TABLE OF CONTENTS This Chapter describes various approaches for the estimation and observation of variables which are not directly measurable. Section 9.1 presents two methods of obtaining the vehicle velocity in the inertial co-ordinate system, a Kalman filter approach and a fuzzy estimator. In Section 9.2 these methods are also employed for the estimation of the yaw rate. In Section 9.4, various approaches for estimating the friction characteristics, and the mass moments of inertia. In Section 9.5, approximation formulas are given for the wheel ground contact forces. The tire side slip constants are adapted with a simple nonlinear approximation equation. Based on the wheel ground contact forces, the roll and pitch angles are approximated. In Section 9.6 the vehicle body side slip angle is estimated using a nonlinear observer. Section 9.7 presents two methods for road gradient estimation.
The vehicle velocity v CoG is obtained via a fusion of the data from all rotational wheel velocities v Rij and the longitudinal acceleration sensor. Via integration of the acceleration a fifth estimate for the vehicle velocity is made available. The estimation must be very accurate, as a basis for the wheel slip calculation (see Section 8.3.2). Some systems only select the maximum rotational wheel speed as the estimate for the vehicle velocity. When all four wheels happen to lock simultaneously, this approach is very inaccurate.
Two alternative estimation methods for the vehicle velocity are regarded, the Kalman filter and the fuzzy estimator.
All sensors...
