An Extended State Space Model and a Kalman Filter Approach

The purpose of this project is to develop a model based observer that estimates the road bank and the road slope angle. An extension of the bicycle model has been used to provide estimations of the respective angles. In addition, a steady state identifier is used to provide an estimation of the bank angle when the vehicle is in a steady state mode. Limited information and approaches for estimating the contribution from the road bank and slope angle to the lateral and longitudinal acceleration have been found in the literature. Neglecting these contributions could yield a false activation of a vehicle’s stability control systems when driving on a banked or sloped road. A vehicle stability control system compensating for the bank and slope contribution would have an enhanced capability to control the vehicle’s behavior. To provide the most accurate vehicle state estimation the extended state space model is combined with a Kalman filter to reduce noise and further improve the estimates. This model provides an estimate of the road bank and as well of the road slope angle. The extended state space model is an extension of the commonly used bicycle model with a separate bank angle state, where the description of the bank angle is implemented as a first order low-pass filter. The use of a linear tire model could restrict the estimation validity, to overcome certain aspects of this issue the bank angle is provided by solving the lateral kinematic equation in steady state for the bank angle. A number of different driving scenarios have been simulated in CarSim to obtain vehicle sensor signals which have been used to analyze and evaluate the proposed system offline. To further evaluate and test the algorithm, several test runs were made at Saab’s test track in Trollhättan, Sweden. At a test from Saab’s test track the measured bank angle contribution made up 50 % of the measured lateral acceleration. At the same test the algorithm estimated the contribution to be 36 % of the measured lateral acceleration. Also, the algorithm was able to detect a change in the bank angle from a positive to a negative angle. The detection and estimations of the road bank and slope angle could be utilized to compensate the lateral and longitudinal acceleration used in the vehicle stability control system. A compensation of these contributions would further the performance of the stability system