关键词: 直径检测, 麻雀搜索算法优化-随机森林回归, 无损检测, 随机森林

Abstract:

To address the issue of diameter reduction in structural components due to corrosion, a non-destructive testing method based on the sparrow search algorithm-optimized random forest regression (SSA-RFR) is proposed. Firstly, the eddy current response signals from specimens with varying diameters are collected under 1 kHz pulsed excitation. Kalman filtering is applied to denoise the raw signals,enhancing their quality.Subsequently, the key feature parameters are extracted via exponential function fitting to serve as effective inputs for modeling.Finally, an SSA-RFR predictive model is developed for regression analysis and prediction of component diameter.Experimental results demonstrate that the mean squared error (MSE) of the proposed method is 0.310 6, representing a reduction of approximately 97.9% compared to the traditional Random Forest Regression model and markedly enhancing prediction accuracy.These findings validate the effectiveness and engineering applicability of the proposed SSA-RFR method for non-destructive testing.

Key words: diameter detection, sparrow search algorithm-optimized random forest regression (SSA-RFR), non-destructive testing, random forest