基于趋势相似性的二甲酚尾气处理过程监测  

Monitoring of Xylenol Tail Gas Treatment Process Based on Trend Similarity

作　　者：许飞鸿 栾小丽 刘飞 XU Fei-hong;LUAN Xiao-li;LIU Fei(Key Laboratory of Advanced Process Control for Light Industry of the Ministry of Education,Jiangnan University,Wuxi 214122,China)

机构地区：[1]江南大学轻工过程先进控制教育部重点实验室,江苏无锡214122

出　　处：《控制工程》2022年第2期231-237,共7页Control Engineering of China

基　　金：国家自然科学基金资助项目(61991402)。

摘　　要：利用工业锅炉对二甲酚的尾气进行处理时,存在爆炸和有毒尾气泄漏的风险,因此对二甲酚尾气处理过程进行故障监测十分必要。然而该过程具有非平稳特征,常规的故障监测方法监测准确率不高。为了解决这个问题,提出了基于趋势相似性特征的故障监测方法,通过滑动时间窗口切割时间序列,计算各时间窗口内数据之间的趋势相似性,并以此作为新的监控特征对该非平稳过程进行监测,从而提高监测准确率。最后,以某二甲酚生产企业采集的现场数据进行验证。结果表明,基于趋势相似性特征的故障监测较常规方法有较高的准确率,且监测准确率随过程的非平稳性增强而提高,验证了所提方法的有效性及实际应用价值。The use of industrial boilers to treat the tail gas of xylenol carries the risk of explosion and toxic tail gas leakage, so fault monitoring of the xylenol tail gas treatment process is necessary. However, the process has non-stationary characteristics, and the monitoring accuracy of conventional fault monitoring methods is not high. In order to solve this problem, a fault monitoring method based on trend similarity feature is proposed to improve the monitoring accuracy. The time series are cut by sliding time windows, the trend similarity between data within each time window is calculated, and it is used as a new monitoring feature for this non-stationary process. Finally, the field data collected from a xylenol plant are used for validation. The results show that the fault detection based on the trend similarity feature has a higher accuracy than the conventional method, and the detection accuracy increases with the non-stationarity of the process, which verifies the effectiveness and practical application value of the proposed method.

关 键 词：工业锅炉 二甲酚尾气处理 故障监测 非平稳过程 趋势相似性 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]