Quantitative assessment of vehicle rollover risk based on TANA method with entropy weight and sensitivity analysis

Abstract

To assess vehicle rollover risk under crosswind conditions, this article establishes a quantitative assessment model based on the Lateral Load Transfer Rate (LTR), and proposes a dimensionality reduction and quantification method that integrates the Entropy Weight Method (EWM) with sensitivity analysis to improve assessment efficiency. Firstly, the EWM combined with the linear weighting strategy derived from global sensitivity analysis is applied to identify the key random variables. Secondly, the Two-point Adaptive Nonlinear Approximation (TANA) model is introduced to replace the traditional Linear Approximation (LA) model, effectively overcoming local oscillation and slow convergence under strongly nonlinear conditions. Finally, the effectiveness of the proposed method is verified through the comparative analysis of the full variables and the reduced variables. The results indicate that the TANA model significantly improves computational efficiency while preserving accuracy. The Entropy-weight and Sensitivity-based dimensionality-reduction strategy enhances computational efficiency by approximately 82% compared to the full-variable LA model, providing theoretical and technical support for the efficient assessment of vehicle rollover risk under crosswind conditions.