A Method to Estimate Critical Tire Properties Using Non-Linear Tire Models

This paper describes a method to estimate lateral and longitudinal cornering stiffness, as well as the maximum tire force. Knowledge of these parameters can be critical for certain vehicle control systems that need to understand the tire’s limits, especially during turning maneuvers and acceleration. In this study, an extended Kalman filter is used along with the Dugoff or Fiala tire model to estimate these parameters. The pros and cons are discussed for estimation with both the Dugoff and Fiala tire models. An inertial measurement unit (IMU), a dual GPS antenna, and wheel speed sensors implemented on a test bed are used to evaluate the performance of the algorithms. A transformation matrix derived from a bicycle model is used to convert acceleration measurements into force measurements. The force measurements are then fed into the extended Kalman filter to estimate the parameters. Although this algorithm is post-processed, it can easily be used in real-time estimation. The experiment is performed on an asphalt surface to test its performance. The effects of vehicle roll can be significant for vehicles that exhibit large roll angles, but it is assumed to be negligible in this study.