Inferential signal generator for instrumented equipment and processes

• US 20030158694A1
• Filed: 03/20/2003
• Published: 08/21/2003
• Est. Priority Date: 11/22/2000
• Status: Active Grant

• Current Assignee: SmartSignal Corporation (Emerson Electric Company)
• Original Assignee: SmartSignal Corporation (Emerson Electric Company)
• Inventors: Wegerich, Stephen W.
• Granted Patent: US 6,876,943 B2
• Time in Patent Office: Days
• US Class Current: 702/127
• CPC Class Codes: G01D 21/00 Measuring or testing not otherwise provided for
• CPC Class Codes: G01D 3/08 with provision for safeguarding the apparatus, e.g. against abnormal operation, against breakdown
• CPC Class Codes: G05B 17/02 electric
• CPC Class Codes: G05B 23/0286 Modifications to the monitored process, e.g. stopping operation or adapting control

Abstract

An improved system and method for producing replacement sensor signals for failed sensors, and inferred sensor signals for non-instrumented physical parameters, in processes and equipment having one or more sensors in place for monitoring physical parameters. A data history from a fully-instrumented prototype provides a representative data set of anticipated operating parameters for forming an empirical model in a computer module for generating “virtual” signals for a process or machine in real-time. Replacement or inferential sensor signals can be advantageously used in downstream control processing or analysis. The replacement or inferred sensor signals generated accordingly may be returned to a control system or display system local to the process or machine, or located at the remote location or yet a different remote location. The system can provide expected parameter values, as well as the differences between expected and input signals; or the system can provide raw measures of similarity between the collection of input signals and the collection of acceptable modeled states. A memory for storing the representative training set, or a transformation thereof, is coupled to a processor. The processor receives from an input data signals embodying real values from sensors actually on the process or machine, and may receive these in real-time. The processor is disposed to take a set of readings of the actual sensors from the input, and generate an estimate of one or more desired inferred sensors, using a linear combination of the representative training set sensor data, as weighted by the result of a measure of similarity of the input sensor data to the representative training set sensor data.

Claims

1. A system for monitoring operational equipment, comprising:

a sensor capable of being mounted on the equipment;

a data acquisition front-end coupled with said sensor for providing a plurality of component signals derived from said sensor; and

an information processor operable in response to said sensor for using a relationship observed between the plurality of component signals and one or more process parameters of the equipment to generate a parametric signal corresponding to the one or more process parameters of the equipment.

2. A system as recited in claim 1, comprising a plurality of sensors.

3. A system as recited in claim 2, wherein said information processor uses a similarity operator to obtain a similarity measure between the plurality of component signals and one of the process parameters to generate the parametric signal.

4. A system as recited in claim 1, wherein said information processor estimates the parametric signal corresponding to one of the process parameters.

5. A system as recited in claim 2, wherein said information processor estimates the parametric signal corresponding to a process parameter being monitored by one of said plurality of sensors.

6. A system as recited in claim 2, wherein said information processor determines a data acquisition failure associated with said sensor such that said data acquisition front-end is inhibited from providing the component signals from said sensor to said information processor, and wherein said information processor estimates the parametric signal corresponding to the component signals from said sensor in response to the data acquisition failure.

7. A system as recited in claim 6, wherein the data acquisition failure corresponds to a failure of said sensor, and said information processor determines the failure of said sensor.

8. A system as recited in claim 6, wherein the component signals from said sensor comprise a process parameter.

9. A system as recited in claim 1, wherein the one or more process parameters of the equipment correspond to physical operating characteristics including temperature, pressure, or displacement of the operational equipment being monitored.

10. An inferential signal generator, comprising:

a memory for storing a set of time-correlated sensor data representative of expected operational states of a system selected from a process or a machine;

a signal acquisition input for receiving sensor values from said system while in operation, descriptive of physical parameters of said system;

a processor disposed to receive the sensor values from said signal acquisition input, and generate at least one output value descriptive of a parameter of the system that is not among parameters measured by the received sensor values, by comparing the similarity of the sensor values from said signal acquisition input to the set of sensor data in said memory.

11. An inferential signal generator as recited in claim 10, wherein said processor generates the at least one output value by a linear combination of the set of sensor data, adjusted according to said similarity of the sensor values to said set of sensor data.

12. An inferential signal generator as recited in claim 11, wherein the similarity of the sensor values from said signal acquisition input to the set of sensor data in said memory is measured by an element by element comparison of a numerical closeness of like sensor values with a full range of values for like sensors in said memory.

13. An inferential signal generator as recited in claim 11, wherein the set of sensory data in said memory comprises a plurality of snapshots of sensor data, each such snapshot comprising time-correlated values from the sensors on the system.

14. An inferential signal generator as recited in claim 13, wherein the sensor values received by said signal acquisition input from said system while in operation comprise a snapshot of time-correlated values from a subset of the sensors on the system.

15. An inferential signal generator as recited in claim 14, wherein said processor generates the at least one output value by a linear combination of the plurality of snapshots in said memory, adjusted according to said similarity of each such snapshot to the snapshot received by said signal acquisition input.

16. An inferential signal generator as recited in claim 10, wherein said signal acquisition input comprises a data bus electrically connected to sensors of said system.

17. An inferential signal generator as recited in claim 10, wherein said signal acquisition input comprises a wireless communications link disposed to receive transmissions containing the sensor values.

18. A control apparatus for a system selected from a process or machine, comprising:

at least one actuator signal line connected to said system for conducting a control signal to said system;

at least one sensor signal line connected to said system for conducting a signal from at least one sensor on said system indicative of a physical parameter of said system;

a memory for storing a set of time-correlated sensor data representative of expected operational states of said system; and

a processor disposed to receive the signal from the at least one sensor signal line, for generating a control signal provided to said actuator signal line, by comparing the similarity of the signal from said sensor signal line to the set of sensor data in said memory to generate at least one computed value descriptive of a second physical parameter of said system, on the basis of which said control signal is generated.

19. A control apparatus as recited in claim 18, wherein said processor generates the at least one computed value from a combination of the set of sensor data, adjusted according to said similarity of the sensor signal to said set of sensor data

20. A control apparatus as recited in claim 19, wherein said similarity is measured by a numerical closeness of the sensor signal from said at least one sensor signal line to like sensor data in said memory.

21. A method for generating at least one inferential sensor signal for a system selected from a machine or process, comprising the steps of:

instrumenting said system with sensors to measure a plurality of physical parameters of said system;

operating said system in at least one operational state;

acquiring sensor data from said sensors for the plurality of physical parameters for the at least one operational state;

storing the acquired sensor data in a memory;

monitoring values of a subset of said sensors while said system is in operation; and

computing at least one inferential sensor signal corresponding to a physical parameter not among those measured by said subset of sensors, by linearly combining the stored sensor data according to a measure of similarity of the monitored sensor values to the stored sensor data.

22. A method according to claim 21, wherein said measure of similarity is an element-by-element comparison of a numerical closeness of values of the monitored sensor values to values of like sensors in the stored sensor data.

23. A computer for generating virtual sensor signals describing a physical system, comprising:

a memory for storing a plurality of reference state vectors, each state vector having as elements sensor values descriptive of a physical state of said system;

a first computing module disposed to receive a current state vector describing said physical system, having as elements sensor values corresponding to a subset of sensors in each of said reference state vectors, for comparing the current state vector to each said reference state vector to determine a measure of similarity therebetween; and

a second computing module disposed to receive the measures of similarity between the current state vector and each said reference state vector, and generate from a combination of said reference state vectors as weighted by the measures of similarity, a virtual signal vector having as elements estimates of those sensor values corresponding to sensors in said reference state vectors remaining out of said subset of sensors.

24. A computer as recited in claim 23, wherein said second computing module generates at least one output value descriptive of a parameter of the physical system to provide a virtual signal in addition to the sensor values corresponding to the subset of sensors of said reference state vectors.

25. A computer as recited in claim 24, wherein said second computing module generates the virtual signal as a linear combination of said reference state vectors adjusted according to the measure of similarity of said reference state vectors to observed sensor data values descriptive of the system.

26. A system for monitoring instrumented equipment, comprising:

a plurality of sensors capable of being monitored from the instrumented equipment;

an information processor;

a data acquisition input to said information processor for receiving operational values descriptive of physical parameters of the system from said plurality of sensors;

a memory operable for being coupled with said signal acquisition input for storing a set of time-correlated sensor data representative of expected operational states, and signals observed during operation of the instrumented equipment;

said information processor generating at least one output value descriptive of a parameter of the instrumented equipment in addition to signals observed from the instrumented equipment; and

said information processor being operable in response to said signal acquisition input using relationships between the expected operational states and the observed signals from the instrumented equipment to generate additional signals corresponding to one or more process parameters of the equipment from the output value descriptive of the parameter of the instrumented equipment in addition to the signals observed.

27. A system as recited in claim 26, wherein said information processor uses a similarity operator to obtain a similarity measure between the expected operational states and the observed signals from the instrumented equipment.

28. A system as recited in claim 26, wherein said information processor determines a data acquisition failure associated with said data acquisition input from at least one of said variety of sensors.

29. A system as recited in claim 26, wherein said information processor uses a similarity operator with data in said memory comprising a plurality of snapshots of the operational states descriptive of the physical parameters of the system from said plurality of sensors, each snapshot comprising time-correlated data from said plurality of sensors.

30. A system as recited in claim 26, wherein the relationships used by said information processor measure similarity as a numerical closeness of the observed signals with the expected operational states.