A load estimation model for axle box bearings of high-speed trains based on vehicle dynamic response
Yang Chen, Xingwen Wu, Maoru Chi, Wubin Cai, Zikui Ma, Xuesong Yao
Abstract
The axle box bearing is a critical component of high-speed train bogies, closely related to vehicle safety. Accurate bearing boundary load is crucial for bearing fatigue life assessment. This paper derives the dynamic mechanism equations of a novel axle box bearing-half vehicle system in detail and develops a bearing load estimation model for straight scenarios based on the augmented Kalman filter algorithm and mechanism equations. Evaluation indices for the load estimation effectiveness are formulated by comparing estimated results with those calculated by a virtual test model. Subsequently, an improved niche genetic algorithm combined with the virtual mapping relationship is proposed to optimize and match key parameters of the estimation model. Validation and analysis of the load estimation model under different operating conditions are conducted. The results suggest that the estimated vertical, lateral, rolling, and yaw loads of the eight axle box bearings in the entire vehicle are consistent with those calculated by the virtual test model, with correlation coefficients above 0.9. The proposed estimation model has good applicability for identifying the service load of axle box bearings under different measured wheel and rail irregularities, carbody weights, and vehicle speeds, which has important implications for monitoring the boundary load of high-speed train axle box bearings.