Apparatus state detector, method for detecting apparatus state, apparatus state detection server and apparatus state detection system; living persons' anomaly detector, living persons' anomaly detection system and method for detecting living persons' anomaly, and apparatus-state database maintenance server

• US 8,930,146 B2
• Filed: 03/18/2009
• Issued: 01/06/2015
• Est. Priority Date: 04/11/2008
• Status: Active Grant

• Current Assignee: Mitsubishi Electric Corporation
• Original Assignee: Mitsubishi Electric Corporation
• Inventors: Katsukura, Makoto, Nakata, Masanori, Kushiro, Noriyuki, Kuroiwa, Takeru, Higuma, Toshiyasu, Hibara, Naoyuki, Ito, Yoshiaki
• Primary Examiner(s): Breene, John
• Assistant Examiner(s): BLOSS, STEPHANIE E
• Application Number:
  US12/919,296
• Publication Number:
  US 20110010106A1
• Time in Patent Office: 2,120 Days
• Field of Search: 702/19, 340/573.1, 340/870.11, 340/3.43
• US Class Current: 702/19
• CPC Class Codes: G01D 4/002 Remote reading of utility meters
• CPC Class Codes: G01R 31/2825 in household appliances or professional audio/video equipment testing LAN's H04L43/50; testing TV systems H04N17/00; testing loudspeakers H04R29/00
• CPC Class Codes: G08B 21/0423 detecting deviation from an expected pattern of behaviour or schedule
• CPC Class Codes: G08B 21/0484 Arrangements monitoring consumption of a utility or use of an appliance which consumes a utility to detect unsafe condition, e.g. metering of water, gas or electricity, use of taps, toilet flush, gas stove or electric kettle
• 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

An apparatus state detector and its associates are provided to save time and effort for learning combinations of operation states of all apparatuses at home and commit fewer estimation errors even when an unknown apparatus starts operating. The apparatus state detector includes measuring means that measures a physical quantity of an environment in which an apparatus is placed, feature-quantity calculation means that calculates a feature quantity of the measured value measured by the measuring means, storage means that stores in advance the feature quantity of each apparatus and an apparatus state associated with the feature quantity in a reference-apparatus entry dictionary, and apparatus-state detection means that searches the reference-apparatus entry dictionary for a feature quantity by using a feature quantity calculated by the feature-quantity calculation means as a search key and detects an apparatus state based on the apparatus state associated with the retrieved feature quantity.

Claims

1. An apparatus state detector that detects states of a single or a plurality of apparatuses comprising:

measuring means that measures a physical quantity of an environment in which the apparatus is placed;

feature-quantity calculation means that calculates a feature quantity of the measured value measured by said measuring means;

storage means that stores, in advance, a plurality of feature quantities of each apparatus and apparatus states associated with the feature quantities, respectively, as dictionary data; and

apparatus-state detection means that searches for a feature quantity from said plurality of feature quantities in said dictionary data by using a feature quantity calculated by said feature-quantity calculation means as a search key and identifies an apparatus state based on the apparatus state associated with a retrieved feature quantity as a result of the search based on stored histories of the feature quantities for each apparatus,

wherein said apparatus-state detection means is configured to

calculate a difference between a first feature quantity, of said stored histories, which is calculated by said feature-quantity calculation means based on said measured value measured by said measuring means at a first time period and a second feature quantity, of said stored histories, which is calculated by said feature-quantity calculation means based on said measured value measured by said measuring means at a second time period,

identify the calculated difference between the first feature quantity and the second feature quantity as a third feature quantity,

using the third feature quantity as the search key, retrieve, from the dictionary data stored in the storage means, a stored feature quantity, from the stored plurality of feature quantities, that corresponds to said third feature quantity, and

identify, a state of the apparatus based on an apparatus state that is stored, in the storage means, in association with the retrieved feature quantity.

2. The apparatus state detector of claim 1, wherein

said measuring means measures a value of current supplied to the apparatus, a flow rate of water used by the apparatus, a flow rate of gas used by the apparatus, illuminance of the environment in which the apparatus is installed, a temperature of the environment in which the apparatus is installed, or an amount of data transfer across a communication network of the environment in which the apparatus is installed, as the physical quantity.

3. The apparatus state detector of claim 1, wherein

said feature-quantity calculation means takes a weighted average of said measured values with a predetermined window width to determine an average measurement value, and defines said average measurement value as said feature quantity.

4. The apparatus state detector of claim 3, wherein

said feature-quantity calculation means obtains a difference value between said average measurement value and said measured value before being averaged, defines the difference value as a differential measurement value, and defines said average measurement value and said differential measurement value as said feature quantity.

5. The apparatus state detector of claim 4, wherein

said feature-quantity calculation means takes a weighted average of said average measurement values with a predetermined window width to determine a second average measurement value,

further obtains a difference value between said second average measurement value and said average measurement value before being averaged and defines the difference value as a second differential measurement value, and

defines said average measurement value, said differential measurement value, said second average measurement value and said second differential measurement value as said feature quantities.

6. The apparatus state detector of claim 1, wherein

said feature-quantity calculation means performs predetermined calculations to extract periodicity of said measured values, and

defines a position at which the periodicity is extracted in the measured values and the intensity at the position for each extracted period as said feature quantity.

7. The apparatus state detector of claim 6, wherein

said feature-quantity calculation means extracts periodicity of said measured values by using Fourier transform or wavelet transform.

8. The apparatus state detector of claim 1, wherein

said apparatus-state detection means calculates a matching degree of the feature quantity calculated by said feature-quantity calculation means and the feature quantity in said dictionary data to identify a search result based on the matching degree.

9. The apparatus state detector of claim 8, wherein

said feature quantity is a multi-dimensional vector value, and

said apparatus-state detection means obtains a difference between the feature quantity calculated by said feature-quantity calculation means and the feature quantity in said dictionary data for every dimension and calculates said matching degree using the difference.

10. The apparatus state detector of claim 8, wherein

said feature quantity is a multi-dimensional vector value, and

said apparatus-state detection means obtains a difference between the feature quantity calculated by said feature-quantity calculation means and the feature quantity in said dictionary data for every dimension, counts the number of the dimensions whose difference is in a predetermined range, and defines a value obtained by dividing a counted result by the total number of dimensions as said matching degree.

11. The apparatus state detector of claim 8, wherein

said feature quantity is a multi-dimensional vector value, and

said apparatus-state detection means obtains an inner product of vectors of the feature quantity calculated by said feature-quantity calculation means and the feature quantity in said dictionary data, and defines a value obtained by dividing the inner product by a norm of each vector as said matching degree.

12. The apparatus state detector of claim 1, wherein

said apparatus-state detection means links said second feature quantity and an apparatus state associated with said feature quantity identified as a search result and stores them as new dictionary data.

13. The apparatus state detector of claim 1, wherein

said apparatus-state detection means judges that a point at which a long period component of said feature quantity varies a predetermined value or more is a point in time when the apparatus has started an operation.

14. The apparatus state detector of claim 1, wherein

said apparatus-state detection means stores the feature quantity that was used as a search key to search for a feature quantity in said dictionary data as feature quantity historical data in said dictionary data identified as a result of the search.

15. The apparatus state detector of claim 14, wherein

said apparatus-state detection means judges whether the apparatus is in a warning state based on said feature quantity historical data.

16. The apparatus state detector of claim 15, wherein

said apparatus-state detection means calculates either of, or both, an average and a variance of said feature quantity historical data and judges that the apparatus is in a warning state when either of, or both, the average and variance exceed a predetermined threshold uniquely assigned to each apparatus.

17. The apparatus state detector of claim 15, wherein

said apparatus-state detection means calculates and stores either of, or both, an average and a variance of said feature quantity historical data, and

when either of, or both, an average and a variance of said feature quantity historical data are newly calculated, said apparatus-state detection means obtains a difference between a stored value and a newly calculated value and judges that the apparatus is in a warning state when the difference exceeds a predetermined threshold uniquely assigned to each apparatus.

18. The apparatus state detector of claim 15, wherein

when an apparatus state is detected to be a warning state, said apparatus-state detection means judges that apparatus states prior to and subsequent to the apparatus state detected by said apparatus-state detection means within a predetermined period of time are warning states.

19. The apparatus state detector of claim 15, wherein

when an apparatus state is detected to be a warning state, said apparatus-state detection means judges that apparatus states prior to and subsequent to the apparatus state detected by said apparatus-state detection means within a predetermined period of time are pre-warning states, and judges that said apparatus states are warning states when the number of determinations of said pre-warning states reaches a predetermined number.

20. The apparatus state detector of claim 1, wherein the first time period represents a time period prior to operation of the apparatus and the second time period represents a time period subsequent to operation of the apparatus.