Method and apparatus for knowledge-based signal monitoring and analysis

• US 5,274,572 A
• Filed: 03/06/1990
• Issued: 12/28/1993
• Est. Priority Date: 12/02/1987
• Status: Expired due to Term

• Current Assignee: Schlumberger Technology Corporation (Schlumberger NV)
• Original Assignee: Schlumberger Technology Corporation (Schlumberger NV)
• Inventors: O'Neill, Dennis M., Moinard, Laurent L., Gingrich, Paul C., Mullarkey, Peter W.
• Primary Examiner(s): Teska, Kevin J.
• Application Number:
  US07/489,053
• Time in Patent Office: 1,393 Days
• Field of Search: 364/550, 364/551.01, 364/422, 364/554, 395/11
• US Class Current: 702/57
• CPC Class Codes: G01V 11/002 Details, e.g. power supply systems for logging instruments, transmitting or recording data, specially adapted for well logging, also if the prospecting method is irrelevant means for transmitting well survey signals E21B47/12; signal transmission systems in general G08C; transmission in general H04B

Abstract

A method and apparatus for monitoring and analysing signal data which uses a network model describing the system under investigation and a runtime agent for acquiring the signal data and accessing the model if an anomaly in the signal data is indicated. The network model describes events of interests and how the events relate to phenomena in the system. The network model is constructed using an object-oriented approach with: observations of the events of interests in the system; situations which describe possible underlying causes of the observations; and relations which specify the logical relationship between the observations and situations. The runtime agent is constructed with an object-oriented approach using observers which monitor the signal data and compute whether an anomaly in the incoming signal data exists. If an anomaly is identified, an "observation" is generated and the network model entered to analyse the observation and estimate a cause of the observation. The method and apparatus are applicable for interpreting phenomena in a wide variety of physical systems and have been exemplarily applied to monitoring the quality of oil well logging and laboratory material test sensor configurations.

Claims

1. A method for monitoring and analyzing a plurality of signals from sensors in a system comprising the steps of:

providing an analysis network defining behavior and problem modes of the system;

establishing one or more observers for each signal and defining criteria for each observer;

monitoring a signal value with a corresponding observer and detecting when the signal value does not meet its defined criteria;

reporting an observation from the detecting observer to the analysis network if said signal value does not meet its defined criteria;

analyzing the observation with the network, including the substeps of invoking one or more additional observers, reporting signal values from each respective observer to the network, and determining whether said one or more additional observers meet respective defined criteria; and

generating an explanation of the observation from the network.

2. The method according to claim 1, wherein the steps are performed to generate said explanation during monitoring of additional signals from said sensors.

3. The method according to claim 1, wherein the measurement signals are derived from a plurality of sensors.

4. The method according to claim 1, including the step of monitoring a number of signal values with respective observers in near real time.

5. The method according to claim 1, the analyzing step including observing additional signal values occurring prior to the observation.

6. The method according to claim 1, the establishing step including establishing at least some of the signals with multiple observers.

7. The method according to claim 1, the establishing step including having a least one observer which monitors multiple signals.

8. The method according to claim 1, prior to said monitoring step including the step of organizing a number of signal values into a data frame and indexing sequential measurement signals into sequential data frames.

9. The method according to claim 8, wherein the identifying index for sequential data frames is time of data measurement.

10. The method according to claim 8, wherein the identifying index for sequential data frames is the physical location of the sensors acquiring the signals.

11. The method according to claim 8, wherein the signal values in each data frame represent separate measurements of the system under investigation.

12. The method according to claim 8, wherein the criteria for reporting an observation includes an algorithm for testing a signal value in a particular data frame and a time interval for reading successive data frames.

13. An apparatus for monitoring and analysing signals from a system comprising:

means for acquiring a plurality of signals as digital values each indicative of phenomena in the system;

observer means coupled to the acquiring means for testing at least one of said signal values for the presence of an anomaly in said at least one signal values and for generating an observation if an anomaly is detected; and

network means containing a number of stored observations and situations linked by specified relationships, activatable if an anomaly is detected by the observer means, for retrieving other signal values and analyzing the observation and other signal values to output a situation as an explanation for the anomaly.

14. The apparatus according to claim 13, the observer means being operable for testing for anomalies in two or more signal values in near real time.

15. The apparatus according to claim 13, the anomaly defined as including algorithm criteria for the signal value and a time interval test for the presence of signal values not meeting the criteria.

16. The apparatus according to claim 13, the network means being operable for outputting an explanation as a best guess at any time during analysis.

17. The apparatus according to claim 13, the network means being operable for outputting an explanation after the network means has completed its analysis.

18. The apparatus according to claim 13, wherein the acquiring means includes a plurality of acquisition sensors and means for converting sensor signals to digital values.

19. The apparatus according to claim 18, wherein the acquiring means includes a computer buffer for receiving the sensor signal digital values.

20. The apparatus according to claim 15, wherein the buffer includes a plurality of indexed data frames for accepting groups of successive signals.

21. The apparatus according to claim 20, wherein the data frames are indexed by time of signal acquisition.

22. The apparatus according to claim 20, wherein the data frames are indexed by location of the sensor acquiring the signal.