Fuzzy Logic Applied to Eddy-Current Inspection of Metallic Automotive Components

Eddy-Current Testing applied to metallic automotive components production is used to monitor and diagnose deviations during the manufacturing phases. The signals are contaminated by noises, making the analysis of signals very difficult and sometimes impossible. The interpretation of these signals relies on subjective judgment, increasing the risk of over conservative or wrong conclusions.

In this work, due to the transient characteristic of the signal, the Wavelet Transform is used to de-noise the original signal. The defect characteristics as defined in the ASME Code, can be obtained with much more accuracy. Also, a Fuzzy Logic inference engine is used to help inspector decision. Experimental data with different defects geometries were used.