TIRE SPINNING DETECTION AND VEHICLE VELOCITY ESTIMATION USING KALMAN FILTER
HOLUB, O.; AUGUSTA, P.; ROUBAL, J.; HURĮK, Z.
Abstract
Track vehicles, besides their advantages like low tractive resistance, have crucial disadvantage comming from the low adhesion coefficient between rail and metal wheel. Thus total tractive or braking force, that the drive can provide, often cannot be transferred from the wheel to the rail. During the acceleration this leads to slipping while during the braking sliding occurs. In both cases distance covered by the wheel perimeter differs from the true distance driven by the vehicle. Identification of this phenomenon can be based on various techniques. The method used within this article is based on thresholding of the difference of measured velocity. The difference is sensitive to sensor noise. Influence of the noise is supressed by Kalman filter, whose performance is further improved via smoothing. In order to obtain best estimate of the true vehicle velocity, all the available measurements of particular axle velocities are utilized. Applicability of this method is demonstrated on real-world data from tram operation.
Coresponding author e-mail: holubo2[at]fel[dot]cvut[dot]cz
Session: Applications and Case Studies
