AE05.pdf

As an alternative pattern recognition has been

This preview shows page 17 - 19 out of 39 pages.

cannot be easily generalized. As an alternative, pattern recognition has been investigated to develop universally applicable acoustic emission analysis methodologies. Supervised and unsupervised pattern recognition for acoustic emission test data analysis has been applied in research settings 26-36 and in industry. 35-47 Some basic pattern recognition techniques are outlined below as an introduction to the thinking and the analysis necessary to implement decision rules for acoustic emission testing. Pattern recognition is well documented in the literature. 1-6 Feature Vector and Pattern Matrix An acoustic emission signal is often described by using a number of its characteristics, or features. Most often a fixed set of d features is measured for any acoustic emission hit or event to be classified. For pattern recognition analysis, it is convenient (for both presentation and mathematical formulation) to represent the acoustic emission hit or event as a feature vector X , where X is a 158 Acoustic Emission Testing F IGURE 9. Typical acoustic emission signal showing some of the features extracted. The features vector for each hit (record) represents that hit as a vector in a multidimensional feature space. Amplitude (relative scale) Time (relative scale) Counts to peak Counts Duration Signal arrival time Amplitude Rise Time Threshold Baseline
Image of page 17

Subscribe to view the full document.

d -dimensional column vector of a signal like that in Fig. 9. Thus, pattern X is a point in a d -dimensional feature space. By this process of feature measurement and representation, an acoustic emission hit or event can be conceived abstractly as a point in multidimensional space. The collection of all the pattern vectors (acoustic emission hits or data points in general) defines the pattern matrix. Thus, the pattern matrix contains the whole data set and is easily viewed in a data table format. The concept of the pattern matrix is shown in Fig. 10, where each row represents one acoustic emission hit (record or pattern vector) composed of extracted features from the corresponding waveform. The selection of acoustic emission features is a key issue for the success of the entire pattern recognition methodology. Feature selection and the design of the feature extractor strongly depend on the application. The ideal set of features (measured by the acoustic emission test system or extracted from waveforms in postprocessing) would produce the same feature vector X for all patterns in the same class and substantially different feature vectors for all patterns in different classes. In addition to that, the feature vector should be composed of uncorrelated acoustic emission features. For this purpose, the covariance or the correlation matrix of acoustic emission features should be examined or subjected to hierarchical clustering to assist the selection. 11,12 Adding or removing a single feature might change the results drastically. The data structure in Fig. 11 represents eight sphere shaped classes well separated in the three-dimensional feature space.
Image of page 18
Image of page 19

{[ snackBarMessage ]}

What students are saying

  • Left Quote Icon

    As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

    Student Picture

    Kiran Temple University Fox School of Business ‘17, Course Hero Intern

  • Left Quote Icon

    I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

    Student Picture

    Dana University of Pennsylvania ‘17, Course Hero Intern

  • Left Quote Icon

    The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

    Student Picture

    Jill Tulane University ‘16, Course Hero Intern