Mass estimation using mapped road grade data in heavy duty vehicles

Abstract: One way of achieving more fuel efficient, more environmentally friendly and user-friendly Heavy-duty vehicles is to develop new embedded control features. Many of these features are based on vehicle mass is known. Methods of finding the current vehicle mass can be implemented in many different ways, including various methods that require user interaction or manual weighing of the vehicle. These should be avoided as they are inefficient and add operations to the driver. Another alternative method is estimation by adaptive filters. This examination is based on this method and assumes that the road grade is known. The method developed consists primarily of a recursive least squares method to estimate against a physical vehicle model in the longitudinal direction. The time-varying input signals are noise reduced by stepwise integration into intervals of 10 seconds and low-pass filtering. For estimation to be carried out, a number of conditions have been set. Verification of the method was carried out through both simulations and by executing it in a vehicle's control system. The results show that the mass is estimated with a relative error of 5 % after 600 seconds of driving. The conclusions include that this method gives a good estimation and it does not stress the vehicle control system so that it becomes unusable. Before using this method, more work should be performed on getting a more accurate model of vehicle, especially on the parameters used in this.