Data processing: measuring – calibrating – or testing – Measurement system – Measured signal processing
Reexamination Certificate
2000-02-03
2003-03-25
Assouad, Patrick (Department: 2857)
Data processing: measuring, calibrating, or testing
Measurement system
Measured signal processing
C708S400000
Reexamination Certificate
active
06539343
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates generally to the field of condition monitoring and diagnosis, and more particularly to methods for system-level condition monitoring and diagnosis using time-frequency analysis mechanisms.
BACKGROUND OF THE INVENTION
Future electro-mechanical machines and structures will increasingly participate in their own service and maintenance using embedded distributed self-diagnostics that are remotely accessible to monitor machine health, detect and isolate subtle performance degradation, and in some cases even reconfigure some machines to adapt to changing operating environments. Traditionally, corrective maintenance and preventative maintenance have been the only two service paradigms. More recently, predictive or condition-based maintenance, enabled by Micro-Electro-Mechanical Systems (MEMS) is emerging as an alternative. Batch-fabricated MEMS sensors are far less expensive than conventional sensors and, thus, encourage larger-scale deployment throughout factories and in vehicles and devices. Condition-based maintenance is just-in-time maintenance based on the actual health of the machine and its components. Since it avoids the cumulative cost of unnecessary service calls associated with preventative maintenance and the occurrence of machine failure and degradation associated with corrective maintenance, condition-based maintenance provides substantial cost savings.
Real-time signal analysis is critical for condition-based monitoring of structures and electro-mechanical systems. An electro-mechanical system or a complex structure comprising multiple moving elements in complex operating regimes can exhibit extremely complex system level responses due to the interaction between the actuating elements and supporting mechanical structures. Fault manifestation in these systems is typically non-stationary in that there is no persistent means or variance over time. The actuating elements such as motors and solenoids produce rich mechanical excitation signals at multiple time and frequency scales. Traditional Fourier spectral analysis, while useful for establishing the signal bandwidth, is unsuitable for analyzing the time-varying properties of the signal that are important for the purpose of fault diagnosis. One problem is that failure modes of system components are difficult to identify and characterize using time-based or frequency-based analysis alone. Another problem is that signals can vary in quality due to external and internal noise as well as multiple signal interference. Furthermore, subtle changes in signals indicative of machine conditions such as motor bearing wear are often buried in larger responses due to, for example, structural resonances. Traditionally, signal demodulation to get rid of irrelevant components, such as performed in the telecommunications industry, is carried out in frequency domains. Because the signals from machine health monitoring applications are typically transitory and time varying, many real-time monitoring applications require more efficient demodulation methods to improve the signal to noise ratio in fault detection.
In light of the foregoing, there is a need for a method for condition-based monitoring of a system and its components using joint time-frequency analysis.
SUMMARY OF THE INVENTION
Accordingly, the present invention is directed to a method for condition-based monitoring of a system and its components using joint time-frequency analysis that substantially obviates one or more of the problems due to limitations and disadvantages of the related art.
In accordance with the purposes of the present invention, as embodied and broadly described, the invention provides a method for constructing time-frequency models of non-stationary signals. The first step is to collect a plurality of training data from at least one sensor in a system of interest, wherein the training data includes an operation fault of the system. The next step is to extract wavelet coefficient features from the training data by performing wavelet analysis. Signal templates of the operating fault are then constructed using the wavelet coefficients.
In another embodiment, the invention provides a method for condition-based monitoring of a system of interest. The first step is to provide signal templates representing a normal operating condition and a faulty operating condition of a system of interest, wherein the signal templates comprise wavelet coefficients. Next, sensor signal segments are constructed by extracting wavelet coefficient features using wavelet analysis from a signal from a sensor monitoring the system of interest. The signal template are then convolved with the segments of the signal from the sensor monitoring the system of interest. Finally, the segments of the sensor signal are classified into the normal and faulty operating conditions of the system based on convolution amplitude.
In still another embodiment, the present invention provides a method for condition-based monitoring of a component of a system using a residual signal. In the first step, a signal model of the nominal response of the component is constructed. Next, a signal from a sensor monitoring the component is detected. And, the signal model is subtracted from the signal from the sensor monitoring the system to create a residual signal.
In another embodiment, the present invention provides a method for condition-based monitoring of a system having a plurality of components. A signal model of the nominal response for each of plurality of components in the system is constructed. Then, signals from a plurality of sensors monitoring the plurality of components are detected. The signal models are subtracted from each signal from the plurality of sensors monitoring the components to create a residual signal for the sensor.
In another embodiment, the present invention provides another method for condition-based monitoring of a system having a plurality of components. In this embodiment, a signal model of the nominal response at each of a plurality of sensors monitoring a plurality of components is constructed. Next, signals from a plurality of sensors monitoring the plurality of components are detected. Then, the signal models are subtracted from the signal from the corresponding sensor monitoring the components.
The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate several embodiments of the invention and together with the description serve to explain the principles of the invention.
REFERENCES:
patent: 5646600 (1997-07-01), Abdel-Malek et al.
patent: 5995910 (1999-11-01), Discenzo
patent: 6014598 (2000-01-01), Duyar et al.
patent: 6105015 (2000-08-01), Nguyen et al.
patent: 6105149 (2000-08-01), Bonissone et al.
patent: 6178386 (2001-01-01), Shertukde et al.
patent: 6181975 (2001-01-01), Gross et al.
patent: 6208951 (2001-03-01), Kumar et al.
patent: 6208953 (2001-03-01), Milek et al.
patent: 6260427 (2001-07-01), Jones et al.
patent: 6272479 (2001-08-01), Farry et al.
patent: 6278962 (2001-08-01), Klimasauskas et al.
patent: 6295510 (2001-09-01), Discenzo
patent: 6356857 (2002-03-01), Qin et al.
patent: 6393373 (2002-05-01), Duyar et al.
Zhang et al., Wavelet Transform Based Sensor Validation, IEEE, 1999.*
Yen et al., “Conditional Health Monitoring Using Vibration Signatures”, IEEE, 1999.*
Gaouda et al., “Wavelet-Based Intelligent System for Monitoring Non-Stationary Disturbances”, IEEE, 2000.*
Kang et al., “Analysis of Vibration Signals for Condition Monitoring of Power Switching Equipment Using Wavelet Transform”, IEEE, 1998.*
Ye et al., “The Abundance of Residuals and Fuzzy Decision-Making for Fault Diagnosis and Sensor Signal Validation”, IEEE, 1996.*
Mufti et al, “Automated Fault Detection and Identification Using a Fuzzy-Wavelet Analysis Technique”, IEEE, 1995.*
Patton et al., “Artificial Intelligence Approaches to Fault Diagnosis”, IEEE, 1999.*
Gertler, “Survey of Model-Based Failure Detection and Isolation In Complex Pl
Manders Eric-J.
Zhao Feng
Assouad Patrick
Finnegan Henderson Farabow Garrett & Dunner L.L.P.
Xerox Corporation
LandOfFree
Methods for condition monitoring and system-level diagnosis... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Methods for condition monitoring and system-level diagnosis..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Methods for condition monitoring and system-level diagnosis... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3029844