Abstract: The present paper is a review of the recent research work on the methodology for predicting the occurrence of failure events for process equipment at home and abroad.The focal point is the study progress and some effective cases in practice,which is based on the principles of approximate and plausible reasonings for uncertain parameters of fuzziness, quantitative metallographic technique and artificial neural networks method.The prediction involves failure probability, serious ness of failure events,damage of structural materials, remaining life and safety,etc．

摘要： 综述了近年来国内外化工设备事故预测技术的研究情况。结合作者及其所在科研组的研究成果，重点介绍基于三项技术原理──根据模糊集理论、用非精确推理方法解决事故预测中模糊不确定问题的技术，定量金相技术和人工神经网络技术──推导出的化工设备缺陷危害严重程度、材料损伤度、可靠度、剩余寿命和安全性等预测方法的研究进展，以及工程应用取得明显成效的案例。