Low adhesion estimation

Low adhesion in wheel-rail contact causes huge problems for scheduling and safety of railway networks. Braking and guidance forces have the same force generation mechanism, via the creep forces at the wheel-rail contact. This paper demonstrates the application of a Kalman filter to estimate the adhesion condition on the rails. The key aspect of this approach is that estimation is carried out on-line, during normal running, i.e. before the brakes have been applied. This will give a full network-wide picture of low adhesion conditions. The results show that the different adhesion levels cause different responses in the railway vehicle, and that a Kalman filter can estimate wheelset forces well, to give an indication of the adhesion level. The work is successfully applied to a Simulink half vehicle model.