Sage Journals: Discover world-class research

Abstract

A study was undertaken to assess the feasibility of identifying the location of a point of excitation acting on an elastic structure based on information obtained from remote transducers. This process was deemed significant as the technique would have applications in identifying machinery in a large, complex plant that was producing spurious forcing conditions, for example out-of-balance or shaft misalignment. A pattern recognition strategy was adopted using the Kohonen network. The Kohonen array was subsequently labelled using both multi-layer perceptrons and Gaussian labelling. The former was found to be computationally inefficient and hence the latter is implemented. It was found that using the system described it is possible to identify a single frequency source of vibration.

Get full access to this article

View all access options for this article.

References

Burrows

J. H.

Optimizing the use of periodic data for machine condition monitoring. In Proceedings of the Third International Conference COMADEM, Southampton, 1991 (Institute for Sound and Vibration Research, Southampton).

Tan

C. C.

Machining signature analysis with adaptive signal processing. Proceedings of the Third International Symposium on Signal processing and its applications, Australia, 1992, pp. 533–536.

Archambeault

Burrows

J. H.

The use of statistical profiles and spectral enveloping techniques in on-condition monitoring programs. Proceedings of Second IMMDC, Los Angeles, Calif., 1991.

Frietmann

E. E. E.

Kerckhoffs

E. J. H.

Wedman

F. W.

Guarding vibration patterns of a mechanical assembly with the help of a neural network. Faculty of Technical Mathematics and Informatics, Delft University of Technology, The Netherlands.

Ory

Glaser

Holdzeppe

The reconstruction of forcing based on measured structural responses. Proceedings of the Second International Symposium on Aeroelasticity and structural dynamics, Aachen, 1985.

Ory

Glaser

Holdzeppe

Quality of modal analysis and reconstruction of forcing functions based on measured output data. Proceedings of the Fourth IMAC, 1986, pp. 850–857.

Zhu

A time domain method for identifying dynamic loads on continuous systems. J. Sound Vibr., 1991, 148(1), 137–146.

Kohonen

Self-organised formation of topologically correct feature maps. Biological Cybernetics, 1982, 43, 59–69.

McClelland

J. L.

Rumelhart

D. E.

Parallel distributed processing, Vol. 1, 1986 (MIT Bradford Press).

10.

Widrow

Lehr

M. A.

Thirty years of the adaptive neural network; perceptron, madaline and back propagation. Proc. IEEE, 1990, 78(9), 1415–1443.

11.

Sprecher

D. A.

On the structure of continuous functions of several variables. Trans. Am. Math. Soc., 1965, 115, 340–355.

12.

Schaeffer

H. G.

MSC/NASTRAN primer, 1979 (Wallace Press).

13.

Cooley

J. W.

Lewis

P. A. W.

Wech

P. D.

The fast Fourier transform and its applications. IEEE Trans. on Educ., 1969, 12(1), 27–34.

14.

Pascoe

S. K.

Mottershead

J. E.

Two new finite element contact algorithms. Comput. Struct., 1989, 32(1), 137–144.

15.

Beale

Jackson

Neural computing (Adam Hilger).

16.

Schalkoff

Pattern recognition (John Wiley, New York).

A Technique for the Calculation of the Location of a Vibration Source

Abstract

Get full access to this article

References