在监测设备运行状态和诊断设备故障时,目前采用的方法是在不同的部位安装若干个传感器拾取一定的信号进行分析.试验表明,诊断结果的准确性不仅与传感器安装的部位有关,而且与传感器的数量、类型以及诊断方法有关.这是因为诊断对象运行工况复杂、影响因素众多,某一故障可能对应若干征兆,而某一征兆也可能对应若干故障,它们之间存在着错综复杂的关系,加上故障和征兆信息的随机性、模糊性,构成了信息的不确定性.因此,故障的多样性、不确定性、相关性以及传播性构成了故障诊断技术上的难点,对于一个复杂的设备系统进行故障诊断,要做到及时准确,单靠某一故障特征量和诊断方法无法完成诊断任务.克服这一困难的方法之一是利用数据集成与信息融合技术进行故障诊断. In the monitoring of equipment operation and diagnosis of equipment failure, the current method is to install a number of sensors in different parts of the pick up a certain number of signals for analysis.Experiments show that the accuracy of the diagnosis results not only with the sensor installed parts, but also with the sensor The number and type of diagnoses and the diagnosis methods.This is because the diagnostic objects have complex operating conditions and many influencing factors, and one symptom may correspond to several symptoms, and one symptom may correspond to several faults, and there are complex relationships between them. Together with the randomness and fuzziness of the fault and symptom information, constitute the uncertainty of the information.Therefore, the diversity, uncertainty, relativity and communication of the fault constitute the technical difficulties in fault diagnosis, and for a complex Equipment system for fault diagnosis, to be timely and accurate, relying solely on a fault feature and diagnostic methods can not complete the diagnostic task.To overcome this difficulty, one of the methods is the use of data integration and information fusion technology for fault diagnosis.