Utility demand forecasting using utility demand matrix

• US 7,647,137 B2
• Filed: 03/13/2007
• Issued: 01/12/2010
• Est. Priority Date: 03/13/2007
• Status: Expired due to Fees

• Current Assignee: Honeywell International Inc.
• Original Assignee: Honeywell International Inc.
• Inventors: Schindler, Zdenek
• Primary Examiner(s): Shechtman; Sean P
• Application Number:
  US11/717,525
• Publication Number:
  US 20080228325A1
• Time in Patent Office: 1,036 Days
• Field of Search: 700/16, 700/83, 700/276, 700/286, 700/291, 700/296, 705/412, 715/963, 702 60- 62, 707/3, 707/6, 707/104.1
• US Class Current: 700/291
• CPC Class Codes: G06Q 10/04 Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem" market predictions or forecasting for commercial activities G06Q30/0202

Abstract

One embodiment of the application provides a method of forecasting utility demand of a utility consuming entity for a future time period. The method includes: selecting a utility demand matrix from a set of utility demand matrixes as a function of a future date, wherein the set of utility demand matrixes are classified by type-of-day, and representing a relation of past utility demand of the entity to time-in-day interval and outdoor temperature in the vicinity of the entity during the type-of-day; identifying a utility demand element from the utility demand matrix as a function of a future time-in-day and a known factor (e.g., future outdoor temperature) thereof; and generating a utility demand forecast of the entity for the future time period as a function of the identified utility demand element.

Claims

1. A machine implemented method of forecasting utility demand of a utility consuming entity for a future time period, the method comprising:

selecting a utility demand matrix from a set of utility demand matrixes as a function of a future date, wherein the set of utility demand matrixes are classified by type-of-day, and representing a relation of past utility demand of the entity to time-in-day and outdoor temperature in the vicinity of the entity during the type-of-day;

identifying a utility demand element from the utility demand matrix as a function of a future time-in-day interval and a known factor thereof; and

generating, by the machine a utility demand forecast of the entity for the future time period as a function of the identified utility demand element.

2. The method of claim 1, wherein the utility consuming entity is a building, or building complex.

3. The method of claim 1, wherein the utility demand is electricity consumption, natural gas consumption, water consumption, heating consumption, or cooling consumption of the utility consuming entity.

4. The method of claim 1, wherein the type-of-day represents a type of workday, weekday, holiday or weekend during spring, summer, fall or winter as defined by the Gregorian calendar.

5. The method of claim 1, wherein a time resolution of the time-in-day in the utility demand matrix is 15, 30, or 60 minutes.

6. The method of claim 1, wherein the utility demand element from the utility demand matrix is a historical data of demand influenced by one or more factors selected from the group consisting of past outdoor temperature, past wind speed, past wind direction, past humidity, and past occupancy during a past time-in-day.

7. The method of claim 1, wherein the known factor is a factor selected from the group consisting of future outdoor temperature, future wind speed, future wind direction, future humidity, and future occupancy during the future time-in-day interval or created from their suitable combination.

8. The method of claim 1, wherein a temperature resolution of the outdoor temperature in the utility demand matrix is one degree Centigrade or Fahrenheit.

9. The method of claim 1, wherein each row of the utility demand matrix represents utility demands during the same time-in-day, and each column of the utility demand matrix represents utility demands at the same outdoor temperature.

10. The method of claim 1, wherein each row of the utility demand matrix represents utility demands at the same outdoor temperature, and each column of the utility demand matrix represents utility demands during the same time-in-day.

11. The method of claim 1, further comprising:

evaluating whether the selected utility demand matrix is stationary; and

updating the set of utility demand matrixes each time when new utility demand data is available.

12. The method of claim 11, wherein updating the set of utility demand matrixes is performed by giving a same weight to the new utility data as to a past utility data, if the selected utility demand matrix is found stationary in the evaluating step.

13. The method of claim 11, wherein updating the set of utility demand matrixes is performed by giving a higher weight to the new utility data as compared to a past utility data, if the selected utility demand matrix found nonstationary in the evaluating step.

14. A machine implemented method of forecasting utility demand of a utility consuming entity for a future time period, comprising:

obtaining access to a set of utility demand matrixes, classified by type-of-day, and representing a relation of past utility demand of the entity to time-in-day and outdoor temperature in the vicinity of the entity during the type-of-day;

receiving a future date, time-in-day, and expected outdoor temperature in the vicinity of the utility consuming entity;

selecting a utility demand matrix from the set of utility demand matrixes as a function of the future date;

matching respectively the future time-in-day and expected outdoor temperature with the time-in-day and the outdoor temperature existing in the selected utility demand matrix;

identifying a utility demand element from the selected utility demand matrix as a function of the time-in-day and the outdoor temperature thereof; and

providing, by the machine, a utility demand forecast of the entity for the future time period as a function of the utility demand element.

15. A system of forecasting utility demand of a utility consuming entity for a future time period, comprising:

means for accessing a set of utility demand matrixes that are classified by type-of-day, and representing a relation of past utility demand of the entity to time-in-day and outdoor temperature in the vicinity of the utility consuming entity during the type-of-day;

a utility demand matrix selecting module, to select a utility demand matrix from the set of utility demand matrixes as a function of a future date;

a utility demand element identifying module, to identify the utility demand element from the selected utility demand matrix as a function of a future time-in-day and expected outdoor temperature; and

a machine, to forecast the utility demand of the utility consuming entity for the future time period as a function of the identified utility demand element.

16. The system of claim 15, wherein the type-of-day represents a type of workday, weekday, holiday or weekend during spring, summer, fall or winter as defined by the Gregorian calendar.

17. The system of claim 16, wherein each row of the utility demand matrix represents utility demands during the same time-in-day, and each column of the utility demand matrix represents utility demands at the same outdoor temperature.

18. The system of claim 15, wherein each row of the utility demand matrix represents utility demands at the same outdoor temperature, and each column of the utility demand matrix represents utility demands during the same time-in-day.

19. A computer-readable medium including instructions that, when at least partially performed by a computer in a system, cause the computer to:

to select a utility demand matrix from a set of utility demand matrixes as a function of an input future date, wherein the set of utility demand matrixes is classified by type-of-day, and represents a relation of past utility demand of the entity to time-in-day and outdoor temperature in the vicinity of the entity during the type-of-day;

to identify a utility demand element from the utility demand matrix, by matching respectively a future time-in-day and expected outdoor temperature with the time-in-day and outdoor temperature in the utility demand matrix; and

to generate a utility demand forecast of the entity for the future time period as a function of the identified utility demand element.

20. The computer-readable medium of claim 19, further comprising instructions that cause the computer:

to evaluate whether the selected utility demand matrix is stationary; and

to update the set of utility demand matrixes each time when new utility demand data is available.

21. The computer-readable medium of claim 20, wherein the instructions cause the computer to update the set of utility demand matrixes by giving a same weight to the new utility data as to a past utility data, if the selected utility demand matrix is found stationary in the evaluating step.

22. The computer-readable medium of claim 20, wherein the instructions cause the computer to update the set of utility demand matrixes by giving a higher weight to the new utility data as compared to a past utility data, if the selected utility demand matrix found nonstationary in the evaluating step.