开发了一种压电加速度传感器在线故障的智能化诊断方法。首先,将压电传感器尽可能全的故障按故障类型、故障原因、故障现象或故障信号分析特征、故障处理方案等信息形成"压电传感器故障信息专家库";其次,对压电传感器进行在线数据采集和信号分析,提炼出与故障有关的信号特征参数;第三,利用一个内嵌的专家系统的搜索与推理机制,根据上述信号特征参数高效、智能地匹配出最相关的故障类型和故障原因;最后,专家系统的诊断结果呈现给现场运行维护工程师,进行人工确认,最终得到故障诊断结果。该方法能够有效指导操作人员排查故障,目前已经移植于一套专用于现场故障诊断的便携式仪器,并成功应用于某核电厂某机组的调试运行中。
An online intelligent fault diagnosis of piezoelectric accelerometer sensors was proposed. First fault information expert library was established with fault types,causes,symptoms,signal analysis features,fault diagnosis program and other information of piezoelectric sensor. Second,online data collection and signal analysis of piezoelectric sensors were done and related failure signal characteristic parameters are extracted. Thirdly,with the searching and inference mechanism of embedded expert system,the most relevant fault type and fault cause were efficiently and intelligently mapped according to the above signal characteristic parameters. Finally,diagnosis results of expert system were presented to the operation and maintenance engineer at field,and manual confirmation is done. This method can effectively guide the operator at field. This method has been ported to a dedicated fault diagnosis portable instrument,and successfully applied to commission of a nuclear plant unit.