高铁轴箱轴承是高铁动车组的关键核心零件，一旦发生故障，将会造成重大的人身安全和财产损失等事故。动态可靠预测轴箱轴承的寿命，是避免突发故障发生进而保障动车组安全运行的关键。寿命动态预测的难点是准确表征轴承性能特征参数演化过程，建立性能特征参数与寿命间的关系。本项目采用时频域特征提取方法，提取轴承振动信号性能特征参数；研究趋势项、随机误差项以及测量误差与个体差异性不确定影响因素的表征模型，阐明轴箱轴承性能特征参数演化规律；研究性能特征参数增量统计特征，建立基于广义Wiener过程的轴箱轴承性能特征参数演化模型；采用EM-遗传算法，提出同批次轴承历史数据和服役轴承性能监测数据融合方法；引入增量随机滤波方法，构建演化模型状态更新与预估机制，提出面向全寿命周期的轴箱轴承服役寿命动态预测方法，通过仿真和试验验证与修正本项目方法。预期成果为高铁轴箱轴承服役寿命高精度预测和动车组安全运行提供新的理论支撑。

Axlebox bearing, as the key core part of high speed train, occurred fault, it will cause major personal safety and property losses and other accidents. It is the key that dynamically and reliably predicting the service life of axlebox bearing for high speed train to avoid sudden accidents and guarantee the security operation of high speed train. The difficulty of dynamic life prediction is to accurately describe the evolution process of axlebox bearing performance characteristic and establish the relationship between performance characteristic and life span. In this project, the time-frequency domain feature extraction method is firstly adopted to extract the performance characteristic from the vibration signal of axlebox bearing. To illuminate the evolution law of the performance characteristic of axlebox bearing, the representation models of the trend term, the random error term and the uncertain affecting factors involving the measurement error and the individual difference are studied respectively. Then the performance evolution model of axlebox bearing is established based on the generalized Wiener process by studying the statistical characteristics of performance characteristic increment. An information fusion method is proposed to fuse the historical data of the same batch of bearings and the performance monitoring data of the service bearing based on the EM-genetic algorithm. Moreover, the state updating and predicting method is constructed for the generalized Wiener process evolution model based on the incremental stochastic filtering theory. Then, a dynamic service life prediction method is proposed for the whole life cycle of the high speed railway axlebox bearing. Finally, the proposed method of this project is verified and modified by the simulation study and engineering experiment. The prospective research can provide a new theoretical basis for high precision and dynamic prediction of the service life of axlebox bearing and the security operation of high speed train.