Valve actuator fault analysis system

• US 8,838,413 B2
• Filed: 05/12/2011
• Issued: 09/16/2014
• Est. Priority Date: 05/12/2011
• Status: Active Grant

• Current Assignee: Saudi Arabian Oil Company (Government of Saudi Arabia)
• Original Assignee: Saudi Arabian Oil Company (Government of Saudi Arabia)
• Inventors: Genta, Pablo Daniel
• Primary Examiner(s): Desta, Elias
• Application Number:
  US13/068,557
• Publication Number:
  US 20120290261A1
• Time in Patent Office: 1,223 Days
• Field of Search: 702/33, 702/179, 702/183, 701/101, 701/29.2
• US Class Current: 702/179
• CPC Class Codes: F16K 37/0083 by measuring valve parameters
• CPC Class Codes: G05B 23/00 Testing or monitoring of control systems or parts thereof monitoring of programme-control systems G05B19/048, G05B19/406
• CPC Class Codes: G05B 23/0245 based on a qualitative model, e.g. rule based; if-then decisions
• CPC Class Codes: G05B 23/0267 Fault communication, e.g. human machine interface [HMI]

Abstract

A valve actuator fault analysis system for detecting, analyzing and reporting valve actuator performance characterized as deviations, malfunctions and failures. The fault analysis system includes an expert system that monitors the valve actuator and related components, such as the power supply and signal line, for determining causal factors of degraded and/or defective performance; generation of fault-analysis reports in a readable text format file; and methods for characterizing deviations, malfunctions and failures, for generating and concatenating event messages and for detecting and characterizing events that take place in valve actuators, peripheral systems and devices and displaying the results on a display device. The system can operate either in conjunction with, or be embedded in the hardware of a pneumatic valve positioner, an electrical valve actuator and/or similar devices, and can receive messages from external sources, process them, and produce fault-analysis reports in text format for display to the user.

Claims

1. A method for detecting and characterizing an occurrence of at least one event during operation of a valve actuator as one of a deviation, malfunction or failure, the method comprising the steps of:

detecting a series of signals received from a plurality of sensors, where each signal represents a condition of a different operational variable associated with the actuator;

determining from the signals one or more of the operational variables that caused the occurrence of the at least one event;

determining at least one relationship between the one or more variables and the event, wherein the at least one relationship includes a time-related succession of intermediate states of operation associated with the valve actuator that are interconnected by transition paths, which are quantified by cause-and-effect relationships and temporal and likelihood factors;

characterizing the event as being one of a normal mode of operation, a deviation, malfunction or failure; and

communicating the at least one relationship through an output device.

2. The method of claim 1, wherein the detecting step comprises the steps of:

converting output signals from the plurality of sensors to a voltage signal;

generating sequences of different signal patterns by means of a programmable signal generator;

normalizing the voltage signal against the generated signal patterns;

comparing the normalized voltage signal during a predetermined time period in order to determine the crossing amplitude level of the signals;

recording each detected crossing level value together with a time stamp in order to produce a time series; and

determining a signature and statistics of the time series.

3. The method of claim 2, wherein the step of generating sequences of different signal patterns by means of a programmable signal generator comprises:

repeating specific operations of the actuator under predetermined fixed conditions;

storing historical data for a plurality of variables and relational paths therebetween; and

learning the states of operation of the valve actuator from the historical data.

4. The method of claim 1, wherein the step of determining from the signals one or more of the operational variables that caused the occurrence of the at least one event comprises the steps of:

comparing the signature and statistics against predetermined patterns and values to identify a present operating zone and the operational variables indicative of states of operation; and

determining the origin, dynamic properties, severity and range of the at least one event for each normalized signal.

5. The method of claim 1, wherein the step of characterizing the event as being one of a normal mode of operation, a deviation, malfunction or failure comprises the step of classifying each of the events and assigning at least one of a plurality of predetermined descriptive messages to each event.

6. The method of claim 1, wherein the step of determining the at least one relationship between the one or more variables and the event comprises the steps of:

identifying recurring events and identifying the trajectory of the events; and

determining a state of operation of the actuator.

7. The method of claim 6, wherein the step of determining the state of operation comprises the steps of:

classifying an event to be a failure when the actuator cannot function;

classifying an event to be a malfunction when the an actuator parameter cannot meet a predetermined parameter range, and

classifying an event to be a deviation when the actuator parameter is outside the predetermined range.

8. The method of claim 6, wherein the step of determining the state of operation comprises concatenating event messages with messages describing possible causal factors and with messages received from external sources and producing textual phrases that follow predetermined grammatical rules.

16. The method of claim 1, wherein the step of communicating the relationship through an output device comprises the step of communicating the state of operation of the actuator on an output device.

17. The method of claim 16, wherein the step of communicating the state of operation comprises the step of displaying the state of operation of the actuator as text messages using a series of grammatical rules on a display device.

18. The method of claim 2, wherein the step of generating sequences of different signal patterns by means of a programmable signal generator comprises:

repeating specific operations of the actuator under specific fixed conditions;

storing historical data for a plurality of variables and relational paths therebetween; and

learning the states of operation of the valve actuator from the historical data.

19. The method of claim 1, wherein the step of determining at least one relationship between the one or more variables and the event comprises the steps of:

providing physical deterministic models, statistical models and event frequency models associated with variables and relational paths therebetween for the actuator and surrounding environment;

comparing metrics associated with the variables of the actuator and surrounding environment as between the deterministic models and statistical models;

selecting variable metrics as between the deterministic and statistical models based on event frequency; and

determining cause and effect relationships between the states of operations based on the comparisons between physical deterministic models, statistical models and event frequency models.

9. The method of step 7, wherein the step of determining the state of operation comprises:

identifying nodes of relational paths associated with the state of operation of the actuator;

generating a table that assigns a unique representative text to each transition path formed between the nodes along the relational paths; and

concatenating the event messages with the representative text according to the predetermined grammatical rules.

10. The method of claim 9, wherein the step of identifying nodes comprises the steps of providing a state transition matrix having a plurality of variables associated with the actuator, each variable being characterized with metrics defining a normal state of operation, a deviation, a malfunction and a failure associated with the operation of the actuator, each variable representing a node;

and identifying transition paths between the nodes that define the state of operation of the actuator.

11. The method of claim 10, wherein the step of providing the state transition matrix includes defining the deterministic metrics associated with each variable based on physical sub-models that contemplate physical characteristics experienced by the actuator.

12. The method of claim 11, wherein the physical characteristics include heat-transfer, electrical variables and mechanical variables associated with the actuator.

13. The method of claim 11, wherein the step of providing the state transition matrix includes defining the statistical metrics associated with each variable based on statistical information including component failure rates.

14. The method of claim 13, further comprising:

comparing, for each variable, the deterministic and the statistical metrics; and

selecting one of the deterministic or statistical metric for each node based on event frequency associated with the actuator.

15. The method of claim 9, wherein the predetermined grammatical rule includes at least one of a subject element, a verb element, and a predicate element.

20. A method of analyzing, detecting and recording an occurrence of at least one event during operation of a valve actuator, the method comprising the steps of:

receiving a plurality of input signals from a plurality of sensors that monitor operation of the actuator and/or surrounding environment;

identifying attributes of the plurality of input signals and classifying the attributes as a potential occurrence of the at least one event;

determining at least one relationship between the identified attributes and the at least one event, wherein the at least one relationship includes a time-related succession of intermediate states of operation associated with the valve actuator that are interconnected by transition paths, which are quantified by cause-and-effect relationships and temporal and likelihood factors; and

verifying the at least one event is characterized as one of a deviation, malfunction or failure associated with the actuator and/or surrounding environment.

21. The method of claim 20, wherein said receiving step comprises receiving input signals from each of the plurality of sensors over at least one of an analog input channel, a discrete input channel, and a digital data channel.

22. The method of claim 20, wherein said identifying step comprises determining if a state of operation associated with an input signal exceeds a predetermined alarm value over a predetermined time.

23. The method of claim 20, wherein said determining step comprises:

characterizing a state of operation associated with an input signal as being degraded;

searching a current state of operation database for all other states of operation related to the state of operation associated with the input signal;

retrieving past events associated to the related states occurring within a predetermined time to the present time;

comparing the state of operation associated with the input signal with operational values associated with the retrieved past events; and

determining from historical data of the retrieve past events whether the state of operation associated with the input signal is a deviation, malfunction or failure associated with the actuator.

24. The method of claim 23, wherein the step of searching a current state of operation database comprises storing a new state of operation if a previous state of operation is not identified.

25. The method of claim 20, further comprising the step of providing a textual report describing current and probable causes of the deviations, malfunctions and failures of the actuator.

26. A fault analyzing system for characterizing deviations in a specified valve actuator power supply, command signals, ambient conditions, usage conditions and actuator performance during operation of a valve actuator, the system comprising:

a memory for storing sensor signals associated with each of the valve actuator power supply, command signals, ambient conditions, usage conditions and actuator performance;

a processor coupled to the memory; and

an expert system program stored in the memory and executable by the processor, the expert system program operable to;

a. calculate a first-order discrete derivative for a measured variable Sn associated with a state of operation of the actuator;

b. calculate a discrete derivative for the measured variable Sn by utilizing a predetermined variable scan time;

c. detect a deviation in measured value of the variable Sn as compared to a specified value of the variable Sn;

d. determine an actuator operating signature according to predetermined criteria;

e. calculate a trajectory signature of the deviation as a function of time;

f. associate messages with the trajectory signature of the deviation; and

g. communicate the associated messages to an output device.

27. The system of claim 26, wherein the specified value of the variable Sn includes a range of values associated with the variable Sn as defined statistically and/or deterministically during operation of the actuator, said range of values associated with the variable being stored in a data structure in the memory.

28. The system of claim 27, wherein the actuator operating signature is defined by a plurality of variables Sn each having a range of predetermined values.

29. The system of claim 28, wherein the plurality of variables Sn are associated with at least one sub-model defined for the actuator, each sub-model having linked groups of variables defining different states of operation of the actuator.

30. The system of claim 29, wherein the messages associated with the trajectory signature of the deviation comprises concatenated messages describing valve actuator faults with messages describing possible causal factors and with messages received from external sources, thereby producing cause-effect descriptive phrases that follow predetermined grammatical rules.