Space-time-frequency joint sampling and Ridgelet transform based on impulse edge derivative excitation for rapid energy-efficient pulsed eddy current testing

Pulsed eddy current (PEC) testing is a vital non-destructive testing technique, but its effectiveness is often limited by suboptimal energy distribution and the challenge of extracting subtle defect signatures. To address these issues, this article proposes a novel method for rapid, energy-efficient inspection of multiscale defects. It integrates a space-time-frequency joint sampling (STF-JS) framework with a Ridgelet transform, powered by a tailored impulse edge derivative excitation (transient duration < 1 μs). This optimized excitation offers ultra-short duration for high-speed inspection, a broadband spectrum for fine-scale detection, low power consumption, and inherent potential to distinguish between inner diameter (ID) and outer diameter (OD) defects. The procedure acquires STF-JS signals via the Hilbert transform, applies a 2D Ridgelet transform for multiscale analysis to construct a test statistic matrix, and uses an iterative algorithm for defect identification, localisation, and classification (ID/OD, circumferential/axial). Experimental validations on steel plates demonstrate reliable multiscale crack detection. Furthermore, a field test using a self-developed in-line inspection (ILI) tool identified a real elbow crack at speeds exceeding 5 m/s, confirming the method’s high-speed, energy-efficient potential for industrial applications.