No abstract

Automated signal classification systems are finding increasing use in many applications for the analysis and interpretation of large volumes of signals. Such systems show consistency of response and help reduce the effect of variabilities associated with human interpretation. This paper deals with the analysis of ultrasonic NDE signals obtained during weld inspection of piping in boiling water reactors. The overall approach consists of three major steps, namely, frequency invariance, multiresolution analysis, and neural network classification. The data are first preprocessed whereby signals obtained using different transducer center frequencies are transformed to an equivalent reference frequency signal. Discriminatory features are then extracted using a multiresolution analysis technique, namely, the discrete wavelet transform (DWT). The compact feature vector obtained using wavelet analysis is classified using a multilayer perceptron neural network. Two different databases containing weld inspection signals have been used to test the performance of the neural network. Initial results obtained using this approach demonstrate the effectiveness of the frequency invariance processing technique and the DWT analysis method employed for feature extraction.

Eddy current-based techniques have been investigated for the inspection of embedded cracks under fastener heads in riveted structures. However, these techniques are limited in their ability to detect cracks that are not perpendicular to induced current flows. Further, the presence of a steel fastener of high permeability produces a strong signal that masks relatively smaller indication from a crack. In this paper, a rotating electromagnetic field is designed to rotate the applied magnetic fields and related eddy currents electrically so that the sensor shows uniform sensitivity in detecting cracks in all radial directions around fastener sites. Giant magnetoresistive sensors are employed to image the normal component of this rotating field, to detect different crack orientations at aluminum and ferromagnetic fastener sites. Numerical model-based studies and experimental validation are presented.Index Terms-Eddy current, giant magnetoresistive sensor, rotating electromagnetic field, fastener hole inspection, NDE.

The solution of partial differential equations (PDE) arises in a wide variety of engineering problems. Solutions to most practical problems use numerical analysis techniques such as finite-element or finite-difference methods. The drawbacks of these approaches include computational costs associated with the modeling of complex geometries. This paper proposes a finite-element neural network (FENN) obtained by embedding a finite-element model in a neural network architecture that enables fast and accurate solution of the forward problem. Results of applying the FENN to several simple electromagnetic forward and inverse problems are presented. Initial results indicate that the FENN performance as a forward model is comparable to that of the conventional finite-element method (FEM). The FENN can also be used in an iterative approach to solve inverse problems associated with the PDE. Results showing the ability of the FENN to solve the inverse problem given the measured signal are also presented. The parallel nature of the FENN also makes it an attractive solution for parallel implementation in hardware and software.

Magnetic flux leakage (MFL) methods are commonly used in the nondestructive evaluation (NDE) of ferromagnetic materials. An important problem in MFL NDE is the determination of flaw parameters such as the flaw length, depth, and shape (profile) from the measured values of the flux density B. Commonly used methods use a forward model in a loop to determine B for a given set of flaw parameters. This approach iteratively adjusts the flaw parameters to minimize the error between the measured and predicted values of B. This article proposes the use of neural networks as forward models. The proposed approach uses two neural networks in feedback configuration—a forward network and an inverse network. The second network is used to predict the profile given the measured value of B, and acts to constrain the solution space. Results of applying these methods to MFL data obtained from a two-dimensional finite-element model, with rectangular flaws of various dimensions, are presented.

We introduce a supervised learning algorithm that gives neural network classification algorithms the capability of leaming incrementally from new data without forgetting what has been learned in earlier training sessions. Schapire's boosting algorithm, originally intended for improving the accuracy of weak leamers, has been modified to be used in an incremental leaming setting. The algorithm is based on generating a number of hypotheses using different distributions of the training data and combining these hypotheses using a weighted majority voting. This scheme allows the classifier previously trained with a training database, to leam from new data when the original data is no longer available, even when new classes are introduced. Initial results on incremental training of multilayer perceptron networks on synthetic as well as real-world data are presented in this paper.

No abstract

No abstract