Acquisition And Analysis Of Vibration Signal Of Rolling Bearing

As a major part of mechanical equipments, it goes without saying that rolling bearing plays an important role in mechanical equipments. According to statistics, there are30%equipment troubles caused by rolling bearings, so taking real-time detection to rolling bearings’working status is very necessary to avoid rolling bearings’troubles and damages to mechanical equipment. There is great significance which can make sure the whole equipment system works in a good condition. In this paper,vibration signals collection system is designed, it collects the vibration signals of rolling bearings and analyzes the collected data,then it extracts the signals’ feature and deals with the signals feature by PLS method and other methods.Finally, the fault recognition model is built to proceed the real time detection to the working condition of rolling bearings.The core job of the paper is that design vibration signals collection system for rolling bearings, extract the signals’feature and build fault detection model.In this paper, a set of vibration signals collection system whose core is STM32F103RB is designed to collect vibration signals of rolling bearings, then analyze the collected data to extract the signals feature and build fault detection model by wavelet packet energy spectrum--PLS method and other methods.The detailed jobs are as followings:(1)The vibration signals collection system can transmit the data in a high-speed and parallel way even in multi-channel. The core control module is the key part,its quality will relate with whether it can collect the signal in a correct and high-speed way. For collected vibration signals,frequency domain analysis methods are adopted to extract signals’feature and build fault detection model.(2) Propose Wavelet packet energy spectrum-PLS fault detection method and build fault detection model.Firstly, make wavelet packet decomposition to vibration signal and extract wavelet packet energy spectrum feature vectors as the input data of PLS in order that build fault recognition model to verify the validity of this method.(3) Propose wavelet packet energy spectrum-KPLS fault detection method and build fault detection model. The KPLS fault detection which base on wavelet packet energy spectrum is a method for dealing with linear data. But during the process of rolling bearings working, there are a lot of nonlinear data, so wavelet packet energy spectrum-KPLS method is proposed.The most efficient way to monitor the working status of rolling bearings is that collect the vibration signals of rolling bearings and analyze them. Thus, the paper proposes the fault recognition method upon wavelet packet energy spectrum which adopts wavelet packet analysis method to extract the energy futures of vibration signals and build PLS&KPLS fault recognition models, detect and recognise the malfunction information of rolling bearings accordingly.Making the malfunction detect models apply to the vibration signals data of rolling bearings supplied by Case Western Reserve University, the experiment verify the validity of the methods.In the end, the article summarizes the whole performance of the paper and expectation to the malfunction detection of rolling bearings.