Slip and Slide Detection and Compensation for Odometery System, Using Adaptive Fuzzy Kalman Filter

Detection and compensation of the slip and slide events, are important factors in providing an effective traction and brake control functions and also in improving the accuracy of the speed and travelled distance data, provided by the train odometer system, respectively. In this paper Kalman and adaptive Fuzzy-Kalman filter algorithms have been designed, implemented and tested for these purposes. For detection and tuning of the filter, in addition to the filter residual, the dynamics of the train, in the form of train acceleration, is considered. The results show that this is quite effective in not only detecting the slip and slide events, as has been the subject of most research in this field, but also in improving and correcting the train speed and positioning data, which is very important for train navigation purposes. A comparison of the results show that the proposed algorithm is more effective in correcting the errors introduced by slip and slide of the wheels in acceleration and deceleration phases of the train movement. The algorithm has been evaluated by implementing on a real train positioning system.