| In recent years, with the constant expansion of the application field of the scroll compressor, the different working conditions, the factors that cause the fault of the scroll compressor become more complicated, the fa ult diagnosis of the scroll compressor becomes more necessary. The vibration test experiment platform of scroll compressor is set up and the analysis of its vibration noise is particularly important. However, due to the complexity of the working environmen t of the scroll compressor, the test system is difficult to build and other factors, it is difficult to meet the requirements of real-time monitoring, realtime diagnosis. In addition of scroll compressor in domestic application time is not long, for its us e in the process of fault database is not perfect and scroll compressor vibration source is more, shell surface signal often non-stationary and nonlinear, so fault feature is difficult to accurately area and fault diagnosis process is more complex. In this paper, based on the experimental platform of vibration test of scroll compressor, the information entropy and entropy distance are combined to realize the fault diagnosis of nonstationary signals. The specific implementation process is as follows:(1) A fault diagnosis method based on information entropy and entropy distance is proposed, which is based on the vibration signal analysis and connect the entropy and Euclidean distance in information theory.(2) To set up the experimental platform, the installa tion of the sensor, the replacement of hardware and the debugging software testing have been finished. Acquisition signal,the normal operation of the characteristics of the standard and the characteristics of the four typical failure criteria is established.(3) Using the experimental platform to simulate four kinds of fault and using MATLAB signal processing toolbox of the powerful and the collected fault signals of threshold based on the wavelet packet denoising, de compose and reconstructto extract the information entropy as the fault feature.(4) The extraction of the fault features, compared with the typical fault, using the entropy distance calculation method to get t he entropy distance, consider the information entropy and entropy distance on the test signal for fault diagnosis.The experimental results show that this method has high accuracy and can distinguish the fault diagnosis of rotor unbalance and bearing fault. Information entropy can reflect the fault types and fault severity, and the entropy d istance curve can further improve the accuracy of diagnosis. At the same time, it can also represent the characteristics of the complex fault, thus providing a new way of thinking for the complex fault diagnosis. It also provides some help for the structur e design, manufacture and installation of the scroll compressor. |
