This paper describes the detection method of the anomalies in signals proposed by Okabe et al. based on their KM_2O-Langevin framework; breakdown points of the stationarity can be detected to give the sign of anomalies. The method was applied to the sound signals to monitor the state of a generator engine and showed its feasibility.
FujimotoH. et al., Trouble Detectin for NIC Machine Using Sound, JSME Dynamics and Design Conference 2000, CD-ROM Proc, 2000, 447, (in Japanese).
2.
TsuchiyaM. and TakagiM., Vibration and Acoustics Fault Diagnosis of Rotational Machine, Trans. JSME C64(614) (1998), 85-02 (in Japanese).
3.
LinJ. et al., Feature extraction based on Morlet wavelet and its application for mechanical fault diagnosis, Journal of Sound and Vibration234(1) (2000), 135-148.
4.
WuJ.-D. et al., Continuous wavelet transform technique for fault signal siagnosis of internal combustion engines, NDT& E INternational39 (2006), 304-311.
5.
ZhangZ. et al., Parasitic Discrete Wavelet Transform and its Problems, Proc. MEXT & OKU 2011 Workshop on Wavelet Theory and Its Applications to Engineering, 2011 (in Japanese).
6.
IdeT., Change-point Detection and Failure Analysis of Sensor Data Using Sparse Structure Learning, ProVISION 65 (2010), Spring, 71-76 (in Japanese).
7.
ArakiH. et al., Abnormality Detection in Multi-Channel Time-Series Data Using Higher-Order Local Autocorrelations, IEEJ Trans. EIS. 129 (2009), no.7 (in Japanese).
8.
OkabeY. et al., Fluctuation-Dissipation Principle in Timeseries Analysis and Experimantal Mathematics, Nippon Hyouron Co., 2002 (in Japanese).
9.
OkabeY., Experimental Mathematics, Asakura Shoten, 2005 (in Japanese).
10.
HondaS. and YonekuraH., Acoustic Diagnosis of Generator Engines with Time Series Analysis Derived from the Fluctuation-Dissipation Principle', J. SICE51(9) (2012), 861-867 (in Japanese).
11.
HondaS. and YonekuraH., Acoustic Diagnosis of Generator Engines with Stationarity Test Derived from the Fluctuation-Dissipation Principle, Proc. SICE 2013.
12.
HondaS. and YonekuraH., `Anomaly Detection with Stationarity Test Derived from Fluctuatin-Dissipation Principle', Proc. SICE 2014. (2014).
