基于交互监测与连通性模型的化工过程故障传播分析

doi:10.11949/0438-1157.20250114

摘要/Abstract

摘要：

复杂化工过程中监测变量存在自相关与互相关的时空耦合关系，导致在故障传播路径识别过程中容易出现冗余信息，造成路径的错误识别。为此，提出一种融合监测数据与过程知识的故障传播路径回溯方法，以基于k近邻的故障传播路径分析方法为框架，引入分布式交互监测以确定故障潜在区域并剔除冗余变量，从工艺过程中提取基于无向邻接矩阵的连通性模型，给故障路径回溯提供逻辑指导。通过Tennessee Eastman过程与合成氨工艺流程的故障案例，与传统传递熵方法、基于k近邻的故障传播路径分析方法相比，验证了所提方法有效地提高了故障路径识别精度和效率，同时减少了冗余备选路径。

关键词: 化工过程, 故障传播路径, k近邻, 分布式监测, 连通性模型

Abstract:

There is a spatiotemporal coupling relationship between autocorrelation and mutual correlation in the monitoring variables in complex chemical processes, which leads to redundant information in the process of fault propagation path identification, resulting in incorrect path identification. Therefore, a fault propagation path backtracking method that integrates monitoring data and process knowledge is proposed. Distributed interactive monitoring is introduced based on the fault propagation path analysis of k-nearest neighbors to determine the potential fault area and eliminate redundant variables. A connectivity model based on an undirected adjacency matrix is extracted from the process to provide logical guidance for fault path backtracking. Compared with the traditional transfer entropy method and the k-nearest neighbor method, the proposed method, through the fault cases of the Tennessee Eastman process and the ammonia synthesis process, effectively improves the accuracy of fault path identification and reduces redundant paths.

Key words: chemical processes, fault propagation path, k-nearest neighbor, interactive monitoring, connectivity model

中图分类号:

TP 274

钱小毅, 王利鑫, 姜兴宇, 孙天贺, 赵毅, 王一飞. 基于交互监测与连通性模型的化工过程故障传播分析[J]. 化工学报, 2025, 76(8): 4155-4164.

Xiaoyi QIAN, Lixin WANG, Xingyu JIANG, Tianhe SUN, Yi ZHAO, Yifei WANG. Fault propagation analysis of chemical process based on interactive monitoring and connectivity model[J]. CIESC Journal, 2025, 76(8): 4155-4164.

图/表 17

参考文献 30

[1]	张中秋, 李宏光, 石逸林. 基于人工预测调控策略的复杂化工过程多任务学习方法[J]. 化工学报, 2023, 74(3): 1195-1204.
	Zhang Z Q, Li H G, Shi Y L. A multi-task learning approach for complex chemical processes based on manual predictive manipulating strategies[J]. CIESC Journal, 2023, 74(3): 1195-1204.
[2]	贠程, 王倩琳, 陈锋, 等. 基于社团结构的化工过程风险演化路径深度挖掘[J]. 化工学报, 2023, 74(4): 1639-1650.
	Yun C, Wang Q L, Chen F, et al. Deep-mining risk evolution path of chemical processes based on community structure[J]. CIESC Journal, 2023, 74(4): 1639-1650.
[3]	黄雄峰, 王鍚, 朱严严, 等. 考虑传播过程的应急网络控制系统故障恢复[J]. 控制理论与应用, 2024, 41(5): 875-884.
	Huang X F, Wang Y, Zhu Y Y, et al. Fault recovery of emergency network control system considering propagation processe[J]. Control Theory & Applications, 2024, 41(5): 875-884.
[4]	申桂香, 栾兰, 张英芝, 等. 加工中心组件故障传播影响力评估[J]. 吉林大学学报(工学版), 2022, 52(1): 63-69.
	Shen G X, Luan L, Zhang Y Z, et al. Fault propagation impact assessment of machining center components [J]. Journal of Jilin University (Engineering and Technology Edition), 2022, 52(1): 63-69.
[5]	Huang S J, Lu N Y, Jiang B, et al. Fault propagation analysis of computer numerically controlled machine tools[J]. Journal of Manufacturing Systems, 2023, 70: 149-159.
[6]	Wang T, Wei X G, Huang T, et al. Modeling fault propagation paths in power systems: a new framework based on event SNP systems with neurotransmitter concentration[J]. IEEE Access, 2019, 7: 12798-12808.
[7]	Peng C, Ouyang Y Y, Gui W H, et al. Equipment fault propagation path identification based on unstable points detection[J]. IEEE Transactions on Industrial Informatics, 2024, 20(2): 1596-1606.
[8]	刘鹏鹏, 左洪福, 苏艳, 等. 基于图论模型的故障诊断方法研究进展综述[J]. 中国机械工程, 2013, 24(5): 696-703.
	Liu P P, Zuo H F, Su Y, et al. Review of research progresses for graph-based models in fault diagnosis method[J]. China Mechanical Engineering, 2013, 24(5): 696-703.
[9]	梁思远, 周金浛, 高占宝, 等. 机电系统健康状态预测和维修决策的双向优化方法[J]. 仪器仪表学报, 2023, 44(1): 131-142.
	Liang S Y, Zhou J H, Gao Z B, et al. Bi-directional optimization method for health state prediction and maintenance decision-making of electromechanical systems[J]. Chinese Journal of Scientific Instrument, 2023, 44(1): 131-142.
[10]	Wang Y, Shang X Y, Peng K. Relocating mining microseismic earthquakes in a 3-D velocity model using a windowed cross-correlation technique[J]. IEEE Access, 2020, 8: 37866-37878.
[11]	Yang B, Li H G, Wen B. A dynamic time delay analysis approach for correlated process variables[J]. Chemical Engineering Research and Design, 2017, 122: 141-150.
[12]	Liu Y, Chen H S, Wu H B, et al. Simplified Granger causality map for data-driven root cause diagnosis of process disturbances[J]. Journal of Process Control, 2020, 95: 45-54.
[13]	李亚晓, 李青山, 王璐, 等. AmazeMap: 基于多层次影响图的微服务故障定位方法[J]. 软件学报, 2024, 35(7): 3115-3140.
	Li Y X, Li Q S, Wang L, et al. AmazeMap: microservices fault localization method based on multi-level impact graph[J]. Journal of Software, 2024, 35(7): 3115-3140.
[14]	郭文强, 李建望, 黄梓轩, 等. 基于互信息-主成分分析-贝叶斯网络的化工过程故障诊断方法[J]. 科学技术与工程, 2023, 23(21): 9144-9150.
	Guo W Q, Li J W, Huang Z X, et al. MPBN-based fault diagnosis method for chemical processes[J]. Science Technology and Engineering, 2023, 23(21): 9144-9150.
[15]	王林, 宋蓓, 张友卫, 等. 考虑父节点的贝叶斯网络故障路径追溯算法[J]. 计算机科学与探索, 2018, 12(11): 1796-1805.
	Wang L, Song B, Zhang Y W, et al. Bayesian-network-based fault path tracking algorithm considering parent nodes[J]. Journal of Frontiers of Computer Science and Technology, 2018, 12(11): 1796-1805.
[16]	冯双, 杨浩, 崔昊, 等. 基于Copula传递熵的设备级和网络级宽频振荡传播路径分析及振荡源定位方法[J]. 电工技术学报, 2023, 11: 1-15.
	Feng S, Yang H, Cui H, et al. Propagation path analysis and oscillation source location method of device-level and network-level broadband oscillation based on Copula transfer entropy[J]. China Industrial Economics, 2023, 11: 1-15.
[17]	Wu Y C, Zhou Y, Song M. Classification of patients with AD from healthy controls using entropy-based measures of causality brain networks[J]. Journal of Neuroscience Methods, 2021, 361: 109265.
[18]	Bauer M, Cox J W, Caveness M H, et al. Nearest neighbors methods for root cause analysis of plantwide disturbances[J]. Industrial & Engineering Chemistry Research, 2007, 46(18): 5977-5984.
[19]	Stockmann M, Haber R, Schmitz U. Source identification of plant-wide faults based on k nearest neighbor time delay estimation[J]. Journal of Process Control, 2012, 22(3): 583-598.
[20]	Liu H, Pi D C, Qiu S Y, et al. Data-driven identification model for associated fault propagation path[J]. Measurement, 2022, 188: 110628.
[21]	马亮, 彭开香, 董洁. 工业过程故障根源诊断与传播路径识别技术综述[J]. 自动化学报, 2022, 48(7): 1650-1663.
	Ma L, Peng K X, Dong J. Review of root cause diagnosis and propagation path identification techniques for faults in industrial processes[J]. Acta Automatica Sinica, 2022, 48(7): 1650-1663.
[22]	Landman R, Jämsä-Jounela S L. Fault propagation analysis by implementing nearest neighbors method using process connectivity[J]. IEEE Transactions on Control Systems Technology, 2019, 27(5): 2058-2067.
[23]	Chen W Y, Yu M Y, Wu P. Rotor-stator rub-impact fault and position identification of aero-engine based on VMD-MF-cepstrum-KNN[J]. Tribology Transactions, 2023, 66(1): 23-34.
[24]	Cecílio I M, Ottewill J R, Fretheim H, et al. Removal of transient disturbances from oscillating measurements using nearest neighbors imputation[J]. Journal of Process Control, 2016, 44: 68-78.
[25]	曹跃, 陈志文, 袁小锋, 等. 部分子块通讯的分布式PCA厂级工业过程监测方法[J]. 控制与决策, 2020, 35(6): 1281-1290.
	Cao Y, Chen Z W, Yuan X F, et al. Distributed PCA for plant-wide processes monitoring with partial block communication[J]. Control and Decision, 2020, 35(6): 1281-1290.
[26]	袁健宝, 王政, 徐一凡, 等. 基于复杂网络边负载分配理论的化工过程级联故障风险传播路径[J]. 化工进展, 2019, 38(8): 3525-3533.
	Yuan J B, Wang Z, Xu Y F, et al. Risk propagation path of cascading fault in chemical process based on edge load distribution in complex network[J]. Chemical Industry and Engineering Progress, 2019, 38(8): 3525-3533.
[27]	Amin M T. An integrated methodology for fault detection, root cause diagnosis, and propagation pathway analysis in chemical process systems[J]. Cleaner Engineering and Technology, 2021, 4: 100187.
[28]	Ricker N L. Optimal steady-state operation of the Tennessee Eastman challenge process[J]. Computers & Chemical Engineering, 1995, 19(9): 949-959.
[29]	郑天标, 肖应旺. 基于核主元分析与核密度估计的非线性过程故障监测与识别[J]. 计算机系统应用, 2022, 31(10): 329-334.
	Zheng T B, Xiao Y W. Nonlinear process fault identification and detection based on KPCA-KDE[J]. Computer System Applications, 2022, 31(10): 329-334.
[30]	朱君烨. 时序数据驱动的化工过程风险动态预警研究: 以合成氨为例[D]. 昆明: 昆明理工大学, 2023.
	Zhu J Y. Study on dynamic early warning of chemical process risk driven by time series data: taking synthetic ammonia as an example[D]. Kunming: Kunming University of Science and Technology, 2023.

节点	名称	符号	节点	名称	符号	节点	名称	符号
1	A进料流量	F1	12	气液分离器液位	L12	23	D进料流量控制阀	V23
2	D进料流量	F2	13	气液分离器压力	P13	24	E进料流量控制阀	V24
3	E进料流量	F3	14	气液分离器底部出口流量	F14	25	A进料流量控制阀	V25
4	ABC总进料流量	F4	15	汽提塔液位	L15	26	ABC总进料流量控制阀	V26
5	压缩机出口再循环流量	F5	16	汽提塔压力	P16	27	压缩机再循环阀	V27
6	反应器进料总流量	F6	17	汽提塔塔底流量	F17	28	排放阀	V28
7	反应器压力	P7	18	汽提塔温度	T18	29	气液分离器底部出口阀	V29
8	反应器液位	L8	19	汽提塔流量	F19	30	汽提塔产品出口阀	V30
9	反应器温度	T9	20	压缩机功率	J20	31	汽提塔水流阀	V31
10	排放速度	F10	21	反应器冷却水出口温度	T21	32	反应器冷却水流阀	V32
11	气液分离器温度	F11	22	冷凝器冷却水出口温度	T22	33	冷凝器冷却水流阀	V33

节点	名称	符号	节点	名称	符号	节点	名称	符号
1	A进料流量	F1	12	气液分离器液位	L12	23	D进料流量控制阀	V23
2	D进料流量	F2	13	气液分离器压力	P13	24	E进料流量控制阀	V24
3	E进料流量	F3	14	气液分离器底部出口流量	F14	25	A进料流量控制阀	V25
4	ABC总进料流量	F4	15	汽提塔液位	L15	26	ABC总进料流量控制阀	V26
5	压缩机出口再循环流量	F5	16	汽提塔压力	P16	27	压缩机再循环阀	V27
6	反应器进料总流量	F6	17	汽提塔塔底流量	F17	28	排放阀	V28
7	反应器压力	P7	18	汽提塔温度	T18	29	气液分离器底部出口阀	V29
8	反应器液位	L8	19	汽提塔流量	F19	30	汽提塔产品出口阀	V30
9	反应器温度	T9	20	压缩机功率	J20	31	汽提塔水流阀	V31
10	排放速度	F10	21	反应器冷却水出口温度	T21	32	反应器冷却水流阀	V32
11	气液分离器温度	F11	22	冷凝器冷却水出口温度	T22	33	冷凝器冷却水流阀	V33

区块编号	区块名称	变量
1	反应器	1，2，3，5，6，7，8，9，21
2	冷凝器和分离器	7，8，9，11，12，13，14，22
3	汽提塔	4，11，12，13，14，15，16，17，18，19
4	压缩机	5，10，11，12，13，20

区块编号	区块名称	变量
1	反应器	1，2，3，5，6，7，8，9，21
2	冷凝器和分离器	7，8，9，11，12，13，14，22
3	汽提塔	4，11，12，13，14，15，16，17，18，19
4	压缩机	5，10，11，12，13，20

方法	准确率/%	潜在路径数/条
传递熵	85.71	16.00
FPA-kNN	82.75	17.00
DIM-CM-FPA-kNN	92.30	13.00