Backlash gap position estimation in automotive powertrains

Backlash in automotive powertrains is a major source of drive-ability limitations. In order to increase the powertrain controller performance, knowledge of the backlash properties (size and current position) is needed. In this paper, a nonlinear estimator for the current angular position in the backlash is developed, based on extended Kalman filtering theory. A linear estimator for fast and accurate estimation of the angular position of a wheel and the engine is also described. It utilizes standard ABS sensors and engine speed sensors, and is based on event based sampling, at each pulse from the sensors. The results show that the backlash position estimate is of high quality, and robust to modeling errors. The performance is increased further when the event based position estimators are used as pre-filters.