Damage detection using outlier analysis

被引：499

作者：

Worden, K

Manson, G

Fieller, NRJ

机构：

[1] Univ Sheffield, Dept Mech Engn, Sheffield S1 3JD, S Yorkshire, England

[2] Univ Sheffield, Sch Math & Stat, Sheffield S1 3JD, S Yorkshire, England

来源：

JOURNAL OF SOUND AND VIBRATION | 2000年 / 229卷 / 03期

关键词：

D O I：

10.1006/jsvi.1999.2514

中图分类号：

O42 [声学];

学科分类号：

070206 ; 082403 ;

摘要：

This paper constitutes a study of a statistical method for damage detection. The lowest level of fault detection is considered so that the methods are simply required to signal deviations from normal condition; i.e., the problem is one of novelty detection. In this paper, the concept of discordancy from the statistical discipline of outlier analysis is used to signal deviance from the norm. The method is demonstrated on four case studies of engineering interest: one simulation, two pseudo-experimental and one experimental. (C) 2000 Academic Press.

引用

页码：647 / 667

页数：21

共 10 条

[1]

Barnett V., 1984, Outliers in Statistical Data, V2nd

[2] NOVELTY DETECTION AND NEURAL-NETWORK VALIDATION [J].

BISHOP, CM .

IEE PROCEEDINGS-VISION IMAGE AND SIGNAL PROCESSING, 1994, 141 (04) :217-222

[3]

Cempel C., 1991, Vibroacoustic Condition Monitoring

[4]

CHEN Q, 1994, P INT C VIBR ENG BEI, P567

[5]

GORDOR BK, 1994, THEISS U SHEFFIELD

[6]

JOHNSON N, 1995, 1551163 VICK SHIPB E

[7] Classification of faults in gearboxes - Pre-processing algorithms and neural networks [J].

Staszewski, WJ ;

Worden, K .

NEURAL COMPUTING & APPLICATIONS, 1997, 5 (03) :160-183

[8] Wavelet signal processing for enhanced Lamb-wave defect detection in composite plates using optical fiber detection [J].

Staszewski, WJ ;

Pierce, SG ;

Worden, K ;

Philp, WR ;

Tomlinson, GR ;

Cutshaw, B .

OPTICAL ENGINEERING, 1997, 36 (07) :1877-1888

[9]

TARRASENKO L, 1995, P 4 IEE INT C ART NE, V409, P442

[10]

WORDEN K, 1997, P 15 IMAC ORL FLOR, P631

← 1 →