Methods and systems for identification and correction of controlled system data

• US 10,002,184 B2
• Filed: 12/08/2014
• Issued: 06/19/2018
• Est. Priority Date: 12/08/2013
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google LLC (Alphabet Inc.)
• Inventors: Matsuoka, Yoky
• Primary Examiner(s): Shmaton, Alexey
• Assistant Examiner(s): Lee, Tsu-Chang
• Application Number:
  US14/563,745
• Publication Number:
  US 20150161515A1
• Time in Patent Office: 1,289 Days
• CPC Class Codes: G05B 13/027 using neural networks only
• CPC Class Codes: G05B 15/02 electric
• CPC Class Codes: G05B 2219/2614 HVAC, heating, ventillation, climate control
• CPC Class Codes: G05B 2219/2642 Domotique, domestic, home control, automation, smart house
• CPC Class Codes: G06F 16/3334 Selection or weighting of terms from queries, including natural language queries
• CPC Class Codes: G06N 3/084 Backpropagation, e.g. using gradient descent

Abstract

Computational methods and systems that collect operational data from an intelligent controller to identify information, or correct information, about a device and system controlled by the intelligent controller are disclosed. Computational methods and systems use a set of operational data and information known about other devices and systems controlled by similar intelligent controllers to process the operational data and generate information, or correct information, about the device and system.

Claims

1. A computer system comprising:

one or more processors;

one or more data-storage devices; and

machine-readable instructions stored in one or more of the one or more data-storage devices that, when executed by the one or more processors, cause the one or more processors to perform operations comprising;

receiving operational data generated by a thermostat;

receiving registration data associated with the thermostat from the one or more data-storage devices;

examining the registration data for defects;

when the registration data includes defects;

generating approximate registration data based on the operational data and a set of operational data and registration data collected from one or more similar thermostats, and

correcting the defects in the registration data with corresponding approximate registration data;

generating commands using the registration data; and

sending the commands to the thermostat to control a heating, ventilation, and air conditioning (HVAC) system.

2. The computer system of claim 1, wherein examining the registration data for defects further comprises:

examining the registration data for inconsistences between the registration data and the operational data; and

examining the registration data for incomplete registration data.

3. The computer system of claim 2, wherein examining the registration data for inconsistencies further comprises:

generating approximate operational data based on the registration data and the set of operational data and registration data collected from the number of similar thermostats; and

comparing elements of the operational data with elements of the approximate operational data.

4. The computer system of claim 3, wherein generating the approximate operational data based on the registration data and the set of operational data and registration data collected from the one or more similar thermostats further comprises:

calculating weights that relate the set of operational data and registration data collected from the one or more similar thermostats using a feed-forward neural network with back propagation; and

calculating approximate operational data from the registration data and the weights.

5. The computer system of claim 1, wherein correcting the defects in the registration data with corresponding approximate registration data further comprises:

when the registration data is inconsistent with the operational data, replacing inconsistent registration data elements with corresponding generated registration data elements; and

when the registration data is incomplete, filling in missing registration data elements with corresponding approximate registration data elements.

6. The computer system of claim 1, wherein generating the approximate registration data based on the operational data and the set of operational data and registration data collected from the one or more similar thermostats further comprises:

calculating weights that relate the set of operational data and registration data collected from the one or more similar thermostats using a feed-forward neural network with back propagation; and

calculating approximate registration data from the operational data and the weights.

7. The computer system of claim 1, wherein the operational data further comprises data collected from thermostat sensors, control schedules, and data generated by the thermostat.

8. The computer system of claim 1, wherein the operations further comprise, when the registration data does not include defects, adding the operational data and registration data associated with the thermostat to the set of operational data and registration data collected from the one or more similar thermostats.

9. A method to be carried out by a computer system that includes one or more processors and one or more data-storage devices, the method comprising:

receiving operational data and registration data associated with a thermostat;

examining the registration data for defects;

when the registration data includes defects;

generating approximate registration data based on the operational data and a set of operational data and registration data collected from a number of similar thermostats, and

correcting the defects in the registration data with corresponding approximate registration data;

generating commands using the registration data; and

sending the commands to the thermostat to control a heating, ventilation, and air conditioning (HVAC) system.

10. The method of claim 9, wherein examining the registration data for defects further comprises:

examining the registration data for inconsistences between the registration data and the operational data; and

examining the registration data for incomplete registration data.

11. The method of claim 10, wherein examining the registration data for inconsistencies further comprises:

generating approximate operational data based on the registration data and the set of operational data and registration data collected from a number of similar thermostats; and

comparing elements of the operational data with elements of the approximate operational data.

12. The method of claim 11, wherein generating the approximate operational data based on the registration data and the set of operational data and registration data collected from the one or more similar thermostats further comprises:

calculating weights that relate the set of operational data and registration data collected from the one or more similar thermostats using a feed-forward neural network with back propagation; and

calculating approximate operational data from the registration data and the weights.

13. The method of claim 9, wherein correcting the defects in the registration data with the corresponding approximate registration data further comprises:

when the registration data is inconsistent with the operational data, replacing inconsistent registration data elements with corresponding generated registration data elements; and

when the registration data is incomplete, filling in missing registration data elements with corresponding approximate registration data elements.

14. The method of claim 9, wherein generating the approximate registration data based on the operational data and the set of operational data and registration data collected from the one or more similar thermostats further comprises:

calculating weights that relate the set of operational data and registration data collected from the one or more similar thermostats using a feed-forward neural network with back propagation; and

calculating approximate registration data from the operational data and the weights.

15. The method of claim 9, wherein the operational data further comprises data collected from thermostat sensors, control schedules, and data generated by the thermostat.

16. A non-transitory computer-readable medium having machine-readable instructions encoded thereon for enabling one or more processors to perform operations comprising:

receiving operational data and registration data from one or more memories, the operational data and registration data associated with a thermostat;

examining the registration data for defects;

when the registration data includes defects;

generating approximate registration data based on the operational data and a set of operational data and registration data collected from a number of similar thermostats, and

correcting the defects in the registration data with corresponding approximate registration data;

generating commands using the registration data; and

sending the commands to the thermostat to control a heating, ventilation, and air conditioning (HVAC) system.

17. The non-transitory computer-readable medium of claim 16, wherein generating the approximate registration data based on the operational data and the set of operational data and registration data collected from the one or more similar thermostats comprises:

calculating weights that relate the set of operational data and registration data collected from the one or more similar thermostats using a feed-forward neural network with back propagation; and

calculating approximate registration data from the operational data and the weights.

18. The non-transitory computer-readable medium of claim 16, wherein correcting the defects in the registration data with corresponding approximate registration data comprises:

when the registration data is inconsistent with the operational data, replacing inconsistent registration data elements with corresponding generated registration data elements; and

when the registration data is incomplete, filling in missing registration data elements with corresponding approximate registration data elements.