Research On Vibration Signal Analysis And Diagnosis Technology Of Power Machinery Transmission Device

With the full implementation of "Made in China 2025",machinery and equipment are becoming more and more popular in daily life and playing an increasingly important role.It is well known that dynamic mechanical transmission as the "heart" of mechanical equipment is directly related to the normal operation of mechanical equipment.Therefore,the analysis and diagnosis of vibration signals of mechanical transmission have great research significance on industrial production efficiency and maintenance cost.This paper mainly starts with the collection of vibration signals of power mechanical transmission and optimizes the design of vibration signal acquisition card.The collected vibration signals are analyzed by wavelet and eigenvectors are extracted.According to the characteristics of eigenvalue,SVM is used to carry out the fault diagnosis identification,and the D_S evidence theory is applied to multi-feature fusion decision to improve the fault diagnosis recognition rate.At last,the correctness and feasibility of the vibration signal analysis and diagnosis technology of power transmission device are verified by the vibration signal data of ball bearing of ZQ-200 reducer.The specific research contents are as follows:First of all,the needs and overall program of the vibration signal data acquisition are analyzed and designed.Vibration sensor selection,sensor measurement point arrangement and vibration signal acquisition card design based on TMS320C6748 are based on requirements and overall solution.Finally,powertrain vibration signal data is collected.Then,the collected vibration signal is pre-processed to eliminate the noise interference such as high frequency and trend term mixed in the acquisition.The appropriate wavelet transform and wavelet basis function are selected to analyze the vibration signals and extract the representative features in time domain,frequency domain and time-frequency domain.Types and causes of common gear and bearing faults are briefly analyzed.Appropriate kernel function and penalty parameters of support vector machine are selected for fault type identification by using the extracted eigenvalues.The D_S evidence theory was used to fuse the single-feature fault recognition results of SVM,and the fault diagnosis accuracy was improved by fusion decision.The technical method proposed in this paper is validated by the vibration signal data of ball bearing of ZQ--200 reducer.The results show that the proposed vibration signal analysis and diagnosis technology has high diagnostic recognition rate for power mechanical transmission failure.Finally,the vibration signal analysis and fault diagnosis system is designed using the powerful Lab VIEW software platform and the vibration signal analysis and diagnosis technology proposed in this paper.The practical data are used to test and verify the system.