Tyre/Road Friction Estimation And Optimal Brake Control Using Dynamic Friction Model

Antilock braking systems (ABS) are closed-loop control devices implemented in ground vehicles that prevent wheel lock-up and improve brake performance during braking. Current ABS controller is built upon iron bird test, which leads to trial and error. Besides, military transport landing procedure is complicated, which make the brake condition varies and unpredictable. It is noneffective to count on iron bird test for designing high performance brake control system. Current control strateges are partially improvement of traditional control stratage, the brake system can still have a tendency to over brake and become unstable when encounters varing brake condition and low friction runway. Considering tyre/road friction plays fundermental role in brake system performance, this thesis proposes an ABS controller designing method using tyre/road dynamic friction model, which lead to higher performance and less iron bird test needings.This thesis addresses the controller designs for ABS in vehicles. Simplified quarter vehicle models with special emphasis on the modeling of tire and dynamic tire/road interaction is utilized to develop and test the proposed controllers. The novelty of the present research is in the utilization of a new dynamic friction tire model for the development and testing of designed controller. Due to complex mechanics of tires, the dynamic friction tire model is significantly more realistic than that of commonly used static friction tire model. Before optimal brake controller and road/tyre friction observer design, a review of recent developments and trends in tyre friction modeling is presented. Further more, a detailed comparison of empirical (semi-empirical) models and analytical models is carried out to demonstrate the effectiveness of the introduced dynamic tire friction model - LuGre friction model. Using this dynamic tire model, a sliding mode scheme for optimal braking and estimation of road/tyre friction is proposed in this thesis. By this means, the difficulty associated with on-line search of the optimal slip and maximum coefficient can be easily overcome, which solves a commonly existed problem conventional ABS controller design. Extensive simulation studies have been conducted for the developed controller and observer to demonstrate the effectiveness of the proposed scheme. Simulation result indicate that the action of the sliding mode controller implemented in the single wheel dynamic model is able to provide the optimal control and can reduce the tendencies of over brake in extreme emergency situations for the specific road conditions. The investigation further demonstrates the effectiveness and convenience of utilizing dynamic fiction tire model for the development of ABS controllers.