Dissertation
Dissertation > Transportation > Rail transport > Locomotive works > Depot, locomotive maintenance and repair

EMU bogie bearing reliability analysis and fault diagnosis technology research

Author NieZuo
Tutor ZhouChunHua
School Central South University
Course Transport equipment and Information Engineering
Keywords EMU bogie bearings Reliability Analysis EMD BP neural network Improved Genetic Algorithm
CLC U269
Type Master's thesis
Year 2011
Downloads 202
Quotes 0
Download Dissertation

EMU efficient means of transport to small groups , large density intercity and suburban railway . Bogie is a high-speed the EMU running gear , decided to train operators speed and running quality . EMU bogie bearings working conditions affecting the safety of railway transport one of the important factors to carry out EMU bogie bearing reliability analysis and fault diagnosis , to ensure the safety of the operators , improve maintenance efficiency and avoiding unnecessary losses have important significance . In this paper, the fault tree analysis method to establish the EMU bogie bearing failure model , and improve its reliability requirements , and a brief introduction EMU Bogie bearing vibration mechanism , fault characteristic frequency . Bearing fault monitoring technologies , vibration monitoring techniques to monitor EMU bogie bearings , and in-depth study of more advanced theories and methods in the field of fault diagnosis . In this paper, two methods of bearing failure diagnosis and monitoring . A time-frequency domain parameters of diagnostic methods , another method is : intelligent diagnostic method , the first vibration signal wavelet packet denoising improve its signal-to-noise ratio , and then based on EMD ( Empirical Mode Decomposition ) method to extract energy characteristics of bearing failure , the fault signal is decomposed to the IMF, the analysis of several important IMF , each IMF component , as BP neural network input vector ; based on genetic algorithm optimization characteristics , combined with improved genetic algorithm BP optimize parameters of the neural network , and then use its bearing failure diagnosis, analysis of the effect of this method diagnosis . The system is based on software as the core of the virtual instrument development , so that the system has scalability , flexible definition , high performance and low maintenance cost advantages . System software development process platform using LabWindows / CVI. Experiment confirmed EMU bogie bearing fault diagnosis system can accurately predict its failure to provide reasonable maintenance recommendations .

Related Dissertations
More Dissertations