一种基于振动图像和结构模型的通用智能故障诊断系统

An intelligent fault diagnosis system based on spectrum image and machinery model

提出一种利用振动频谱图像提取旋转机械故障特征的方法。针对旋转机械设备种类多样性及专家系统诊断对象范围窄的特点,研究基于结构模型的多征兆模糊产生式规则的诊断、知识表示和以关系数据库为基础的知识库的建立与维护机制,开发基于振动图像和设备结构模型的故障诊断专家系统,系统在满足知识表示和推理模型的有效性和实用性需求下,具有构造简单,推理效率高等优点,大大提高了旋转机械故障诊断专家系统的适应性和通用性。

A fault feature extraction method of rotating machinery based on vibration spectrum image recognition is put forward in order to recognize fault automatically. On the basis of the variety of rotating machinery and the limitation of expert system which can only be applied in several cases, the expression of multiple symptom fuzzy diagnosis knowledge based on machinery model and the establishment of safeguarding mechanism of rule base are studied. And a kind of fault diagnosis expert system based on vibration spectrum image and machinery model is developed. Besides validity and practicability of knowledge expression and reasoning model, the expert system has the merits of being simple to construct and high reasoning efficiency. Furthermore, the adaptability and universality of fault diagnosis expert system for rotating machinery are greatly increased.