Friction condition characterization for rail vehicle advanced braking system

• Published in: Mechanical systems and signal processing Vol. 134; p. 106324
• Main Authors: Shrestha, Sundar, Spiryagin, Maksym, Wu, Qing
• Format: Journal Article
• Language: English
• Published: Berlin Elsevier Ltd 01.12.2019; Elsevier BV
• Subjects: Adhesion; Adhesion observer; Algorithms; Braking; Braking systems; Computer simulation; Control algorithms; Control theory; Error compensation; Mathematical models; Optimization; Parameter estimation; Post-processing; Rail brake system; Robust estimator; Simulation; Stability; Variations; Wagons; Wheel slide protection control; Rail brake system; Post-processing; Adhesion observer; Wheel slide protection control; Robust estimator
• ISSN: 0888-3270; 1096-1216
• DOI: 10.1016/j.ymssp.2019.106324

•Wheel slide protection control algorithm eliminating drawbacks of existing methods.•Control algorithm based on multiple-mode shifting during operations.•Accurate online observer to estimate the available adhesion between wheel and rail.•Adhesion force model is modelled using adhesion-creep curve from experimental data.•The proposed approach is compared with the existing approaches. This paper discusses the improvement in rail vehicle braking systems and proposes a new control algorithm for optimum utilization of the available adhesion between wheel and rail during braking conditions. Unlike conventional controllers, the proposed controller algorithm is responsive to the change of operational and environmental parameters between wheel and rail. It is designed based on multiple mode shifting during operations as the parameters change. Furthermore, adhesion is a process which cannot be measured directly; it has to be estimated from related measurable parameters and can be expressed in a quantitative term such as adhesion coefficient. Thus, an online observer to estimate the available adhesion is also proposed in this paper. The observer uses feedback responses to compensate for any error and ensures better accuracy by overcoming steady-state fluctuations and an impractical initial error. For numerical simulation, a wagon model considering longitudinal dynamics is developed. The adhesion force is then modelled by a proper definition of an adhesion-creep characteristics curve that was achieved from the experimental data collected from field measurements. Two sets of simulation were carried out. In the first simulation, the performance of the proposed observer was compared with the existing observer under different adhesion conditions. The comparison suggests that the proposed observer estimates realistic results even under different adhesion conditions as well as during switching between these conditions. In the second simulation, the performance of the proposed control algorithm was compared with the conventional algorithm. The comparison suggests that the proposed control algorithm is able to optimally utilize the available adhesion between wheel and rail to ensure the shortest possible braking distance while maintaining vehicle stability.