tire model is just one section of an entire control strategy that is being developed by General Motors in partnership with the University of Waterloo. The approach allows for estimated and measured vehicle information to provide input excitation for the tire model when driven with real-world conditions that enabling tire estimations. The tire model would then provide the controller information indicating the expected tire capacity and compares it with the instantaneous loading. The adaptive tire model has been tested with flat road experimental cases and the results provided reasonable estimates. The experimentation was performed with a fully instrumented research vehicle that used in-wheel force transducers, and later repeated with a completely different non-instrumented fully electric vehicle. The concepts and investigation presented here has initiated the ground work for a real-time implementation of a full adaptive tire model. Further work is still required to evaluate the influence of a range of operating conditions, tire pressure, and of different tire types. However, the findings indicate that this approach can produce reasonable results for the specified conditions examined.