Abstract: Structural welding defects are vital safety hidden troubles to the crane, and acoustic emission(AE) technology can be used to monitor the defect changing process. Because of the complex and changeable application conditions for the truck crane, it is difficult to obtain the real AE signals of welding defect and the recognizing methods are not mature. Through adding ancillary testing specimens with welding defects to the principal structure, the breakage process in welded seam of structure was simulated, and the typical AE data during damage process were collected. The signal attenuation was tested, and in combination with characteristic parameter analysis methods and spectral characteristic analysis methods, the typical AE signals were analyzed. The result shows that AE technology can be used well for the accurate positioning of welding defects; and some characteristic parameters such as duration time, frequency centroid and peak frequency can be used to classify the defect signals and the noise.