Systems and methods for determining the uncertainty in parameters of an energy use model

• US 9,355,069 B2
• Filed: 12/20/2013
• Issued: 05/31/2016
• Est. Priority Date: 12/20/2013
• Status: Active Grant

• Current Assignee: Tyco Fire & Security GmbH (Johnson Controls International plc)
• Original Assignee: Johnson Controls Technology Company (Johnson Controls International plc)
• Inventors: ElBsat, Mohammad N., Wenzel, Michael J.
• Primary Examiner(s): Chen, Alan
• Application Number:
  US14/137,627
• Publication Number:
  US 20150178633A1
• Time in Patent Office: 893 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G05B 13/048 using a predictor
• CPC Class Codes: G05B 15/02 electric
• CPC Class Codes: G05B 2219/2642 Domotique, domestic, home control, automation, smart house
• CPC Class Codes: G05F 1/66 Regulating electric power
• CPC Class Codes: G06F 17/16 Matrix or vector computation , e.g. matrix-matrix or matrix-vector multiplication, matrix factorization matrix transposition G06F7/78
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks

Abstract

Systems and methods for determining the uncertainty in parameters of a building energy use model are provided. A disclosed method includes receiving an energy use model for a building site. The energy use model includes one or more predictor variables and one or more model parameters. The method further includes calculating a gradient of an output of the energy use model with respect to the model parameters, determining a covariance matrix using the calculated gradient, and using the covariance matrix to identify an uncertainty of the model parameters. The uncertainty of the model parameters may correspond to entries in the covariance matrix.

Claims

1. A method for determining an uncertainty in parameters of a building energy use model, the method comprising:

receiving an energy use model for a building site, the energy use model having one or more model parameters, wherein the energy use model is used to predict an energy consumption of the building site as a function of the one or more model parameters and one or more predictor variables;

calculating a gradient of an output of the energy use model with respect to the model parameters;

determining a covariance matrix using the calculated gradient; and

using the covariance matrix to identify an uncertainty of the model parameters.

2. The method of claim 1, wherein the uncertainty of the model parameters is a function of entries in the covariance matrix.

3. The method of claim 1, wherein at least one of the model parameters is a balance point parameter defining a range of outside air temperatures within which the output of the energy use model depends on the outside air temperature.

4. The method of claim 1, wherein at least one of the predictor variables is a function of a balance point parameter of the energy use model, wherein determining the covariance matrix comprises:

identifying, in the energy use model, a predictor variable that is a function of the balance point parameter;

identifying a regression coefficient associated with the identified predictor variable; and

using a function of the identified regression coefficient as a gradient of the output of the energy use model with respect to the balance point parameter.

5. The method of claim 4, wherein the balance point is a threshold temperature value defining a range of outside air temperatures within which the output of the energy use model depends on the outside air temperature;

wherein the function of the identified regression coefficient is a product of the identified regression coefficient and a variable representing a total time duration during which the outside air temperature is within the outside air temperature range.

6. The method of claim 1, wherein determining the covariance matrix comprises determining a standard error of regression for the energy use model.

7. The method of claim 1, further comprising:

obtaining a plurality of data points, each of the data points comprising a value for the one or more predictor variables and an associated energy consumption value for the building site; and

estimating the model parameters using the plurality of data points.

8. The method of claim 7, wherein obtaining the plurality of data points comprises, for each of the data points:

receiving at least one of;

an observed temperature value and an observed enthalpy value; and

calculating the value of the predictor variable using the observed temperature value or the observed enthalpy value.

9. The method of claim 1, wherein the one or more predictor variables comprise at least one weather-related predictor variable,

wherein the weather-related predictor variable is at least one of:

cooling degree days, heating degree days, cooling energy days, heating energy days, temperature, and enthalpy.

10. The method of claim 1, wherein the energy use model is at least one of:

a non-linear energy use model, a piecewise linear energy use model, a three-parameter energy use model, and a five-parameter energy use model.

11. The method of claim 1, further comprising:

updating the uncertainty of the model parameters;

applying inputs to the energy use model;

conducting a performance analysis using the energy use model, wherein a result of the performance analysis is a function of one or more of the model parameters;

providing an output using the result of the performance analysis; and

determining an uncertainty of the output using the updated uncertainty of the model parameters.

12. The method of claim 1, further comprising:

using the uncertainty of the model parameters to perform a multivariate uncertainty analysis of the model parameters.

13. The method of claim 12, wherein the multivariate uncertainty analysis allows for a visualization of a correlation between two or more of the model parameters.

14. The method of claim 12, wherein the multivariate uncertainty analysis allows for performing a multivariate peer analysis of the model parameters.