when using pulsed eddy current technology to detect the wall thickness of pressure equipment, if the defect size to be detected is smaller than the sensor size, the actual thickness to be detected is much greater than the true thickness of the equipment. Through the simulation analysis, it is found that the detection thickness obtained by using a single characteristic quantity inversion is the average thickness of the specimen within the receiving range of the sensor, rather than the real thickness of the specimen defect. In order to solve this problem, firstly, according to the minimum feature number of defect type, the mapping relationship between multiple feature quantities and defect features of specimen is established. Then, PCA technology is used to determine the best combination of feature quantities. Finally, a detection method based on multiple feature quantities is proposed to inverse the thickness of specimen. In this paper, the feasibility of the method is verified by the simulation data, and the experimental verification is carried out by using the flat bottom blind hole test piece. The results show that the wall thickness inversion error of flat bottom blind hole test piece can be controlled within 5%, which indicates that the method has a wide range of engineering application value.