Advanced Forecasting and Classification Technique for Condition Monitoring of Rotating Machinery
Prediction and classification of particular faults in rotating machinery, based on a given set of measurements, could significantly reduce the overall costs of maintenance and repair. Usually the vibration signal is sampled with a very high frequency due
- PDF / 341,624 Bytes
- 10 Pages / 430 x 660 pts Page_size
- 44 Downloads / 176 Views
Saint-Petersburg State University, Applied Mathematics and Control Processes [email protected] 2 Saint-Petersburg State University, Applied Computational Physics [email protected]
Abstract. Prediction and classification of particular faults in rotating machinery, based on a given set of measurements, could significantly reduce the overall costs of maintenance and repair. Usually the vibration signal is sampled with a very high frequency due to its nature, thus it is quite difficult to do considerably long forecasting based on the methods, which are suitable for e.g. financial time series (where the sampling frequency is smaller). In this paper new forecasting and classification technique for particular vibration signal characteristics is proposed. Suggested approach allows creating a part of control system responsible for early fault detection, which could be used for preventive maintenance of industrial equipment. Presented approach can be extended to high frequency financial data for the prediction of “faults” on the market. Keywords: fault analysis and prevention, artificial neural networks, artificial intelligence, rotating machinery, ball bearing failures, predictive monitoring.
1 Introduction The faults in particular parts of industrial equipment could cause serious problems such as production losses, expensive repair procedures or even the personnel injures. Therefore the problem of fault analysis and prevention is very important. One of the main reasons of breakdowns of rotating machinery is the bearing failures. Plenty of papers, during the last decades, were devoted to the analysis of such kinds of faults by different methods of vibration analysis [11, 12, 13, 15]. The aim of this paper is to present the combination of forecasting and classification techniques, which could be used for the fault analysis and prevention. In case the sampling rate is measured in kHz (for the dataset used in this paper sampling rate was equal to 40kHz), it is possible to estimate that prediction, achieved by the forecasting based on the pure time signal from the vibration sensors, will be only some microseconds ahead, which seems to be useless in terms of practical applications. Therefore in order to achieve applicable prediction for overcoming the problem described above should be introduced. H. Yin et al. (Eds.): IDEAL 2007, LNCS 4881, pp. 37–46, 2007. © Springer-Verlag Berlin Heidelberg 2007
38
I. Mokhov and A. Minin
Let us consider the measurements obtained from rotating equipment containing ball bearings. Wide range of vibration analysis techniques could be used for analysis of the given measurements. The main idea of the traditional approaches is to analyze the peaks existence for particular frequencies and their multipliers (for more information regarding calculation of these frequencies and analysis methods [4, 5]) This paper will be devoted to the method which could be used for estimation of some particular defect frequency components evolution by using the artificial intelligence methods. The arti
Data Loading...
