Abstract: This study conducted inversion analysis of pipeline wall thickness with coatings based on pulsed eddy current testing technology. First, parametric models with varying aluminum layer thickness, insulation thickness, pipeline wall thickness, and thinning thickness were established using COMSOL software, with late-stage slopes of eddy current signals extracted as characteristic values. Then, the gradient boosting decision tree (GBDT) algorithm was used to train simulation data and construct a wall thickness inversion model. Finally, a pulsed eddy current testing platform was built to validate the model′s inversion effectiveness by collecting actual eddy current signals. Experimental results demonstrated that the established model could effectively invert pipeline wall thickness under different coating thicknesses.