基于测点数据变化趋势告警判定的大型立磨故障诊断专家系统

Fault diagnosis expert system for large vertical mill with alarming judgement based on measured data variation trend

目前对大型立磨的故障诊断主要靠专家的经验判断,主观性强,故障排查效率低,实时性、准确性和智能化较差。本文研究了大型立磨状态监测及故障诊断专家系统的关键问题,针对状态监测中测点数据趋势类告警难以判定的问题,采用最小二乘法拟合出数据序列斜率,根据斜率的相对变化值来判定测点数据变化趋势的告警情况;根据专家故障诊断推理流程构建故障树,提取出诊断规则;根据状态检测和故障诊断的实体关系模型,构建了专家知识库;基于C#语言、SQL server数据库开发了大型立磨状态监测及故障诊断专家系统。通过实例验证表明,该系统能够对立磨异常状态及时报警,通过推理快速准确找出故障原因,并对推理过程作出清晰的解释。

At present, the fault diagnosis of the large vertical mill mainly depends on the experience knowledge of field experts, which is characterized by strong subjectivity, low efficiency, as well as poor realtime performance, accuracy and intelligence. In the paper, the key issues of the state monitor and fault diagnosis expert system for the large vertical mill were studied. Aiming at the difficulty in judging the alarming of the measured data variation trend during state monitor, the least square method was used to fit the slope of data sequence. According to the relative variation of the slope, the alarming of the measured data variation trend was judged. According to the fault diagnosis inference process of experts in the field, the fault tree was constructed and the diagnosis rules were extracted. Based on the entity relationship model of state monitor and fault diagnosis, the expert knowledge base was constructed. In addition, the state monitor and fault diagnosis expert system for the large vertical mill was developed based on language C and SQL server database. The example showed that the system alarmed the abnormal state in time, found out the reason of the fault quickly and accurately, and made clear explanation to the reasoning process.