ABSTRACT
ABSTRACT: With increasing automation of driving tasks, reliable information on the road conditions is required to ensure safe automated transport. A method is presented to determine the maximum coefficient of friction between tire and road surface during driving based on measurements of the vehicle’s dynamic state using a model-based approach. Particle filtering is applied as a state estimation technique which is able to consider measurement noise and model uncertainties within a Bayesian framework. Almost only standard sensors as installed in a vehicle with electronic stability control (ESC) are used. The sensor information required includes wheel speeds, longitudinal and lateral chassis acceleration, the yaw rate, the steering wheel angle and the longitudinal velocity. Using real vehicle measurements, possibilities and limitations of the presented approach are discussed. It is demonstrated that the tire and road conditions can be estimated in many driving conditions with a high confidence.
