Using a digital twin to facilitate environment-specific prognostic-surveillance operations for engineering assets in the field
During operation, the system receives time-series signals from sensors in the asset while the asset is operating. Next, the system obtains real-time environmental parameters for an environment in which the asset is operating. The system then selects an environment-specific inferential model for the asset based on the real-time environmental parameters, wherein the environment-specific inferential model was trained on a golden system while the golden system was operating in an environment that matches the real-time environmental parameters. Next, the system uses the environment-specific inferential model to generate estimated values for the received time-series signals based on correlations among the received time-series signals, and performs a pairwise-differencing operation between actual values and the estimated values for the received time-series signals to produce residuals. Finally, the system determines from the residuals whether the asset is operating correctly.
1. A method for performing environment-specific prognostic-surveillance operations for an asset, comprising:
receiving time-series signals from sensors in the asset while the asset is operating;
obtaining real-time environmental parameters for an environment in which the asset is operating;
selecting an environment-specific inferential model for the asset based on the real-time environmental parameters, wherein the environment-specific inferential model was trained on a golden system while the golden system was operating in an environment that matches the real-time environmental parameters, wherein a configuration of the golden system matches a configuration of the asset, and wherein the golden system was certified to be operating correctly;
using the environment-specific inferential model to generate estimated values for the received time-series signals based on correlations among the received time-series signals;
performing a pairwise-differencing operation between actual values and the estimated values for the received time-series signals to produce residuals; and
determining from the residuals whether the asset is operating correctly.
2. The method of claim 1 , wherein determining from the residuals whether the asset is operating correctly involves:
performing a sequential probability ratio test (SPRT) on the residuals to produce SPRT alarms; and
determining from the SPRT alarms whether the asset is operating correctly.
3. The method of claim 1 , wherein prior to receiving the time-series signals, the method further comprises generating a set of environment-specific inferential models by:
gathering environment-specific time-series signals from the golden system while the golden system operates in an environmental-testing facility, which is cycling through different combinations of environmental parameters;
training a different environment-specific inferential model using time-series signals for each of the different combinations of environmental parameters; and
storing the different environment-specific inferential models to facilitate subsequent environment-specific prognostic-surveillance operations.
4. The method of claim 1 ,
wherein using the environment-specific inferential model to generate the estimated values for the received time-series signals involves generating estimated values for a sliding time window of values for the received time-series signals;
wherein the method further comprises using a telemetry parameter synthesis (TPSS) technique to project the estimated values for the received time-series signals into the future based on the time window; and
wherein the pairwise-differencing operation is performed between the projected estimated values and the actual values for the received time-series signals.
5. The method of claim 1 , wherein the real-time environmental parameters include one or more of the following:
an ambient temperature;
a relative humidity; and
an altitude.
6. The method of claim 1 , wherein the asset comprises one of the following:
a component in a power-generation system; and
a component in a power-transmission system.
7. The method of claim 1 , wherein the inferential model comprises a Multivariate State Estimation Technique (MSET) model.
8. A non-transitory, computer-readable storage medium storing instructions that when executed by a computer cause the computer to execute a method for performing environment-specific prognostic-surveillance operations for an asset, the method comprising:
receiving time-series signals from sensors in the asset while the asset is operating;
obtaining real-time environmental parameters for an environment in which the asset is operating;
selecting an environment-specific inferential model for the asset based on the real-time environmental parameters, wherein the environment-specific inferential model was trained on a golden system while the golden system was operating in an environment that matches the real-time environmental parameters, wherein a configuration of the golden system matches a configuration of the asset, and wherein the golden system was certified to be operating correctly;
using the environment-specific inferential model to generate estimated values for the received time-series signals based on correlations among the received time-series signals;
performing a pairwise-differencing operation between actual values and the estimated values for the received time-series signals to produce residuals; and
determining from the residuals whether the asset is operating correctly.
9. The non-transitory, computer-readable storage medium of claim 8 , wherein determining from the residuals whether the asset is operating correctly involves:
performing a SPRT on the residuals to produce SPRT alarms; and
determining from the SPRT alarms whether the asset is operating correctly.
10. The non-transitory, computer-readable storage medium of claim 8 , wherein prior to receiving the time-series signals, the method further comprises generating a set of environment-specific inferential models by:
gathering environment-specific time-series signals from the golden system while the golden system operates in an environmental-testing facility, which is cycling through different combinations of environmental parameters;
training a different environment-specific inferential model using time-series signals for each of the different combinations of environmental parameters; and
storing the different environment-specific inferential models to facilitate subsequent environment-specific prognostic-surveillance operations.
11. The non-transitory, computer-readable storage medium of claim 8 ,
wherein using the environment-specific inferential model to generate the estimated values for the received time-series signals involves generating estimated values for a sliding time window of values for the received time-series signals;
wherein the method further comprises using a TPSS technique to project the estimated values for the received time-series signals into the future based on the time window; and
wherein the pairwise-differencing operation is performed between the projected estimated values and the actual values for the received time-series signals.
12. The non-transitory, computer-readable storage medium of claim 8 , wherein the real-time environmental parameters include one or more of the following:
an ambient temperature;
a relative humidity; and
an altitude.
13. The non-transitory, computer-readable storage medium of claim 8 , wherein the asset comprises one of the following:
a component in a power-generation system; and
a component in a power-transmission system.
14. The non-transitory, computer-readable storage medium of claim 8 , wherein the inferential model comprises an MSET model.
15. A system that performs environment-specific prognostic-surveillance operations for an asset, comprising:
at least one processor and at least one associated memory; and
a prognostic-surveillance mechanism that executes on the at least one processor, wherein during operation, the prognostic-surveillance mechanism:
receives time-series signals from sensors in the asset while the asset is operating;
obtains real-time environmental parameters for an environment in which the asset is operating;
selects an environment-specific inferential model for the asset based on the real-time environmental parameters, wherein the environment-specific inferential model was trained on a golden system while the golden system was operating in an environment that matches the real-time environmental parameters, wherein a configuration of the golden system matches a configuration of the asset, and wherein the golden system was certified to be operating correctly;
uses the environment-specific inferential model to generate estimated values for the received time-series signals based on correlations among the received time-series signals;
performs a pairwise-differencing operation between actual values and the estimated values for the received time-series signals to produce residuals; and
determines from the residuals whether the asset is operating correctly.
16. The system of claim 15 , wherein while determining from the residuals whether the asset is operating correctly, the prognostic surveillance mechanism:
performs a sequential probability ratio test (SPRT) on the residuals to produce SPRT alarms; and
determines from the SPRT alarms whether the asset is operating correctly.
17. The system of claim 15 , wherein prior to receiving the time-series signals, the prognostic-surveillance mechanism generates a set of environment-specific inferential models, during which process, the prognostic-surveillance mechanism:
gathers environment-specific time-series signals from the golden system while the golden system operates in an environmental-testing facility, which is cycling through different combinations of environmental parameters;
trains a different environment-specific inferential model using time-series signals for each of the different combinations of environmental parameters; and
stores the different environment-specific inferential models to facilitate subsequent environment-specific prognostic-surveillance operations.
18. The system of claim 15 ,
wherein while using the environment-specific inferential model to generate the estimated values for the received time-series signals, the prognostic-surveillance mechanism generates estimated values for a sliding time window of values for the received time-series signals;
wherein the prognostic-surveillance mechanism additionally uses a TPSS technique to project the estimated values for the received time-series signals into the future based on the time window; and
wherein the pairwise-differencing operation is performed between the projected estimated values and the actual values for the received time-series signals.
19. The system of claim 15 , wherein the real-time environmental parameters include one or more of the following:
an ambient temperature;
a relative humidity; and
an altitude.
20. The system of claim 15 , wherein the asset comprises one of the following:
a component in a power-generation system; and
a component in a power-transmission system.