| Abstract |
This paper describes a fuzzy identification model for a rail traction load simulator utilising the prediction of variables causing changes in the friction-creep curve. The model has fuzzy inputs corresponding to rail conditions (presence of water, sand, oil and solid debris) and crisp inputs for vehicle and wheel speed, weight and time. After calculation of reference values for the friction-creep curve parameters, the model makes adjustments according to the values of intermediate variables
corresponding to contact pressure and creepage. The final friction-creep curve is obtained by interpolation. Results are presented of friction-creep curves under the influence of sudden changes in
water, oil & sand contamination and train speed. |
