时序数据分析的复杂化工过程异常智能溯源研究

Research on intelligent traceability of abnormalities in complex chemical processes based on time series data analysis

对复杂化工过程异常工况进行智能推理溯源是实现安全关口前移、降低灾难性事故发生的有效途径。提出了一种基于Spearman-Apriori的化工过程异常智能溯源分析方法,旨在研究复杂化工过程异常工况发生的前置原因,并形成一种智能决策模型。针对化工工艺参数之间耦合性强、关联关系分析难度大的特点,引入Spearman相关系数,通过Spearman实时在线分析过程参数间的相关关系,并设置强关联阈值将Spearman相关系数分析与Apriori算法进行关联耦合,利用Apriori算法中的支持度和置信度二维挖掘各参数之间的超强关联规则。将该方法应用于合成氨工艺中合成工段的异常工况智能推溯,并选取氢氮比、管路工艺气流量、给水换热器冷凝剂流量等8个关键监测指标,研究发现氢氮比增大和给水换热器冷凝剂流量升高分别是导致合成塔入口压力超压、合成塔第一床层温度过低两组异常工况的前置原因,该分析结果与实际生产工艺相符,证明该方法可以有效地对化工过程异常原因进行推溯并筛选主要影响因素。研究为使用生产过程大数据实现化工过程异常智能溯源提供了理论基础,为进一步完善过程风险精细化管控提供了新思路。

Intelligent reasoning and tracing of abnormal working conditions in complex chemical processes is an effective way to achieve the forward movement of safety checkpoints and reduce the occurrence of catastrophic accidents.This study proposes an intelligent traceability analysis method for chemical process anomalies based on Spearman-Apriori,aiming to study the pre-causes of abnormal operating conditions in complex chemical processes and form an intelligent decision-making model.In light of the characteristics of strong coupling between chemical process parameters and great difficulty in correlation analysis,A Spearman correlation coefficient is introduced to analyze the correlation between process parameters in real-time online through Spearman and set a strong correlation threshold to associate Spearman correlation coefficient analysis with Apriori algorithm and use the support and confidence of Apriori algorithm to mine the super strong association rules between parameters in two dimensions.Applying this method to the intelligent tracing of abnormal working conditions in the synthesis section of the ammonia synthesis process,and selecting 8 key monitoring indicators such as hydrogen nitrogen ratio,pipeline process gas flow rate,and condensate flow rate in the feedwater heat exchanger,the study finds that an increase in hydrogen nitrogen ratio and a rise in condensate flow rate in the feedwater heat exchanger are the leading causes of two abnormal working conditions:overpressure at the inlet of the synthesis tower and low temperature in the first bed of the synthesis tower,The analysis results are consistent with the actual production process,proving that this method can effectively trace the causes of abnormal chemical processes and screen the main influencing factors.This study provides a theoretical basis for using production process big data to achieve intelligent traceability of chemical process anomalies and provides new ideas for further improving process risk refinement control.