Systems and methods for improving the accuracy of appliance level disaggregation in non-intrusive appliance load monitoring techniques

• US 9,612,286 B2
• Filed: 02/06/2012
• Issued: 04/04/2017
• Est. Priority Date: 02/04/2011
• Status: Active Grant

• Current Assignee: Bidgely, Inc.
• Original Assignee: Bidgely, Inc.
• Inventors: Rodney, Uri, Bapat, Subodh, Gupta, Abhay, Garud, Vivek
• Primary Examiner(s): VON BUHR, MARIA N
• Application Number:
  US13/366,850
• Publication Number:
  US 20130110621A1
• Time in Patent Office: 1,884 Days
• Field of Search: 700286, 700291, 702 60, 702 61, 702179-185, 702188, 705412
• CPC Class Codes: G01D 2204/24 Identification of individual loads, e.g. by analysing current/voltage waveforms
• CPC Class Codes: G01D 4/004 Remote reading of utility meters to a fixed location
• CPC Class Codes: G01R 31/34 Testing dynamo-electric machines
• CPC Class Codes: G06F 15/00 Digital computers in general details G06F1/00 – G06F13/00; Data processing equipment in general
• CPC Class Codes: G06Q 30/0254 based on statistics
• CPC Class Codes: Y02B 70/30 Systems integrating technologies related to power network operation and communication or information technologies for improving the carbon footprint of the management of residential or tertiary loads, i.e. smart grids as climate change mitigation technology in the buildings sector, including also the last stages of power distribution and the control, monitoring or operating management systems at local level
• CPC Class Codes: Y02B 90/20 Smart grids as enabling technology in buildings sector smart grids supporting the management or operation of end-user stationary applications in general, or like technologies with no associated climate change mitigation effect Y04S20/00
• CPC Class Codes: Y04S 20/242 Home appliances
• CPC Class Codes: Y04S 20/30 Smart metering, e.g. specially adapted for remote reading

Abstract

The present invention generally teaches systems and methods for creating appliance signatures based upon whole house composite load profiles. Methods may includes steps such as identifying primitive elements including transients and absolute steady state levels; clustering the primitive elements along multiple dimensions to form impulses; combining impulses to form simple bundles; combining simple bundles with each other or impulses to form complex bundles; and determining specific appliance signatures that substantially match the complex bundles. Methods may also include steps such as determining transitions within the whole house composite load profile; determining household specific appliance state machines for each appliance in the household; and disaggregating the whole house composite load profile into individual appliance energy loads by assigning the determined transitions to the determined household specific appliance state machines.

Claims

1. A system for detecting individual appliance energy loads from a whole house composite load profile, comprising:

a transitions generator, that receives the whole house composite load profile and identifies transitions within the whole house composite load profile;

a detector in selective communication with the transitions generator, that receives transitions from the transitions generator and determines a household specific appliance state machine for each appliance in the household, the household specific appliance state machine comprising transitions associated with different functions, if any, of the appliance, thereby modeling an inner mechanism of the appliance; and

a solver in selective communication with the detector, which receives as input all of the transitions identified by the transitions generator and the household specific state machines from the detector, and parses all transitions received from the detector as a whole, thereby assigning each transition to a household specific appliance state machine, and disaggregating the whole house composite load profile into individual appliance energy loads.

2. The system of claim 1, further comprising a transitions broker in selective communication with the solver, the transitions broker configured to assign transitions to the determined household specific state machine, in situations where the transition may apply to multiple household specific state machines.

3. The system of claim 2, wherein the transitions broker is in selective communication with a repository of non-electrical information, which is used to assist in properly assigning transitions to the determined household specific state machines.

4. The system of claim 3, wherein the repository of non-electrical information comprises one or more types of information selected from the group consisting of:

third party data, heuristics, and user input.

5. The system of claim 4, wherein the third party data comprise information selected from the group consisting of:

household square footage;

household occupancy, temperature data, heating degree days, cooling degree days, neighborhood data, and municipality data.

6. The system of claim 4, wherein the user input is received from any user of the system in identifying an appliance determined to correspond to specific energy usage.

7. A method for creating an appliance signature based upon a whole house composite load profile, comprising:

identifying primitive elements comprising transients and absolute steady state levels;

combining primitive elements to form impulses, using multi-dimensional clustering to group primitive elements along multiple dimensions;

combining impulses to form simple bundles;

combining simple bundles with other simple bundles or with impulses to form complex bundles, the complex bundles combining simple bundles associated with different functions of an appliance;

determining, based upon the complex bundles, specific appliance signatures that substantially match the complex bundles, and identifying the specific appliance signatures from the whole house composite load profile.

8. The method of claim 7, wherein the multi-dimensional clustering clusters primitive elements according to one or more dimensions selected from the group consisting of:

amplitude, periodicity, frequency, pulse width of impulses, on-time of impulses, rising edge slope, falling edge slope, time of day, time of week, and time of month.

9. A method for detecting individual appliance energy loads from a whole house composite load profile, comprising:

determining transitions within the whole house composite load profile;

determining household specific appliance state machines for each appliance in the household, the household specific appliance state machine comprising transitions associated with different functions, if any, of the appliance, the determined transitions being parsed together in order to model an inner mechanism of the appliance;

disaggregating the whole house composite load profile into individual appliance energy loads by assigning each of the determined transitions to the determined household specific appliance state machines.

10. The method of claim 9, wherein the transitions are determined within the whole house composite load profile by:

identifying minimum and maximum power levels within an identified temporal window;

comparing the minimum and maximum power levels with the previous or last known power levels; and

identifying a transition if the difference between the minimum and maximum power levels and the previous or last know power levels is greater than an allowable noise margin, wherein the amplitude of the transition is the amplitude of the change in power levels.

11. The method of claim 9, wherein the transitions are assigned based in part upon communication with a repository of non-electrical information used to assist in properly assigning transitions to the determined household specific state machines.

12. The method of claim 11, wherein the repository of non-electrical information comprises one or more types of information selected from the group consisting of:

third party data, heuristics, and user input.

13. The method of claim 12, wherein the user input is received from a user in identifying the determined appliance that corresponds to specific energy usage.

14. The method of claim 13, wherein the user input comprises a user, when presented with information regarding appliances matched to particular energy usages, confirming or negating such match.

15. The method of claim 13, wherein the user input comprises a user, when presented with information regarding particular energy usage, identifying an appliance that caused the particular energy usage.

16. The method of claim 15, wherein the user selects the appliance that caused the particular energy usage from a set of potential appliances presented to the user.

17. The method of claim 9, wherein the step of disaggregating the whole house composite load profile by assigning the determined transitions to the determined household specific appliance state machines comprises assigning the determined transitions based upon one or more rules selected from the group consisting of:

(i) assigning transitions to only one state machine;

(ii) assigning transitions active state machines over non-active state machines; and

(iii) assigning transitions to more complex state machines over less complex state machines.

18. A method of incentivizing users to replace inefficient appliances with efficient appliances based upon energy usage data measured from the user'"'"'s specific inefficient appliances, comprising:

determining an appliance load signature from the user'"'"'s specific appliance, determined from a whole house composite load signature, wherein such determination is based at least in part on;

determining transitions within the whole house composite load profile;

determining household specific appliance state machines for each appliance in the household, the household specific appliance state machine comprising transitions associated with different functions, if any, of the appliance, the determined transitions being parsed together in order to model an inner mechanism of the appliance;

disaggregating the whole house composite load profile into individual appliance energy loads by assigning each of the determined transitions to the determined household specific appliance state machines;

determining an appliance signature from an alternative appliance;

presenting to the user the difference in energy usage and associated costs between the user'"'"'s specific appliance and the alternative appliance.

19. The method of claim 18, further comprising:

generating a personalized incentive for the user based at least in part on the difference in energy usage and associated costs between the user'"'"'s specific appliance and the alternative appliance.

20. The method of claim 18, further comprising:

presenting to the user a purchase price for the alternative appliance and the time required for the user to see a return on investment if the alternative appliance is purchased.

21. The method of claim 20, further comprising:

identifying potential funding sources for the user to obtain credit or other funding in order to purchase the alternative appliance.