Data processing: vehicles – navigation – and relative location – Vehicle control – guidance – operation – or indication – With indicator or control of power plant
Reexamination Certificate
2000-12-29
2002-09-24
Wolfe, Willis R. (Department: 3747)
Data processing: vehicles, navigation, and relative location
Vehicle control, guidance, operation, or indication
With indicator or control of power plant
C701S111000, C702S082000, C714S048000
Reexamination Certificate
active
06456928
ABSTRACT:
FIELD OF THE INVENTION
The present invention is related generally to monitoring systems, and more specifically, to devices and methods for monitoring system performance or parameters to detect and isolate present and future predicted failures.
BACKGROUND OF THE INVENTION
Numerous systems or components in use today are subject to physical failure. Many of these systems have a monitoring mechanism for monitor the operation of the system. The monitoring mechanisms often have one or more monitoring points. One illustrative system is a combustion engine. Inevitably, combustion engines fail. Some engine failures may be prevented through replacement of parts, standard service procedures, and/or major overhauls. Engine failures or component failures may cause more harm than necessary due to a lack of real time information about the nature of the component failure and the lack of a real time suggestion as to a course of action which could ameliorate the component failure. The prevention of engine failure may not be carried out due to the lack of knowledge of current engine parameters, lack of knowledge of trends in engine parameters, and lack of knowledge of current component failures.
In combustion engines used in automobiles, unexpected or puzzling engine failures are at best inconvenient, at worst presenting safety issues. Failure in combustion engines used in industrial applications may cause unplanned down time and loss of production. Combustion engines used in aircraft may cause major loss of life upon unexpected failure.
What would be desirable, therefore, are systems and methods for better detecting and isolating system or component failures. Systems and methods for predicting future system or component failures before they occur would also be advantageous.
SUMMARY OF THE INVENTION
The present invention includes methods and devices for detecting faults or failure modes in systems that are subject to failure. The present invention is described with reference to an engine, but numerous other applications are contemplated. An engine may have numerous monitor points, with each point having an actual parameter value, and the engine having numerous failure modes. The monitor points may be measurable values which can include, for example, continuous values such as engine oil pressure, or binary state values such as ON or OFF, HI ALARM or NOT HI ALARM.
The present invention may use a range of acceptable values for each monitor point parameter. Values outside of this range may generate a deviation result or signal. Some ranges are binary values, such as OFF, which have no acceptable values other than the desired binary value. Other ranges are continuous values, such as two endpoints for a range, or a midpoint together with a margin about the midpoint. Another range is a high or low limit for a value, with any value exceeding the limit being cause for a deviation condition. In some embodiments, exceeding different limits for the same measured quantity can result in different deviation outputs.
Numerous ambiguity groups may be created by a user to work with the present invention, or may be included with a product according to the present invention. The ambiguity groups can include a number of failure modes, or faults that can be a cause, or the cause, of out of range values or deviations emanating from the monitor point parameter values. One class of failure modes is mechanical failures, for example, a ruptured fuel line. In one embodiment, an ambiguity group includes one or more failure modes, the monitor point parameter deviation that would result from the failure mode or combination of failure modes, and the probability of the combination of failure modes occurring together to produce the deviation signal. The ambiguity groups can be clustered together about common deviation signals than can be caused by the failure modes. In one example, all ambiguity groups that could result in lower than expected RPM could be grouped together in vectors or collections of ambiguity groups. In one embodiment, the ambiguity groups are collected together in a table.
In operation, the engine failure detector can collect data from sensor values, and operate upon the sensor values to form parameter values for monitor points. Some monitor points are directly coupled to sensors while other monitor points are synthesized from multiple sensor values to form virtual or inferred monitor points.
The parameter value may be obtained, and compared with the desired parameter value range by a monitor point comparator function. If the comparator function decides that the actual parameter value exceeds the acceptable range, a deviation signal may be generated. In some embodiments, a presumptive deviation signal is filtered prior to outputting the deviation signal. The filtering is often used to reduce the number of false alarms.
Once a deviation signal is generated, a fault isolator functionality can operate on the provided deviation signal or signals. In one embodiment, all the ambiguity groups having a particular deviation signal are collected together. Given the collected ambiguity groups, one or more of the ambiguity groups can be selected as the most likely ambiguity groups to have generated the deviation signal. In one embodiment, the ambiguity group having the highest probability is selected. In another embodiment, the possible ambiguity groups are analyzed to determine intersecting failure modes, if any. It is contemplated that the ambiguity groups may be Boolean ANDed together to determine common failure modes.
The failure mode or modes selected as most likely to have caused the deviation signal may be output to a human operator, and may be fed to an automatic control function which may determine a course of action to alert or ameliorate the condition caused by the failure mode.
Some embodiments of the invention have failure predictors as well. A history of values for monitor point parameters may be collected by a data historian functionality. Some embodiments record the data in a sliding window. The historical data can be extrapolated or projected out into the future, and a prediction made as to whether the projected monitor point parameter value trends will exceed the range of acceptable values for that monitor point parameter. If a deviation condition is predicted, a projected deviation signal and the projected time for that deviation may be output for use by a human operator and/or fed back to an engine control system to ameliorate or even postpone the predicted deviation signal condition. In many embodiments, the predicted deviation signal is analyzed to isolate a fault in a manner similar to the fault isolation performed for a present deviation signal. In some embodiments, the early deviation signal predicted is used to suggest preventative maintenance earlier than otherwise planned.
REFERENCES:
patent: 5099436 (1992-03-01), McCown et al.
patent: 5586252 (1996-12-01), Barnard et al.
patent: 6223143 (2001-04-01), Weinstock et al.
“AGATE Propulsion Monitoring and Diagnostic System”; Honeywell Technology Center; presented in Jul. 1999.
Honeywell International , Inc.
Wolfe Willis R.
LandOfFree
Prognostics monitor for systems that are subject to failure does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Prognostics monitor for systems that are subject to failure, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Prognostics monitor for systems that are subject to failure will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2830540