RSFT-RBF-PSO: a railway wheel profile optimisation procedure and its application to a metro vehicle

Abstract

With the expansion of urbanisation, more and more metro vehicles shuttle on dedicated railway lines, leading to serious wheel wear and reduced passenger comfort, it is therefore of interest to design a wheel profile that matches well with the dedicated lines. This paper presents a wheel profile optimisation procedure consists of 3 steps. Firstly, taking the LM profile, which widely used in China metro vehicles, as an example, the rotary-scaling fine-tuning (RSFT) method is introduced to generate a large number of candidate profiles. Secondly, to quickly and reliably establish the relationship between the candidate profiles and the objectives obtained by multi-body dynamics simulation (MBS), the radial basis function (RBF) is introduced to reduce the number of MBS runs. Thirdly, the established RBFbased function is treated as the objective function of the particle swarm optimisation (PSO) algorithm to develop a wheel profile that simultaneously considers wheel wear and ride comfort for the B-type metro vehicles. Finally, this paper compares the wheel-rail contact characteristics, critical speed, and simulated long-term wear distribution of the LM profile and the optimised profile named as LMopt. The results show that the LMopt profile has better performance, demonstrating that the RSFT-RBF-PSO procedure is of engineering significance.