A Thermo-Wear Dual-Parameter Wiener Model Approach for Predicting the Remaining Useful Life of Wet Friction Components

Abstract Wet friction components perform the critical function of transmission system and controlling torque within clutches, and their lifespan directly determines the stability and safety of wet clutches. Moreover, their primary failure modes are wear failure and thermal failure. In this context, this study conducts accelerated life testing on wet friction components to collect multi-dimensional degradation data and identifies corresponding failure thresholds. Subsequently, we construct the temperature and wear degradation parameters by integrating subjective and objective weighting approaches, enabling a comprehensive characterisation of the component's degradation state. Thereafter, the Copula function is employed to analyse the dependency structure between the two degradation parameters, leading to the derivation of the joint probability density function of the remaining useful life (RUL). Finally, based on the inverse Gaussian distribution, a dual-parameter wiener degradation model is developed. The model parameters are estimated using maximum likelihood estimation and Bayesian updating techniques, enabling effective RUL prediction for wet friction components. The results demonstrate that, within the critical degradation phase of 88 h-110 h, the dual-parameter wiener degradation model yields a root mean square error (RMSE) of 3.053, which is significantly lower than those of the single-parameter models. In this stage, the prediction error is reduced by up to 46.83%. Furthermore, 72.73% of the model’s prediction errors fall within the 0-10% range. Therefore, these findings confirm the accuracy and advantage of the proposed model in predicting the RUL of wet friction components