Comprehensive evaluation of wheel-rail equivalent conicity status based on an improved combination weighting and cloud model
Qian Jin Xiao, Zhongxu Xu, Chao Chang · Vehicle System Dynamics · 2025
This study addresses the lack of a comprehensive evaluation framework for equivalent conicity in high-speed train wheel-rail dynamics. It proposes a methodology integrating hybrid weighting and cloud modelling to evaluate wheel-rail equivalent conicity status. Fourteen dynamic indices are selected to construct a hierarchical model. Subjective weights are determined using an enhanced Analytic Hierarchy Process (AHP), and objective weights via entropy weighting, combined through an improved game-theoretic approach. Dynamic simulations of a high-speed train under various wheelset equivalent conicity conditions are conducted using Simpack software, based on measured wheel profile data from a maintenance cycle. The dynamic indicators are calculated, and an evaluation grade set is established alongside standard cloud models. For each indicator, a cloud model is generated, and the comprehensive certainty degree of each equivalent conicity scheme is quantitatively assessed using membership degrees based on the Hausdorff distance. Model validation is achieved through the bootstrap resampling method to assess robustness and parameter sensitivity, complemented by an independent field dataset to confirm real-world applicability. This approach delivers a reliable framework for evaluating wheelset equivalent conicity status, enhancing the stability and safety of high-speed rail operations.