Algorithm for inverse determination of derailment coefficient by using instrumented wheelsets

Abstract

The main indicator of the running safety of railway vehicles is the ratio between lateral and vertical forces in the wheel-rail interaction, known as derailment coefficient. Its exact determination has a huge significance in the development and certification phases of railway vehicles. The most reliable determination of wheel-rail interaction forces is based on experimental testing, by using instrumented wheelsets, which is otherwise prescribed in appropriate international regulations. This paper presents a unique algorithm for inverse determination of wheel-rail interaction forces, i.e., derailment coefficient, based on the measurement signals obtained from instrumented wheelsets. The blind signal separation (BSS) and method of independent component analysis (ICA) are applied. Verification is carried out on the example of the wheelset of freight wagon, based on the wheel's FEM model. The obtained results confirmed a high efficiency and accuracy of the developed algorithm, whereby estimated error of inverse identification is less than 5%.