IP Library Granted Patent US 11,921,137
Granted Patent B2
US 11,921,137 · App. 17/651,306 · Granted Mar 5, 2024

Anomaly detection apparatus, method, and program

Inventors: Tenta Sasaya (Tokyo, JP); Takashi Watanabe (Yokohama, JP); Toshiyuki Ono (Kawasaki, JP)
Assignees: KABUSHIKI KAISHA TOSHIBA; Toshiba Infrastructure Systems & Solutions Corporation
G01R23/16
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 11,921,137
App. No.
17/651,306
Granted
Mar 5, 2024
Kind
B2
Abstract

According to one embodiment, an anomaly detection apparatus includes a processing circuit configured to calculate a reconstruction error of an input signal being a time-series signal, calculate cycle information indicating cyclic property of the reconstruction error, and determine presence/absence of an anomaly signal in the input signal on the basis of the cycle information or the cycle information and the reconstruction error.

Claims (25)

1. An anomaly detection apparatus comprising at least one processing circuit configured to:

calculate a reconstruction error of an input signal being a time-series signal,

calculate cycle information indicating cyclic property of the reconstruction error, and

determine presence/absence of an anomaly signal in the input signal on the basis of the cycle information or the cycle information and the reconstruction error.

2. The anomaly detection apparatus according to claim 1 , wherein the at least one processing circuit is configured to acquire a reference signal synchronized with a cycle of the input signal, and calculates similarity between the reconstruction error and the reference signal as the cycle information.

3. The anomaly detection apparatus according to claim 1 , wherein the cycle information is a feature quantity generated by executing Fourier transform for a vector generated by arranging values of the reconstruction error.

4. The anomaly detection apparatus according to claim 1 , wherein the cycle information is a degree of dispersion of two or more times selected on the basis of magnitude of values of the reconstruction error.

5. The anomaly detection apparatus according to claim 1 , wherein the at least one processing circuit is configured to;

divide the input signal into two or more short-time signals on the basis of reference information,

calculate the reconstruction error for each of the short-time signals, and

calculate, as the cycle information, a degree of dispersion of timings in the respective short-time signals for the reconstruction errors selected in the respective short-time signals.

6. The anomaly detection apparatus according to claim 5 , wherein

the input signal is a voltage value of AC voltage,

the reference information is a voltage value of the AC voltage or the number of phases of the AC voltage,

each of the short-time signals includes the input signal of a phase for one cycle of the AC voltage, and

each of the timings in the respective short-time signals indicates a phase of the AC current.

7. The anomaly detection apparatus according to claim 1 , wherein the at least one processing circuit is configured to calculate a smoothed signal on the basis of the reconstruction error, and corrects the reconstruction error by subtracting the smoothed signal from the reconstruction error.

8. The anomaly detection apparatus according to claim 1 , wherein the at least one processing circuit is configured to execute weighting for the reconstruction error on the basis of the cycle information, and determine presence/absence of the anomaly signal on the basis of the cycle information and the weighted reconstruction error.

9. A method, implemented by at least one processing circuit of an anomaly detection apparatus, comprising:

calculating a reconstruction error of an input signal being a time-series signal;

calculating cycle information indicating cyclic property of the reconstruction error; and

determining presence/absence of an anomaly signal in the input signal on the basis of the cycle information or the cycle information and the reconstruction error.

10. A non-transitory computer-readable storage medium storing a program for causing a computer to execute:

a function of calculating cycle information indicating cyclic property of the reconstruction error; and

a function of determining presence/absence of an anomaly signal in the input signal on the basis of the cycle information or the cycle information and the reconstruction error.

Assignments (2)
MERGER Recorded Jul 29, 2025
From: TOSHIBA INFRASTRUCTURE SYSTEMS & SOLUTIONS CORPORATION
To: KABUSHIKI KAISHA TOSHIBA
Reel/Frame 072239/0263 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Feb 16, 2022
From: SASAYA, TENTA; WATANABE, TAKASHI; ONO, TOSHIYUKI
To: KABUSHIKI KAISHA TOSHIBA; TOSHIBA INFRASTRUCTURE SYSTEMS & SOLUTIONS CORPORATION
Reel/Frame 059026/0359 →
Priority Claims (1)
JP 2021-150373 · Sep 15, 2021 · national
Continuity (1)
Related Publication 20230079397A1 · Mar 16, 2023