Information fusion for vehicular systems parameter estimation using an extended regressor in a finite time estimation algorithm

In this paper, we present an extension to a recently developed continuous-time, finite-time parameter estimation structure to perform data fusion. The regression elements of the finite-time algorithm are used to carry additional information. Their parameters are also parameters in the dynamics of additional sensors. This additional information will help in estimating these parameters. The algorithm can easily augment an adaptive observer. This new data fusion structure is employed in the context of vehicle mass and road gradient estimation. The estimator in its original form makes use of vehicle speed over ground and driving force information and the modification is demonstrated with the inclusion of an accelerometer aligned to the longitudinal direction in the vehicle frame of reference. Such an accelerometer could be part of the array in an onboard IMU like those used to control vehicle safety systems, whose outputs are broadcast on the onboard vehicle CAN bus. The modified algorithm has been tested with practically relevant data, confirming that the new technique produces numerically correct results and significantly improves parameter estimates over the algorithm in its existing form.