Water-leak state estimation system, method, and recording medium

Patent metadata

Field	Value
Publication number	US-10228301-B2
Application number	US-201615573039-A
Country	US
Kind code	B2
Filing date	Mar 10, 2016
Priority date	May 13, 2015
Publication date	Mar 12, 2019
Grant date	Mar 12, 2019

Abstract

This invention provides a water-leakage state estimation system configured to estimate a state of a water leakage in a specific area of a water distribution network. A learning unit is configured to: receive labeled data, which is labeled so as to separate past flow rate data into abnormal values and normal values, and past environment state condition data; build a prediction model for predicting the normal values in the labeled data through learning; and determine a score parameter defining a length of a period involving data to be verified through learning as well. A water-leakage estimation unit is configured to: compare predicted flow rate data obtained by supplying current environment condition data into the prediction model and current flow rate data to produce error values; and calculate an average value of the error values in the period of a window width defined by the score parameter to estimate a water-leakage score representing a state of the water-leakage in the specific area.

First claim

The invention claimed is:

1. A water-leakage state estimation system, which is configured to estimate a state of a water leakage in a specific area of a water distribution network, the water-leakage state estimation system comprising:
a learning circuitry configured to:
receive labeled data, which is labeled so as to separate past flow rate data representing a water use amount measured in the past in the specific area into abnormal values and normal values, and past environment condition data representing environment conditions at measurement time instants of the past flow rate data;
build, through learning, a prediction model for predicting the normal values in the labeled data in response to the past environment condition data; and
determine, through learning as well, a first score parameter required for using the prediction model to determine a water leakage, the first score parameter indicating a window width defining a length of a period involving data to be verified; and

a water-leakage estimation circuitry configured to:
compare predicted flow rate data obtained by supplying current environment condition data representing current environment conditions into the prediction model and current flow rate data representing a water use amount currently measured in the specific area with each other to produce error values; and
calculate an average value of the error values in the period of the window width defined by the first score parameter to estimate a water-leakage score representing a state of the water-leakage in the specific area.

2. The water-leakage state estimation system according to claim 1 , wherein the learning circuitry is configured to select an optimal first score parameter out of a plurality of candidate first score parameters as the first score parameter and to acquire the optimal first score parameter.

3. The water-leakage state estimation system according to claim 2 , wherein:
the learning circuitry is configured to determine, through learning as well, a second score parameter required for using the prediction model to determine a water leakage, the second score parameter indicating a threshold defining a magnitude of a deviation exhibited between the predicted flow rate data and the current flow rate data when the state is considered as a water leakage; and
the water-leakage estimation circuitry is configured to compare the error values and the threshold defined by the second score parameter for binarization to obtain hinge transform error values, to thereby estimate the water-leakage score representing the state of the water leakage in the specific area.

4. The water-leakage state estimation system according to claim 3 , wherein the learning circuitry is configured to select an optimal second score parameter out of a plurality of candidate second score parameters as the second score parameter and to acquire the optimal second score parameter.

5. The water-leakage state estimation system according to claim 1 , further comprising a labeling circuitry configured to add a water-leakage label for the separation between the abnormal values and the normal values to the past flow rate data, to thereby generate a combination of water-leakage information representing the water-leakage label and the past flow rate data as the labeled data.

6. The water-leakage state estimation system according to claim 1 , wherein:
the past environment condition data includes past calendar information and past weather information; and
the current environment condition data includes current calendar information and current weather information.

7. The water-leakage state estimation system according to claim 1 , wherein the learning comprises heterogeneous mixture leaning.

8. The water-leakage state estimation system according to claim 3 , further comprising an output device configured to visualize and output the water-leakage score together with values in a time series of the predicted flow rate data and values in a time series of the current flow rate data so that a relationship between the first score parameter and the second score parameter is recognizable.

9. A method of estimating a state of a water leakage in a specific area of a water distribution network through use of a water-leakage state estimation system, the method comprising:
receiving labeled data, which is labeled so as to separate past flow rate data representing a water use amount measured in the past in the specific area into abnormal values and normal values, and past environment condition data representing environment conditions at measurement time instants of the past flow rate data;
building, through learning, a prediction model for predicting the normal values in the labeled data in response to the past environment condition data;
determining, through learning as well, a first score parameter required for using the prediction model to determine a water leakage, the first score parameter indicating a window width defining a length of a period involving data to be verified;
comparing predicted flow rate data obtained by supplying current environment condition data representing current environment conditions into the prediction model and current flow rate data representing a water use amount currently measured in the specific area with each other to produce error values; and
calculating an average value of the error values in the period of the window width defined by the first score parameter to estimate a water-leakage score representing a state of the water-leakage in the specific area.

10. A non-transitory computer-readable recording medium having recorded thereon a water-leakage state estimation program for causing a computer to estimate a state of a water leakage in a specific area of a water distribution network, the water-leakage state estimation program causing the computer to execute:
a learning procedure for:
receiving labeled data, which is labeled so as to separate past flow rate data representing a water use amount measured in the past in the specific area into abnormal values and normal values and past environment condition data representing environment conditions at measurement time instants of the past flow rate data;
building, through learning, a prediction model for predicting the normal values in the labeled data in response to the past environment condition data; and
determining, through learning as well, a first score parameter required for using the prediction model to determine a water leakage, the first score parameter indicating a window width defining a length of a period involving data to be verified; and

a water-leakage estimation procedure for:
comparing estimated flow rate data obtained by supplying current environment condition data representing current environment conditions into the prediction model and current flow rate data representing a water use amount currently measured in the specific area with each other to produce error values; and
calculating an average value of the error values in the period of the window width defined by the first score parameter to estimate a water-leakage score representing a state of the water-leakage in the specific area.

Assignees

• Nec Corp

Inventors

• Kusumura Yukitaka
• TARASENKO Sergey
• Eto Riki
• Muraoka Yusuke
• Fujimaki Ryohei

Classifications

• G01M3/28 Primary

  for pipes, cables or tubes; for pipe joints or seals; for valves {; for welds} · CPC title