Fatigue Life Prediction of Cableway Wire Rope: Integrating Stress Analysis \& Gradient Boosting Regression Tree

This article proposes an intelligent prediction method that integrates stress analysis and gradient boosting regression tree (GBRT) for predicting the remaining life of cableway wire ropes under complex loads and environmental coupling. Firstly, based on the S-N curve and Miner's rule, a physical life model considering stress amplitude and load cycle is established. Furthermore, the gradient boosting regression tree method is introduced to construct a data-driven life correction model with multidimensional features such as stress distribution, load history, cumulative damage, and environmental factors as inputs, in order to compensate for the shortcomings of traditional physical models in nonlinear damage accumulation, material discreteness, and timevarying environmental factors. The experimental results show that the proposed fusion method significantly reduces the prediction error under various load conditions, verifying its effectiveness and engineering applicability in steel wire rope health monitoring and predictive maintenance.

Keywords: Steel wire rope, S-N curve, fatigue life, gradient boosting regression tree, remaining useful life, physics-informed machine learning.