Fault Diagnosis Method Research On Axle Box Bearing Of High Speed Train Without Fault Data

The axle box bearing of high-speed train is installed in the axle box,which is the key component for high-speed train to transmit traction.It changes the sliding friction between the wheel set shaft and bearing seat into rolling friction to reduce the friction loss.In the process of train running,the deterioration of the healthy state of the axle box bearing will produce irreversible defects and effect the operation safety,so it is necessary to monitor its health in real time.Because the axle box bearing of high-speed train has the characteristics of unfixed base installation,complicated mechanical background and many exciting sources,so the fault diagnosis research results of axle box bearing of high speed train are few.In order to promote the development of fault diagnosis of axle box bearing of high speed train,a fault diagnosis method for axle box bearings of high-speed trains with health data is presented in this paper.A fault diagnosis method for axle box bearing of high-speed train is presented in this paper.The guiding ideology of this method is the universal information fusion theory.The core content of this method is data preprocessing,feature extraction,feature dimensionality reduction and data modeling for the health data of axle box vibration signals measured by high speed trains under operating conditions.And then obtaining the safety domain boundary and performance degradation data model of the axle box bearing of high speed train.The main innovation of this method is putting forward a set of data "cleaning" method for the vibration data of the axle box of high speed train,presenting a 320 dimensions eigenvector that can fully characterize the healthy state of the axle box bearing,and proposing a method based on t-SNE to reduce the boundary function order of security domain and estimate the boundary of security domain.The advantage of this method is that it overcomes the limitation that the actual fault data is difficult to obtain and avoid studying the complex mechanical structure and mutual influence mechanism between high-speed train components.This method has strong engineering application value.The study of this method can improve the safety of train operation,reduce the maintenance cost,cancel the planned inspection,timely notify the support chain at all levels,and capture the state information in time before the potential catastrophic failure,shorten the maintenance of the entire life cycle of the equipment,which provides technical support for the industrial demand of intelligent operation and maintenance of the axle box bearing of high-speed train.