Patent US10804702

Self-organizing demand-response system
[+] Show Details

US10804702

1. A method of controlling an energy device, said method comprising: (8) (6)

receiving at each energy device in a pool of energy devices a service context from a central entity that commands each energy device to provide said service to a grid operator;

applying a heuristic at a first device agent executing on a first computer representing a first energy device in order to identify a second energy device in said pool;

receiving at said first computer a power model of said second energy device identified by said heuristic, wherein a power model of said first energy device exists on said first computer;

performing an optimization on said first computer to calculate a value of said first and second energy devices working together to provide said service;

forming a cluster of said first and second energy devices; and

outputting a first control action from said first device agent to provide a first power level to said first energy device that partially depends upon said power model of said second energy device and that provides said service to said grid operator.

2. A method as recited in claim 1 wherein said first and second energy devices are thermal energy devices. (0)

3. A method as recited in claim 1 wherein said first and second energy devices are electrical energy devices. (0)

4. A method as recited in claim 1 wherein said performing an optimization includes performing optimizations in order to rank said energy devices of said pool in terms of the value to provide said service when an energy device is combined with said first energy device, and selecting said second energy device from said ranked energy devices. (1) (0)

6. A method as recited in claim 1 further comprising: (0) (1)

7. A method as recited in claim 1 wherein said first and second energy devices are peers in said pool of energy devices. (0)

8. A method as recited in claim 1 wherein said first computer is local to said first energy device. (0)

9. A method as recited in claim 1 further comprising: (0) (1)

10. A method as recited in claim 1 further comprising: (1) (3)

executing a second device agent on a second computer representing said second energy device;

receiving at said second computer a power model of said first energy device;

outputting a second control action from said second device agent to provide a second power level to said second energy device that partially depends upon said power model of said first energy device and that provides said service to said grid operator.

11. A method as recited in claim 10 further comprising: (0) (2)

12. A system for controlling an energy device, said system comprising: (0) (3)

a central entity computer;

a pool of energy devices, each of said devices receiving a service context from said central entity computer that commands each energy device to provide said service to a grid operator, said pool including a first energy device and a second energy device;

a first device agent executing on a first computer representing said first energy device arranged to apply a heuristic in order to identify said second energy device in said pool, wherein a power model of said first energy device exists on said first computer, said device agent being further arranged to receive a power model of said second energy device identified by said heuristic, to perform an optimization on said first computer to calculate a value of said first and second energy devices working together to provide said service, and to output a first control action from said first device agent to provide a first power level to said first energy device that partially depends upon said power model of said second energy device and that provides said service to said grid operator.

13. A method of controlling an energy device in a self-reliance mode, said method comprising: (7) (4)

applying a heuristic at a first device agent executing on a first computer representing a first energy device in order to identify a second energy device in a pool of energy devices that provides energy complementary to said first energy device;

forming a cluster of said first and second energy devices in which a net energy to or from said cluster is minimized;

receiving at said first computer a power model of said second energy device, wherein a power model of said first energy device exists on said first computer; and

outputting a first control action from said first device agent to provide a power level to said first energy device that partially depends upon said power model of said second energy device.

14. A method as recited in claim 13 wherein said first and second energy devices are thermal energy devices. (0)

15. A method as recited in claim 13 wherein said first and second energy devices are electrical energy devices. (0)

16. A method as recited in claim 13 wherein said forming a cluster includes performing optimizations in order to rank said energy devices of said pool in terms of the value to provide said minimized net energy when an energy device is combined with said first energy device, and selecting said second energy device from said ranked energy devices. (1) (0)

20. A method as recited in claim 13 further comprising: (1) (2)

executing a second device agent on a second computer representing said second energy device;

receiving at said second computer a power model of said first energy device;

24. A method as recited in claim 20 further comprising: (0) (2)

21. A method as recited in claim 13 wherein said first and second energy devices are peers in said pool of energy devices. (0)

22. A method as recited in claim 13 wherein said first computer is local to said first energy device. (0)

23. A method as recited in claim 13 wherein said heuristic is a proximity heuristic and wherein said second energy device is the closest energy device of said pool to said first energy device that provides said complementary energy. (0)

25. A system for controlling an energy device in a self-reliance mode, said system comprising: (0) (2)

26. A method of providing a service to a grid operator, said method comprising: (7) (5)

forming a plurality of clusters of energy devices out of a pool of energy devices, each cluster being formed by exchanging models between energy devices of said each cluster and optimizing said each cluster using said models to provide said service;

sending a request from a service manager to said each cluster for a bid that partially provides said service, said each bid being based upon a current multiplier;

comparing a total of all bids from said clusters to a bid from said service manger to provide said service;

determining that said total of all bids from said clusters is close enough to said service manager bid; and

each energy device of each of said clusters implementing a control policy based upon said optimizing of said each cluster, wherein said service is provided.

27. A method as recited in claim 26 wherein said energy devices are thermal energy devices. (0)

28. A method as recited in claim 26 wherein said energy devices are electrical energy devices. (0)

29. A method as recited in claim 26 further comprising: (0) (1)

30. A method as recited in claim 26 wherein each of said implemented control policies of said each cluster partially depends upon a power model of at least one other energy device in said each cluster. (0)

31. A method as recited in claim 26 wherein said multiplier represents a value desirable by said each cluster. (0)

32. A method as recited in claim 26 further comprising: (0) (1)

33. A method as recited in claim 26 wherein said clusters include at least one standalone energy device. (0)

34. A system for providing a service to a grid operator, said system comprising: (0) (2)

35. A method of providing a service to a grid operator using a distributed architecture, said method comprising: (6) (6)

executing a plurality of service manager processes, each service manager process representing one of said clusters and executing on a computer local to one of said clusters;

implementing a peer-to-peer consensus-seeking technique in order to iteratively request bids from said clusters using said service manager processes, wherein each of said bids partially providing said service, and wherein said each bid being based upon a current multiplier;

comparing a total of all bids from said clusters to a bid from one of said service manager processes that represents a service manger to provide said service;

determining that said total of all bids from said clusters is close enough to said service manager bid; and

each energy device of each of said clusters implementing a control policy based upon said optimizing of said each cluster, wherein said service is provided.

36. A method as recited in claim 35 wherein said energy devices are thermal energy devices. (0)

37. A method as recited in claim 35 wherein said energy devices are electrical energy devices. (0)

38. A method as recited in claim 35 further comprising: (0) (1)

39. A method as recited in claim 35 wherein each of said implemented control policies of said each cluster partially depends upon a power model of at least one other energy device in said each cluster. (0)

40. A method as recited in claim 35 wherein said multiplier represents a value desirable by said each cluster. (0)

42. A method as recited in claim 35 wherein said clusters include at least one standalone energy device. (0)

43. A system of providing a service to a grid operator using a distributed architecture, said system comprising: (0) (2)

a plurality of clusters of energy devices, each cluster being formed by exchanging models between energy devices of said each cluster and optimizing said each cluster using said models to provide said service, wherein each energy device of each of said clusters being arranged to implement a control policy based upon said optimizing of said each cluster, wherein said service is provided; and

a plurality of computers, each computer representing one of said clusters and executing a service manager process, wherein each service manager process being arranged to implement a peer-to-peer consensus-seeking technique in order to iteratively request and receive bids from other of said service manager processes to partially provide said service based upon a current multiplier, to compare a total of all of said bids from said clusters to a bid from one of said service manager processes that represents a service manger to provide said service, to determine that said total of all bids from said clusters is close to said service manager bid, and to direct said energy devices to implement said each control policy based upon said each bid from said clusters.

44. A method of controlling an energy device, said method comprising: (0) (6)

receiving at each energy device in a pool of energy devices a technical constraint of one of said devices or of a power grid from a central entity that commands each energy device to satisfy said technical constraint;

applying a heuristic at a first device agent executing on a first computer representing a first energy device in order to identify a second energy device in said pool;

receiving at said first computer a power model of said second energy device identified by said heuristic, wherein a power model of said first energy device exists on said first computer;

performing an optimization on said first computer to calculate the ability of said first and second energy devices working together to satisfy said technical constraint;

forming a cluster of said first and second energy devices; and

45. A system for controlling an energy device, said system comprising: (0) (3)

a central entity computer;

a pool of energy devices, each of said devices receiving a technical constraint from said central entity computer that commands each energy device to satisfy said technical constraint, said pool including a first energy device and a second energy device;

a first device agent executing on a first computer representing said first energy device arranged to apply a heuristic in order to identify said second energy device in said pool, wherein a power model of said first energy device exists on said first computer, said device agent being further arranged to receive a power model of said second energy device identified by said heuristic, to perform an optimization on said first computer to calculate the ability of said first and second energy devices working together to satisfy said technical constraint, and to output a first control action from said first device agent to provide a first power level to said first energy device that partially depends upon said power model of said second energy device and that satisfies said technical constraint.

View Abstract and Specification Size

PDF with Images and Document Face >

Full Text Publication >

Patent US10804702

Self-organizing demand-response system

Patent Matrix® Search