铁磁性材料在各领域生产设备中广泛应用，然而在生产制造和使用过程中，受力学负荷及环境介质影响，易产生裂纹、腐蚀等缺陷，威胁工业设备安全运行。基于此，开展了三轴信息融合的缺陷量化方法研究，为交流电磁场检测技术对不同缺陷的分类及反演提供技术支持。基于交流电磁场检测技术（alternating current field measurement, ACFM）原理，建立不同缺陷交流电磁场仿真模型，分析裂纹和腐蚀缺陷三轴磁通量信号与对应缺陷尺寸的关系，确定正反演规律，同时设计搭建实验平台，扩充数据量并建立三轴信息融合数据库，借助神经网络开展裂纹和腐蚀缺陷的自动分类与量化研究。结果显示，裂纹和腐蚀缺陷三轴磁通量信号与缺陷尺寸呈显著相关性，成功构建了准确反映该关系的量化模型，神经网络训练中，模型对裂纹和腐蚀缺陷的自动分类准确率达100%，缺陷量化反演精度在94%以上。基于交流电磁场检测技术的三轴信息融合缺陷量化方法可行且有效，能够准确对铁磁性材料中的裂纹和腐蚀缺陷进行自动分类和量化，为工业设备中铁磁性材料缺陷的检测提供了可靠技术手段，有助于提升工业设备运行的安全性与稳定性。

Ferromagnetic materials are widely applied in production equipment in various fields. However, during the manufacturing and operation processes, they are easily affected by mechanical loads and environmental media, and defects such as cracks and corrosion are likely to be generated, which threaten the safe operation of industrial equipment. Based on this, the defect quantification method of three-axis information fusion was carried out to provide technical support for the classification and inversion of different defects by ACFM (alternating current field measurement) technique. Based on the principle of the ACFM technique, simulation models of alternating current magnetic fields for different defects were established. The relationship between the three-axis magnetic flux signals of crack and corrosion defects and the corresponding defect sizes was analyzed, and the forward and inverse laws were determined. Meanwhile, an experimental platform was designed and built, the data volume was expanded, and a three-axis information fusion database was established. With the help of neural networks, the automatic classification and quantification research of crack and corrosion defects was carried out. The results show that there is a significant correlation between the three-axis magnetic flux signals of crack and corrosion defects and the defect sizes. A quantification model that accurately reflects this relationship has been successfully constructed. During the neural network training, the automatic classification accuracy of the model for crack and corrosion defects reaches 100%, and the defect quantification inversion accuracy is above 94%. The defect quantification method of three-axis information fusion based on the ACFM technique in this research is feasible and effective. It can accurately automatically classify and quantify the crack and corrosion defects in ferromagnetic materials, provides a reliable technical means for the detection of defects in ferromagnetic materials in industrial equipment, and helps to improve the safety and stability of the operation of industrial equipment.