MINIMISATION OF DECISION ERRORS IN A PROBABILISTIC NEURAL NETWORK FOR CHANGE POINT DETECTION IN MECHANICAL SYSTEMS

Abstract Probabilistic neural nets have been applied in the detection of structural damage. These networks, which rely upon approximating the multivariate density of the training data, have been shown to be effective in some applications. However, quantification of decision errors, which must ultimately be used to rank their effectiveness, has received little attention. In this paper, a two-dimensional probabilistic pattern classifier (PPC) based on an L 2 detector is studied. Each dimension is modelled as a Gaussian random variable in order to directly study the effects introduced by a commonly used density estimator. The region of acceptance is examined for different parameters of the kernel density estimator. The detector behaviour of the original PPC is compared to theory. It is demonstrated that performance is affected by errors introduced by a kernel constructed using a finite set of data, and also from theoretical limitations inherent in the test statistic. A modification of the test statistic is suggested to improve the sensitivity for structural damage detection.

[1]  Adapting the classical kernel density estimator to data , 1990 .

[2]  E. J. Williams,et al.  STRUCTURAL DAMAGE DETECTION BY A SENSITIVITY AND STATISTICAL-BASED METHOD , 1998 .

[3]  Ming L. Wang,et al.  A health monitoring system for large structural systems , 1998 .

[4]  S. Law,et al.  Structural Damage Detection from Incomplete and Noisy Modal Test Data , 1998 .

[5]  Ephrahim Garcia,et al.  Structural Damage Identification: A Probabilistic Approach , 1998 .

[6]  M. Rosenblatt.  Remarks on a Multivariate Transformation , 1952 .

[7]  S. Kay.  Fundamentals of statistical signal processing: estimation theory , 1993 .

[8]  D. W. Scott,et al.  On Locally Adaptive Density Estimation , 1996 .

[9]  Asymptotically optimal bandwidth selection Of the kernel density estimator under The proportional hazards model , 1993 .

[10]  Donald F. Specht,et al.  Probabilistic neural networks , 1990, Neural Networks.

[11]  Charles R. Farrar,et al.  BRIDGE MODAL PROPERTIES USING SIMPLIFIED FINITE ELEMENT ANALYSIS , 1998 .

[12]  Joseph P. Noonan,et al.  False Alarm Probability of a DWT-Based Estimation Algorithm , 1996, Digit. Signal Process..

[13]  C. R. Farrar,et al.  Statistical damage identification techniques applied to the I-40 bridge over the Rio Grande River , 1998 .

[14]  Tommy Öberg,et al.  Adaptive detection of known signals in additive noise by means of kernel density estimators , 1997, IEEE Trans. Inf. Theory.

[15]  Rune Brincker,et al.  Proceedings of the 10th International Modal Analysis Conference , 1993 .

[16]  Keith Worden,et al.  CONFIDENCE BOUNDS FOR FREQUENCY RESPONSE FUNCTIONS FROM TIME SERIES MODELS , 1998 .

[17]  B. Silverman.  Density estimation for statistics and data analysis , 1986 .

[18]  T. L. Paez,et al.  Neural networks for damage identification , 1997 .

[19]  Anne S. Kiremidjian,et al.  Method for Probabilistic Evaluation of Seismic Structural Damage , 1996 .

[20]  T. L. Paez,et al.  Statistical analysis of modal parameters using the bootstrap , 1996 .

[21]  Carl W. Helstrom,et al.  Probability and stochastic processes for engineers , 1984 .

[22]  Jenq-Neng Hwang,et al.  Nonparametric multivariate density estimation: a comparative study , 1994, IEEE Trans. Signal Process..