Method and system for forecasting power requirements using granular metrics

• US 9,366,704 B2
• Filed: 05/28/2014
• Issued: 06/14/2016
• Est. Priority Date: 11/29/2011
• Status: Active Grant

• Current Assignee: Generac Power Systems Incorporated (Generac Holdings Incorporated)
• Original Assignee: Neurio Technology Inc. (Generac Holdings Incorporated)
• Inventors: Guo, Zhenyu, Haghighat-Kashani, Ali, Cheam, Janice Tze-Nee, Hallam, Jonathan Mark
• Primary Examiner(s): Bui, Bryan
• Application Number:
  US14/288,630
• Publication Number:
  US 20140336960A1
• Time in Patent Office: 748 Days
• Field of Search: 702/60
• US Class Current: 1/1
• CPC Class Codes: G01R 21/133 by using digital technique
• CPC Class Codes: G06F 16/212 with details for data modelling support
• 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
• CPC Class Codes: G06Q 50/06 Energy or water supply
• CPC Class Codes: H02J 13/00002 characterised by monitoring
• CPC Class Codes: H02J 13/00016 using a wired telecommunication network or a data transmission bus
• CPC Class Codes: H02J 13/00026 involving a local wireless network, e.g. Wi-Fi, ZigBee or Bluetooth
• CPC Class Codes: H02J 13/00028 involving the use of Internet protocols
• CPC Class Codes: H02J 3/003 Load forecast, e.g. methods or systems for forecasting future load demand
• CPC Class Codes: Y02D 10/00 Energy efficient computing, e.g. low power processors, power management or thermal management
• CPC Class Codes: Y02E 60/00 Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
• CPC Class Codes: Y04S 10/30 State monitoring, e.g. fault, temperature monitoring, insulator monitoring, corona discharge
• CPC Class Codes: Y04S 10/50 Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
• CPC Class Codes: Y04S 40/124 using wired telecommunication networks or data transmission busses
• CPC Class Codes: Y04S 40/126 using wireless data transmission

Abstract

A method for modeling power usage within a macrogrid uses data relating to the behavioral patterns and states (“BA”) of the users, data relating to external impacts on power usage and disaggregated power consumption data in at least one premises within the macrogrid (forming “power usage model data”) and thereafter a method of forecasting and predicting future power requirements within the macrogrid uses such power usage model data.

Claims

1. A computer implemented method of modeling power usage within a macrogrid, the macrogrid being within a region, the method comprising:

a) obtaining, by a processor, disaggregated power consumption data from at least one premises within the macrogrid to determine at least one behavioral pattern according to usage behaviors for at least two appliances within the premises based on;

storing, in a database, power signal clips of model appliances that are comparable to the appliances;

creating, for each of the appliances, a model power draw based on a Hidden Markov Model (HMM);

estimating a Factorial HMM for the appliances using an Explicit Duration HMM with Difference Observations, the estimating based on the model power draws;

tuning one or more parameters of said Factorial HMM based on a comparison, of signal clips of power consumed at the premises that include power signal clips of the appliances, with the power signal clips of the model appliances;

calculating a difference between a total power draw of the appliances at a first time and the total power draw of the appliances at a preceding time; and

determining whether a state of a maximum of one of the appliances has changed, based on said difference;

b) obtaining, by the processor, data relating to the at least one behavioral pattern and a state of a user;

c) obtaining, by the processor, data relating to at least one external impact on the user'"'"'s power usage within the premises, said external impact originating from an environment of the user and not from the user;

d) using, by the processor, the obtained data from one or more of a) to c) to create an individual consumer forecast of power usage, said consumer forecast being collectively aggregated across a plurality of users;

e) using, by the processor, data from one or more of a) to d) to perform a demographic analysis;

f) collecting, by the processor, macrogrid aggregate power consumption data for the region;

g) determining, by the processor, power consumption requirements across the macrogrid for the region using data from at least one of a) to e); and

h) determining, by the processor, future power requirements within the macrogrid, using the data obtained in at least one of a) to g).

2. The method of claim 1 wherein said determining the future power requirements is performed for at least one the following:

decreasing and tracking consumer usage of a power resource, decreasing consumer costs for power, maximizing revenue for a power provider, minimizing consumer discomfort, and identifying future peak usage times and requirements.

3. The method of claim 1 wherein said determining whether the state of a maximum of one of the appliances has changed is also based on a duration of how long said one appliance has been in the state up to said first time.

4. A computer implemented method of modeling power usage within a macrogrid for a region, the method comprising:

a) periodically receiving, by a processor, granular power consumption data from a statistically significant portion of a target population of power users within the macrogrid for the region;

b) using, by the processor, the power consumption data to perform a consumer load disaggregation to determine usage behaviors for at least two appliances at a premises based on;

storing, in a database, power signal clips of model appliances that are comparable to the appliances;

creating, for each of the appliances, a model power draw based on a Hidden Markov Model (HMM);

estimating a Factorial HMM for the appliances using an Explicit Duration HMM with Difference Observations, the estimating based on the model power draws;

tuning one or more parameters of said Factorial HMM based on a comparison, of signal clips of power consumed at the premises that include power signal clips of the appliances, with the power signal clips of the model appliances;

calculating a difference between a total power draw of the appliances at a first time and the total power draw of the appliances at a preceding time; and

determining whether a state of a maximum of one of the appliances has changed, based on said difference;

c) obtaining, by the processor, data relating to at least one behavioral pattern and a state of a user;

d) obtaining, by the processor, data relating to external impacts on the user'"'"'s power usage within the premises, said external impact originating from an environment of the user and not from the user;

e) using, by the processor, obtained data from at least one of a) to d) to generate an individual consumer forecast of power usage, the consumer forecast being collectively aggregated across a plurality of users;

f) using, by the processor, data from one or more of a) to e) to perform a demographic analysis;

g) collecting, by the processor, aggregate power consumption data for a region associated with the macrogrid;

h) determining, by the processor, power consumption requirements across the macrogrid for the region using data from at least one of more of a) to g); and

i) determining, by the processor, future power requirements within the macrogrid, using the data obtained in at least one of a) to h).

5. The method of claim 4 wherein periodically comprises receiving data within a five minute interval or less.

6. The method of claim 5 wherein said external impacts are selected from any one of:

weather patterns and predictions, calendar month, economic indicators, world events, television programming, equipment age and maintenance, and population mobility.

7. The method of claim 5 wherein the at least one behavioral pattern comprises data related to user behavioral states selected from one or more of the following states:

sleeping, just awoken, on vacation, working from home, charging electric vehicles, and extended absence.

8. The method of claim 4 wherein periodically comprises receiving data within a one hour interval or less.

9. The method of claim 4 wherein said determining the future power requirements is performed for at least one the following:

decreasing and tracking consumer usage of a power resource, decreasing consumer costs for power, maximizing revenue for a power provider, minimizing consumer discomfort, and identifying future peak usage times and requirements.

10. The method of claim 4 wherein the target population is selected from at least one of:

consumers, commercial operations and industrial operations.

11. The method of claim 4, wherein the statistically signification portion is a population sample which is larger than or equal to a smallest possible sample for which the behavioral patterns of the entire population can be observed and modeled within a required accuracy.

12. A system for modeling power usage within a macrogrid for determining future power requirements within the macrogrid, the system comprising a server and one or more databases, the server comprising a processor and memory, the memory comprising computer executable instructions which, when executed by the processor cause the server to:

a) obtain disaggregated power consumption data from at least one premises within the macrogrid to determine at least one behavioral pattern according to usage behaviors for at least two appliances within one of said premises based on;

storing, in a database, power signal clips of model appliances that are comparable to the appliances;

creating, for each of the appliances, a model power draw based on a Hidden Markov Model (HMM);

estimating a Factorial HMM for the appliances using an Explicit Duration HMM with Difference Observations, the estimating based on the model power draws;

tuning one or more parameters of said Factorial HMM based on a comparison, of signal clips of power consumed at said one of said premises that include power signal clips of the appliances, with the power signal clips of the model appliances;

calculating a difference between a total power draw of the appliances at a first time and the total power draw of the appliances at a preceding time; and

determining whether a state of a maximum of one of the appliances has changed, based on said difference;

b) obtain data relating to the at least one behavioral pattern and a state of a user;

c) obtain data relating to at least one external impact on the user'"'"'s power usage within the premises, said external impact originating from an environment of the user and not from the user;

d) use the obtained data from one or more of a) to c) to create an individual consumer forecast of power usage, said consumer forecast being collectively aggregated across a plurality of users;

e) use data from one or more of a) to d) to perform a demographic analysis;

f) collect macrogrid aggregate power consumption data for the region;

g) determine power consumption requirements across the macrogrid for the region using data from at least one of a) to e); and

h) determine future power requirements within the macrogrid, using the data obtained in at least one of a) to g).

13. The system of claim 12, wherein said external impacts are selected from any one of:

weather patterns and predictions, calendar month, economic indicators, world events, television programming, equipment age and maintenance, and population mobility.

14. The system of claim 12 wherein the at least one behavioral pattern comprises data related to user behavioral states selected from one or more of the following states:

sleeping, just awoken, on vacation, working from home, charging electric vehicles, and extended absence.

15. The system of claim 12 wherein the at least one behavioral pattern comprises data related to user behavioral patterns selected from one or more of the following patterns:

sleeping hours, weekend schedules, weekday schedules, departure to work and arrival home from work, and amount of time spent away from home.

16. The system of claim 12 wherein a forecast period is selected as being one of:

minutes, hours, days, months, seasonal and years.

17. The system of claim 12 further comprising instructions that cause the server to use the forecast to:

regulate power supply and distribution, or identify timing and location of future upgrades.

18. The system of claim 12 further comprising instructions that cause the server to use the forecast to dynamically evaluate power consumption patterns in a selected demographic sector of the population.

19. The system of claim 12 further comprising instructions that cause the server to use the forecast to identify at least one demographic categorization with a higher predictability value than other demographic categorizations previously known.

20. The system of claim 12 further comprising instructions that cause the server to interpolate a sample user forecast onto larger populations with at least one diverse demographic variation.

21. The system of claim 12 further comprising instructions that cause the server to use the forecast to identify one or more periods in a user'"'"'s billing cycle in which a potential for energy conservation is high for that user.

22. The system of claim 12 further comprising instructions that cause the server to use the forecast to identify a power consumption reduction and budget reduction plan.

23. A system for forecasting and predicting power usage within a macrogrid, the system comprising a server and one or more databases, the server comprising a processor and memory, the memory comprising computer executable instructions which, when executed by the processor cause the server to:

a) periodically receive granular power consumption data from a statistically significant portion of a target population of power users within the macrogrid for a region;

b) use the power consumption data to perform a consumer load disaggregation to determine usage behaviors for at least two appliances in a premises within the macrogrid based on;

storing, in a database, power signal clips of model appliances that are comparable to the appliances;

creating, for each of the appliances, a model power draw based on a Hidden Markov Model (HMM);

estimating a Factorial HMM for the appliances using an Explicit Duration HMM with Difference Observations, the estimating based on the model power draws;

tuning one or more parameters of said Factorial HMM based on a comparison, of signal clips of power consumed at the premises that include power signal clips of the appliances, with the power signal clips of the model appliances;

calculating a difference between a total power draw of the appliances at a first time and the total power draw of the appliances at a preceding time; and

determining whether a state of a maximum of one of the appliances has changed, based on said difference;

c) obtain data relating to at least one behavioral pattern and a state of a user;

d) obtain data relating to external impacts on the user'"'"'s power usage within the premises, said external impacts originating from an environment of the user and not from the user;

e) use obtained data from at least one of a) to d) to generate an individual consumer forecast of power usage, the consumer forecast being collectively aggregated across a plurality of users;

f) use data from one or more of a) to e) to perform a demographic analysis;

g) collect aggregate power consumption data for a region associated with the macrogrid;

h) determine power consumption requirements across the macrogrid for the region using data from at least one of more of a) to g); and

i) determine future power requirements within the macrogrid, using the data obtained in at least one of a) to h).