US 11,789,436 B2
Diagnosing device, diagnosing method, and program
Ichiro Nagano, Yokohama (JP); Mayumi Saito, Tokyo (JP); Keiji Eguchi, Tokyo (JP); and Kuniaki Aoyama, Takasago (JP)
Assigned to Mitsubishi Heavy Industries, Ltd., Tokyo (JP)
Appl. No. 17/781,612
Filed by MITSUBISHI HEAVY INDUSTRIES, LTD., Tokyo (JP)
PCT Filed Dec. 24, 2020, PCT No. PCT/JP2020/048494
§ 371(c)(1), (2) Date Jun. 1, 2022,
PCT Pub. No. WO2021/140942, PCT Pub. Date Jul. 15, 2021.
Claims priority of application No. 2020-000427 (JP), filed on Jan. 6, 2020.
Prior Publication US 2023/0004153 A1, Jan. 5, 2023
Int. Cl. G05B 23/02 (2006.01)
CPC G05B 23/0235 (2013.01) [G05B 23/0221 (2013.01); G05B 23/0275 (2013.01)] 7 Claims
OG exemplary drawing
 
1. A diagnosing device for detecting an anomaly in a gas turbine, the device comprising:
a Mahalanobis distance calculating unit that calculates a Mahalanobis distance (MD value) of a detected value received from a measurement equipment provided at the gas turbine; and
an anomaly determination unit that determines a presence or absence of an anomaly based on the MD value, wherein
the anomaly determination unit determines the presence or absence of an anomaly of the gas turbine in a manner that a determination that there is no anomaly is more likely to occur when the number of samples per unit space is smaller than when the number of samples per unit space is large, wherein
the anomaly determination unit is configured to correct the MD value to a second MD value, based on a cumulative probability in a t-distribution having a degree of freedom according to the number of samples and a cumulative probability in a normal distribution corresponding to the cumulative probability in the t-distribution, and determine the presence or absence of an anomaly, based on a result of comparing the second MD value with a predetermined threshold value.