Data processing: measuring – calibrating – or testing – Measurement system – Statistical measurement
Reexamination Certificate
2002-04-12
2004-06-08
Hoff, Marc S. (Department: 2857)
Data processing: measuring, calibrating, or testing
Measurement system
Statistical measurement
C702S183000
Reexamination Certificate
active
06748341
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
A method and device for machinery diagnostics and prognostics.
2. Prior Art
The machine-health-monitoring problem is widespread, very costly to the nation, and starting to be recognized. In 1996, the National Science Foundation (NSF) convened a workshop on the subject, inviting 37 experts from academia, government, and industry to participate. With respect to the current state of the art, it was suggested by at least some of the attendees that the current design of manufacturing and monitoring processes does not necessarily exploit the state-of-the-art in digital signal processing. In fact, it can be argued that little machine health monitoring research has actually been done within the signal processing community that specifically addresses the needs and nature of these types of problems.
Participants in the aforesaid NSF meeting also reported that for many critical applications, machinery condition assessment has the potential to save billions of dollars while dramatically increasing safety and reliability. Examples include power generation turbines and critical equipment in nuclear reactors or on large oilrigs, where unscheduled failure can result in lost revenue approaching a million dollars per day. Failure during operations of aircraft machines or power train components in helicopters often results in loss of life along with the equipment. For example, it was reported that between 1985 and 1992, the U.S. Navy lost 67 airframes and 84 lives due to material-related failure in helicopters. Failure during operations of critical machinery on a Navy capital ship during wartime could endanger the security of the nation. Methods for machinery condition assessment which provide warning in time to cease operations or schedule maintenance offer immense value in such applications, and methods and devices for monitoring the condition of such machines is already routinely used or eagerly sought.
The maintenance of turbine machines alone costs the Navy hundreds of millions of dollars per year. The cost and security risk of unscheduled failure is high, so prophylactic maintenance is routinely practiced. At great expense, in both dollars and downtime, critical components are routinely replaced long before their mean time to failure in order to reduce the risk of failure during operations. Unfortunately, it is suspected that the majority of failures are due to problems introduced by faulty maintenance; that is, the routine maintenance itself is the dominant cause of failure. For both cost reduction and for improved reliability, the Navy would prefer to adopt an “if it ain't broke, don't fix it” philosophy. But this can only be done without endangering Navy operations with reliable machinery condition assessment and early detection of precursors to equipment failure. Much the same conditions and logic prevail in the Air Force, as well as among other organizations operating high-value equipment.
Current technologies for machine health monitoring are usually based on the concept of individual variables exceeding or approaching upper and lower limits. Existing machine health monitoring systems typically monitor machine variables and notify an operator or maintenance personnel when a particular variable exceeds or is approaching a limit. The method and device of the present invention also employs this limit method, but only as a minor ancillary function. The principal and unique function of the device and method of the present invention is not just to monitor individual machine variables (e.g., pressure, temperature) relationship to fixed limits, but instead to monitor all of their relationships to each other and to recognize developing abnormal relationships far in advance of any individual variable approaching a fixed limit.
SUMMARY OF THE INVENTION
It is a first object of the invention to provide a device and method for obtaining an accurate assessment of the normality of measurable machine variables and thus predict future machine operability.
It is a further object of the invention to provide a method, which may be used in a computer environment, to diagnose a problem affecting machine operability.
It is another object of the invention to provide a computer readable medium bearing instructions that cause a computer to provide an accurate assessment of machine variable relationships and thus to predict future machine operability.
It is yet another object of the invention to provide means for accomplishing the above objectives in “real-time,” so that the predictions can be made while the machine is operating based on the measured values of the current operating variables being produced by the machine.
The above objectives are met by automatically developing a set of accurate multivariate transfer function models (TFMs) of machine variables and their relationships over time that are deemed most important by the user, machine expert, or machine operator. A TFM is simply a statistical model or function that converts “inputs” to “outputs”, as illustrated in FIG.
1
. Comparison of TFM predictions of “normal” machine variable values and relationships to actual observed machine variable and relationships is then accomplished by abnormality detection algorithms in accordance with the present invention. The combination of TFMs and abnormality detection algorithms are so powerful, the device and method of the present invention detects small abnormalities (on the order of 0.01% of normal operating values) with high probability while keeping false alarm rates low. The benefits are both reduced wasted maintenance resources due to false alarms and time-based actions and reduced catastrophic failures due to missed subtle indicators.
The present invention discloses, in one aspect, a computer readable medium whose contents include a set of instructions that cause a computer system to perform an accurate machine health prediction upon receipt of raw system variables data from the machine, the computer readable medium comprising: (a) a data filtering component comprising a filtering algorithm operable for eliminating “spikes” in the raw system variables data while leaving legitimate “ups and downs” in the data intact, thereby improving the predictive performance of transfer function models (TFMs); (b) a TFM estimator component comprising a multivariate non-linear transfer function estimation algorithm that uses historical or real-time calibration data comprised of values of machine operating variables from one or more normally operating machines to define “normal” and to construct statistical multivariate non-linear TFMs of “normal” machine variables; (c) a TFM modelbase comprising (i) a plurality of TFMs; and (ii) the operational condition variable values employed to develop the TFMs; (d) a Predictor component that finds the best TFMs in the TFM modelbase for the monitored machine variables and the current set of real-time machine operational conditions and uses those to predict one-step-ahead values for the monitored machine variables; (e) a Comparator component that compares what an operating machine is producing in the way of monitored variable values to what the appropriate TFMs predict for the same variables, producing the algebraic differences, called residuals; and (f) a prognosticator component that receives the residuals for each monitored machine variable from the Comparator component and conducts statistical tests on the residuals to categorize each machine variable as normal or abnormal. The Prognosticator uses the results of the tests on each machine variable to calculate an overall probability of machine abnormality (PMA), the value of which is predictive of machine failure.
In another aspect, the present invention provides a method for predicting machine failure comprising the steps of: (a) presenting a machine having an operability that is dependent on the values of a set of machine variables; (b) presenting a computer means comprising the computer readable medium described above; (c) measuring the set of machine variables, the
Hoff Marc S.
Miller Craig Steven
Petit Michael G.
LandOfFree
Method and device for machinery diagnostics and prognostics does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and device for machinery diagnostics and prognostics, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and device for machinery diagnostics and prognostics will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3354680