Systems and methods for adaptively updating equipment models

• US 10,317,864 B2
• Filed: 09/01/2017
• Issued: 06/11/2019
• Est. Priority Date: 12/22/2014
• Status: Active Grant

• Current Assignee: Tyco Fire & Security GmbH (Johnson Controls International plc)
• Original Assignee: Johnson Controls Technology Company (Johnson Controls International plc)
• Inventors: Boettcher, Andrew J., Vitullo, Steven R., Drees, Kirk H., Wenzel, Michael J.
• Primary Examiner(s): Tran, Vincent H
• Application Number:
  US15/694,033
• Publication Number:
  US 20180011459A1
• Time in Patent Office: 648 Days
• CPC Class Codes: G05B 19/042 using digital processors G05B19/05 takes precedence
• CPC Class Codes: G05B 2219/2642 Domotique, domestic, home control, automation, smart house
• CPC Class Codes: G05B 23/024 Quantitative history assessment, e.g. mathematical relationships between available data; Functions therefor; Principal component analysis [PCA]; Partial least square [PLS]; Statistical classifiers, e.g. Bayesian networks, linear regression or correlation analysis; Neural networks

Abstract

A system for generating and using a predictive model to control building equipment includes building equipment operable to affect one or more variables in a building and an operating data aggregator that collects a set of operating data for the building equipment. The system includes an autocorrelation corrector that removes an autocorrelated model error from the set of operating data by determining a residual error representing a difference between an actual output of the building equipment and an output predicted by the predictive model, using the residual error to calculate an autocorrelation for the model error, and transforming the set of operating data using the autocorrelation. The system includes a model generator module that generates a set of model coefficients for the predictive model using the transformed set of operating data and a controller that controls the building equipment by executing a model-based control strategy that uses the predictive model.

Claims

1. A system for generating and using a predictive model to control building equipment, the system comprising:

building equipment operable to affect one or more variables in a building;

an operating data aggregator configured to collect a first set of operating data for the building equipment during a first time period;

an autocorrelation corrector configured to remove an autocorrelated model error from the first set of operating data by;

determining a residual error representing a difference between an actual output of the building equipment and an output predicted by the predictive model;

using the residual error to calculate an autocorrelation for the model error; and

transforming the first set of operating data using the autocorrelation;

a model generator module configured to generate a first set of model coefficients for the predictive model using the transformed first set of operating data; and

a controller configured to control the building equipment by executing a model-based control strategy that uses the predictive model.

2. The system of claim 1, further comprising a model update module configured to:

generate a new set of model coefficients for the predictive model;

determine whether the new set of model coefficients improves a fit of the predictive model to a set of operating data relative to the first set of model coefficients and

replace the first set of model coefficients with the new set of model coefficients in the predictive model in response to a determination that the new set of model coefficients improves the fit of the predictive model.

3. The system of claim 1, further comprising a model update module configured to:

generate a new set of model coefficients for the predictive model;

determine whether the new set of model coefficients improves a fit of the predictive model to a set of operating data relative to the first set of model coefficients; and

retain the first set of model coefficients in the predictive model in response to a determination that the new set of model coefficients does not improve the fit of the predictive model.

4. The system of claim 1, further comprising a model update module configured to:

adjust the first time period to define an adjusted time period based on a current time; and

iteratively update the first set of model coefficients using a set of the operating data corresponding to the adjusted time period.

5. The system of claim 1, wherein the controller comprises a demand response module configured to use the predictive model to generate a control output for the building equipment using a model-based control methodology.

6. The system of claim 1, wherein:

the operating data aggregator is configured to collect a second set of operating data for the building equipment during a second time period; and

the autocorrelation corrector configured to remove an autocorrelated model error from the second set of operating data; and

the model generator module configured to generate a second set of model coefficients for the predictive model using the transformed second set of operating data.

7. The system of claim 6, further comprising:

a test statistic module configured to generate a test statistic based on a difference between the first set of model coefficients and the second set of model coefficients; and

a critical value module configured to calculate a critical value for the test statistic.

8. The system of claim 6, further comprising:

a hypothesis testing module configured to perform a statistical hypothesis test comprising comparing the test statistic with the critical value to determine whether the predictive model has changed; and

a model update module configured to adaptively update the predictive model in response to a determination that the test statistic exceeds the critical value.

9. The system of claim 8, wherein calculating the critical value comprises:

identifying a parameter representing a predetermined likelihood that the statistical hypothesis test improperly rejects a null hypothesis that the predictive model has not changed; and

using an inverse cumulative distribution function for the test statistic to determine, based on the parameter, the critical value for the test statistic such that the test statistic has the predetermined likelihood of exceeding the critical value when the predictive model has not changed.

10. A system for detecting a fault in a predictive model for building equipment, the system comprising:

building equipment operable to affect one or more variables in a building;

an operating data aggregator module configured to collect a first set of operating data for the building equipment during a first time period and a second set of operating data for the building equipment during a second time period;

an autocorrelation corrector configured to remove an autocorrelated model error from the first set of operating data and the second set of operating data by;

determining a residual error representing a difference between an actual output of the building equipment and an output predicted by the predictive model;

using the residual error to calculate an autocorrelation for the model error; and

transforming the first set of operating data and the second set of operating data using the autocorrelation;

a model generator module configured to generate a first set of model coefficients for the predictive model using the transformed first set of operating data and a second set of model coefficients for the predictive model using the transformed second set of operating data;

a fault detection module configured to generate a fault indication in response to a determination that the predictive model has changed, the determination based on the first and second sets of model coefficients; and

a controller configured to control the building equipment by executing a model-based control strategy that uses the predictive model.

11. The system of claim 10, wherein generating the fault indication comprises:

generating a fault event indicating that the predictive model has changed; and

appending to the fault event a statistical confidence that the predictive model has changed, the statistical confidence based on a parameter of a statistical hypothesis test.

12. The system of claim 10, wherein the model generator module is configured to:

adjust at least one of the first time period and the second time period to define an adjusted time period based on a current time; and

iteratively update at least one of the first set of model coefficients and the second set of model coefficients using a set of the operating data corresponding to the adjusted time period.

13. The system of claim 10, wherein the controller comprises a demand response module configured to use the predictive model to generate a control output for the building equipment using a model-based control methodology.

14. The system of claim 10, further comprising:

a test statistic module configured to generate a test statistic based on a difference between the first set of model coefficients and the second set of model coefficients; and

a critical value module configured to calculate a critical value for the test statistic.

15. The system of claim 14, further comprising a hypothesis testing module configured to perform a statistical hypothesis test comprising comparing the test statistic with the critical value to determine whether the predictive model has changed.

16. The system of claim 15, wherein calculating the critical value comprises:

identifying a parameter representing a predetermined likelihood that the statistical hypothesis test improperly rejects a null hypothesis that the predictive model has not changed; and

using an inverse cumulative distribution function for the test statistic to determine, based on the parameter, the critical value for the test statistic such that the test statistic has the predetermined likelihood of exceeding the critical value when the predictive model has not changed.

17. A method for identifying changes in a predictive model for building equipment, the method comprising:

operating building equipment to affect one or more variables in a building;

collecting a first set of operating data for the building equipment during a first time period;

removing an autocorrelated model error from the first set of operating data by;

determining a residual error representing a difference between an actual output of the building equipment and an output predicted by the predictive model;

using the residual error to calculate an autocorrelation for the model error; and

transforming the first set of operating data using the autocorrelation;

generating a first set of model coefficients for the predictive model using the transformed first set of operating data;

in response to a determination that the predictive model has changed, triggering an action comprising at least one of generating a fault indication and adaptively updating the predictive model; and

controlling the building equipment by executing a model-based control strategy that uses the predictive model.

18. The method of claim 17, further comprising:

collecting a second set of operating data for the building equipment during a second time period;

removing an autocorrelated model error from the second set of operating data; and

generating a second set of model coefficients for the predictive model using the transformed second set of operating data.

19. The method of claim 18, wherein adaptively updating the predictive model comprises:

determining whether the second set of model coefficients improves a fit of the predictive model to a set of operating data relative to the first set of model coefficients;

replacing the first set of model coefficients with the second set of model coefficients in the predictive model in response to a determination that the second set of model coefficients improves the fit of the predictive model; and

retaining the first set of model coefficients in the predictive model in response to a determination that the second set of model coefficients does not improve the fit of the predictive model.

20. The method of claim 17, wherein generating the fault indication comprises:

generating a fault event indicating that the predictive model has changed; and

appending to the fault event a statistical confidence that the predictive model has changed, the statistical confidence based on a parameter of a statistical hypothesis test.