Research On Fault Diagnosis Of Rolling Bearing Based On Dual-tree Complex Wavelet Packet And EMD

Rolling bearing is the common basic element of rotating machinery and equipment.Its working condition is related to the reliability of the whole operation.It is a research hotspot in the field of equipment fault diagnosis of how to identify the problem accurately and develop the corresponding maintenance strategy when the rolling bearing fails,The vibration signal of the rolling bearing has nonlinear and non-stationary characteristics.The frequency and energy characters of the vibration signal will change when the different parts of the bearing fails.The fault type of the bearing can be identified by analyzing the change of the energy characteristic of the vibration signal.A fault diagnosis model of rolling bearing based on dual-tree complex wavelet packet,EMD and RBF neural network is established in this paper.The vibration signal is denoised by the singular value difference spectrum method to reduce the influence of noise,then,two algorithms on dual-tree complex wavelet packet and empirical mode decomposition are used to decompose the outer ring,the rolling element and the inner ring fault signal of the rolling bearing and extract the energy characteristic samples,and provide the data support for the next fault type identification.RBF neural network has a strong adaptive learning,nonlinear approximation ability,can achieve the classification of fault types.In this paper,the fault recognition model of RBF neural network is realized by MATLAB programming.The energy feature samples based on dual-tree complex wavelet packet and EMD algorithm are taken as input samples of network model respectively,and the fault type of rolling bearing is identified by simulation output.The whole algorithm is implemented on the dataset provided by the electrical engineering laboratory of Keith West Reservoir University,The simulation results show that the diagnostic model based on the dual-tree complex wavelet packet and the empirical mode decomposition and RBF neural network can effectively identify the different faults of the rolling bearing.