An intelligent diagnostics system, that integrates data management methods, signal processing, artificial neural networks, and decision making using fuzzy logic for the automation of steam generator eddy current test (ECT) data analysis, is presented. The following key issues were identified and developed for establishing a robust analysis system: (1) Digital eddy current test data calibration, compression and representation; (2) flaw classification and decision confidence estimation usign fuzzy logic; (3) development of robust neural networks for tube defect sizing; (4) system integration for database management, compilation of a trained neural network library, and a decision module. The fuzzy logic based flaw detection system was the first to utilize information from multi-frequency eddy current data for flaw detection. The system was tested extensively usign ECT data to establish its ability to identify defect types and to estimate defect parameters.
展开▼
