基于简化因果图的工业过程故障根本变量诊断

在对工业过程故障进行根本原因诊断时,由于过程的自身特性和反馈控制等因素的干扰,使得变量因果图过于复杂从而使故障传播路径难以解释且不能找到导致故障的根本变量。提出一种简化因果图的方法,通过两步走对收敛交叉映射法构建的因果图实现简化,保留主要的故障传播路径。首先采用模糊综合评判法判别因果图中不确定性的关系;然后通过求解最大生成树,得到赋权无向图,并根据变量间因果关系选取根节点,分析赋权无向图获得新路径,从而将其改进成赋权有向图。在田纳西—伊斯曼过程进行验证实验,并与传统收敛交叉映射法进行比较,结果表明所提出方法的有效性。     

一种新颖的深度因果图建模及其故障诊断方法

为了实现复杂工业过程故障检测和诊断一体化建模, 提出了一种新颖的深度因果图建模方法. 首先, 利用循环神经网络建立深度因果图模型, 将Group Lasso稀疏惩罚项引入到模型训练中, 自动地检测过程变量间的因果关系. 其次, 利用模型学习到的条件概率预测模型对每个变量建立监测指标, 并融合得到综合指标进行整体工业过程故障检测. 一旦检测到故障, 对故障样本构建变量贡献度指标, 隔离故障相关变量, 并通过深度因果图模型的局部因果有向图诊断故障根源, 辨识故障传播路径. 最后, 通过田纳西?伊斯曼过程进行仿真验证, 实验结果验证了所提方法的有效性.     

Hybrid approach to casual analysis on a complex industrial system based on transfer entropy in conjunction with process connectivity information

Industrial processes often encounter disturbances that propagate through the process units and their control elements, leading to poor process performance and massive economic losses. Thus, one major concern in the chemical industry is the detection of disturbances and identification of their propagation path. Causal analysis based on process data is frequently applied to identify causal dependencies among process measurements and thereby obtain the propagation path of disturbances. One significant challenge in data-based causal analysis is investigating industrial systems with a high degree of connectivity due to multiple causal pathways. This paper proposes a new hybrid approach for detecting causality based on the transfer entropy (TE) method by incorporating process connectivity information using an explicit search algorithm. Based on the hybrid approach, initially, the TE is only calculated for pathways that are considered as direct pathways based on the process topology. Then, the direct transfer entropy (DTE) is employed to discriminate spurious and/or indirect pathways obtained by the initial TE results. To facilitate the DTE calculation, the search algorithm is invoked once again to extract the intermediate pathways. This concept is demonstrated on an industrial board machine. In particular, the propagation path of an oscillation due to valve stiction within multiple control loops in the drying section of the machine is studied. Finally, the results are discussed and evaluated.     

基于因果规则的电力营销系统故障定位算法

现有的故障定位算法无法有效地应用于带有负载均衡机制的因果关系频繁变动的复杂系统。为此,本文提出一种基于因果规则的故障定位算法（CRFLA）。首先利用改进的因果关联兴趣度度量方法自适应地学习出故障和事件之间因果规则,然后根据得到的因果规则中故障原因集对已发生事件集的影响程度进行根因推断。该方法考虑了因果关系的同时无需明确具体的因果网络结构,并且能够灵活地结合先验知识。利用电力营销系统中真实生产环境产生的数据进行故障定位,实验结果表明,CRFLA优于传统的方法,能够迅速、有效地定位故障根因。     

Leveraging on causal knowledge for enhancing the root cause analysis of equipment spot inspection failures

Causal correlation data over the equipment spot-inspection operation and maintenance (O&M) records and fault investigation sheets potentially reflect the state related to the causal effect of equipment failures. Various factors influence equipment failures, making it difficult to effectively analyze the main cause of the problems. Mining and leveraging these causal data from the equipment spot inspection records will undoubtedly significantly improve the root cause analysis of the fault in the O&M system. Hence, this paper introduces causal knowledge in equipment fault O&M for the first time and proposes to exploit causal knowledge for enhancing root cause analysis of equipment spot inspection failures. Specifically, an equipment fault O&M knowledge graph with causal knowledge called CausalKG is constructed to provide knowledge support for the causal analysis of faults. That is, CausalKG consists of spot-inspection knowledge graph (SIKG) and causal relationship knowledge (CRK) in equipment fault O&M. Further, a CausalKG-ALBERT knowledge reasoning model is designed. The model transforms CausalKG into network embeddings based on relational graph convolutional networks. In turn, it combines the Q&A mechanism of the language model ALBERT to mine the root cause knowledge of equipment failures. The case study confirms that incorporating the CRK is more effective than directly using the SIKG for causality reasoning; The model can fully use causal relationship knowledge to enhance the reliability of root cause analysis. This method is valuable to help engineers strengthen their causal analysis capabilities in preventive equipment maintenance.     

Simplified Granger causality map for data-driven root cause diagnosis of process disturbances

Root cause diagnosis is an important step in process monitoring, which aims to identify the sources of process disturbances. The primary challenge is that process disturbances propagate between different operating units because of the flow of material and information. Data-driven causality analysis techniques, such as Granger causality (GC) test, have been widely adopted to construct process causal maps for root cause diagnosis. However, the generated causal map is over-complicated and difficult to interpret because of the existence of process loops and the violation of statistical assumptions. In this work, a two-step procedure is proposed to solve this problem. First, a causal map is built by adopting the conditional GC analysis, which is viewed as a graph in the next step. In this graph, each vertex corresponds to a process variable under investigation, while the weight of the edge connecting two vertices is the F-value calculated by conditional GC. This graph is then simplified by computing its maximum spanning tree. Thus, the results of the causality analysis are transformed into a directed acyclic graph, which eliminates all loops, highlights the root cause variable, and facilitates the diagnosis. The feasibility of this method is illustrated with the application to the Tennessee Eastman benchmark process. In the investigated case studies, the proposed method outperforms the conditional GC test and provides an easy way to identify the root cause of process disturbances.     

工业控制系统的信息物理因果流模型

工业控制系统向智能控制的发展随着人工智能的因果革命也应该进行因果建模的思考。通过提出工业过程控制系统的单、多层信息物理结构,引入信息、物理因果流,建立过程控制系统的信息物理因果流模型,为依据因果关系设计工业控制系统、分析控制运行机制、实施系统故障预测和监控等提供理论基础和描述框架。     

A practical method for identifying the propagation path of plant-wide disturbances

Disturbances in a continuous process often travel along the product stream and affect the performance of the process. To isolate the root cause and find the path along which the disturbance propagates requires an understanding of the cause-and-effect relationships between multiple variables. One way of identifying these relationships is through the time delays between process variables. A practical and robust approach is proposed that uses the cross-correlation function to estimate the time delay between process measurements and derives a qualitative model of the propagation path in the form of a causal map. The approach was applied to an industrial case study of a process affected by a plant-wide disturbance and was able to decide between two alternative root cause explanations. It was also applied successfully to a process with a recycle. The advantages of the proposed method are its ease of implementation and automation as well as simple interpretation of the resulting causal map.     

Information transfer methods in causality analysis of process variables with an industrial application

This paper studies mutual information and transfer entropy for detection of cause and effect relationships between industrial process variables. Mutual information quantifies the amount of dependency between process variables, while transfer entropy detects the direction of information flow between the variables. The paper overviews the existing definition and limitations of these two quantities and proposes an algorithm based on combining and extending these two quantities for more reliable identification of causal relationship between process variables. Detection of causal relationships between plant variables is useful for diagnosis of the root cause of a distributed fault in the process. It also helps predicting the effect variables. The proposed method is illustrated through an industrial case study.     

SFEP文本因果关系提取及其与SFN转化研究

为将系统故障演化过程(system fault evolution process，SFEP)的文本描述转化为空间故障网络(space fault network，SFN)结构，用于故障分析，本文提出SFEP文本因果关系提取方法，及其与SFN基本结构的转化方法。首先给出SFEP中事件的几种典型因果关系。随后提出因果关系与SFN基本结构的转化流程。本文方法围绕着关键字和因果关系组模式展开，通过模型的不断学习补充和丰富关键字和组模式。最终使方法具备将SFEP文本转化为SFN结构的能力。以飞机起落架故障发生过程文本为例进行了应用，实验结果表明该方法可用于SFEP文本中的因果关系分析，并得到了理想的SFN。完善的关键字和组模式有利于使用计算机智能处理SFEP的SFN。     

Fault tolerant control for a class of nonlinear system with actuator faults

In this work, a fault tolerant control scheme is proposed for a class of nonlinear system with actuator faults. In this fault tolerant control strategy, an estimator is designed to estimate both the system states and the fault signal simultaneously. Based on these estimations, the control law is constructed to achieve the fault tolerant control for the nonlinear system considered. It is shown that the estimation error and the system state can be guaranteed to be bounded. The obtained theoretic results have been verified through the simulation examples on the three‐tank system.     

从结构推断到根因识别——工业过程故障根因诊断研究综述

由于现代工业过程的复杂结构,变量间普遍存在紧密耦合,故障往往在变量间广泛传播,为过程运维带来挑战.针对该问题,工业根因诊断(industrial root cause diagnosis, IRCD)技术应运而生,其从异常变量中确定故障根因,便于针对性故障处理. IRCD包含两个主要步骤:结构推断和根因识别.前者建立变量间的信息传递结构;后者根据传递结构定位根因.然而,现有IRCD综述多侧重于结构推断,未对根因识别步骤进行调研,且未建立起各类IRCD模型与过程特性间的系统关联.为此,从结构推断和根因识别两个层级展开IRCD的研究综述.首先,依据推断准则的异同,归纳4类经典结构推断模型;其次,考虑到过程的高维度、非线性、非平稳性质以及机理知识的效用,对结构推断模型的变种及适用场景进行梳理;随后,对根因识别方法进行归类,包括纯数据驱动、知识与数据融合驱动的范式,涵盖6类典型方法,并分析它们的优势与不足;最后,讨论IRCD技术中存在的挑战,并给出未来研究方向,为后续研究提供参考.     

A novel robust data reconciliation method for industrial processes

Data reconciliation has played a significant role in rectifying process data which can meet the conservation laws in industrial processes. Generally, the actual measurements are often easily contaminated by different gross errors. Thus, it is essential to build robust data reconciliation methods to alleviate the impact of gross errors and provide accurate data. In this paper, a novel robust estimator is proposed to improve the robustness of data reconciliation method, which is based on a new robust estimation function. First, the main robust properties are analyzed with its objective and influence functions for the proposed robust estimator. Then, the effectiveness of the new robust data reconciliation method is demonstrated on a linear numerical case and a nonlinear example. Moreover, it is further used to a practical industrial evaporation production process, which also demonstrates that the process data can be better reconciled with the proposed robust estimator.     

面向故障间格兰杰因果发现的霍克斯过程研究

现有因果关系建模方法应用于故障事件序列时,难以有效引入因果先验,使得算法结果过于稠密,同时在稀疏、时间精度低的数据上因果关系可靠性较差。将不同故障类型事件的因果关系建模为基于霍克斯过程的格兰杰因果关系,提出一种面向故障序列的格兰杰因果发现的霍克斯过程模型。将霍克斯过程拓展到离散时间域,解决低时间精度数据的建模问题,并通过构造基于贝叶斯信息准则的目标函数,保证因果结构稀疏性,进而利用基于EM算法与爬山法的迭代优化算法引入因果先验,提高模型的可靠性。实验结果表明,该方法在由不同参数生成的模拟数据上均表现突出,且在两个通信网络的真实数据集中,F1评分相比ADM4、MLE-SGL、TSSO和PCMCI算法提升15.18%以上。而通过引入根因标注和因果依赖性先验,算法的F1评分进一步提升22.43%以上,验证了引入先验的有效性。     

工业过程故障根源诊断与传播路径识别技术综述EI北大核心CSCD

故障根源诊断与传播路径识别是故障诊断框架下的关键核心问题,是保障工业过程安全生产及获得可靠产品质量的有效手段,是当前过程控制领域的研究热点.该技术的研究不仅丰富了故障诊断理论,而且对故障诊断技术在工程中的推广与应用具有重要意义.阐述了基于知识、数据及知识与数据联合驱动的故障根源诊断与传播路径识别方法的基本思想、适用条件和优劣特点,分类概述了相关方法的研究现状.探讨了该领域亟待解决的问题及未来的发展方向,包括:1)“三个维度”视角下的工业过程故障根源诊断与传播路径识别;2)基于制造大数据分析与因果关系挖掘的工业过程质量精准追溯;3)面向传播、耦合、多重并发特性的工业过程复合故障分布式诊断;4)基于多源异构动态信息融合的工业过程异常工况时空追溯可视化.     

Causal connectivity of evolved neural networks during behavior

To show how causal interactions in neural dynamics are modulated by behavior, it is valuable to analyze these interactions without perturbing or lesioning the neural mechanism. This paper proposes a method, based on a graph-theoretic extension of vector autoregressive modeling and 'Granger causality,' for characterizing causal interactions generated within intact neural mechanisms. This method, called 'causal connectivity analysis' is illustrated via model neural networks optimized for controlling target fixation in a simulated head-eye system, in which the structure of the environment can be experimentally varied. Causal connectivity analysis of this model yields novel insights into neural mechanisms underlying sensorimotor coordination. In contrast to networks supporting comparatively simple behavior, networks supporting rich adaptive behavior show a higher density of causal interactions, as well as a stronger causal flow from sensory inputs to motor outputs. They also show different arrangements of 'causal sources' and 'causal sinks': nodes that differentially affect, or are affected by, the remainder of the network. Finally, analysis of causal connectivity can predict the functional consequences of network lesions. These results suggest that causal connectivity analysis may have useful applications in the analysis of neural dynamics.     

Design and analysis of a fault isolation scheme for a class of uncertain nonlinear systems

This paper presents a unified fault isolation scheme for handling both process faults and sensor faults in a class of uncertain nonlinear systems. The proposed fault diagnosis architecture consists of a fault detection estimator and a bank of isolation estimators, each corresponding to a particular fault type. The design of the fault isolation decision scheme is based on the derivation of appropriate adaptive thresholds for each fault isolation estimator. Fault isolability conditions characterizing the class of process faults and sensor faults that are isolable by the proposed scheme are derived. A rigorous isolability analysis is presented via the use of the so-called fault mismatch functions, which are defined between pairs of possible faults. A simulation example is used to illustrate the proposed fault isolation scheme.     

Empirical Analysis of Software Fault Content and Fault Proneness Using Bayesian Methods

We present a methodology for Bayesian analysis of software quality. We cast our research in the broader context of constructing a causal framework that can include process, product, and other diverse sources of information regarding fault introduction during the software development process. In this paper, we discuss the aspect of relating internal product metrics to external quality metrics. Specifically, we build a Bayesian network (BN) model to relate object-oriented software metrics to software fault content and fault proneness. Assuming that the relationship can be described as a generalized linear model, we derive parametric functional forms for the target node conditional distributions in the BN. These functional forms are shown to be able to represent linear, Poisson, and binomial logistic regression. The models are empirically evaluated using a public domain data set from a software subsystem. The results show that our approach produces statistically significant estimations and that our overall modeling method performs no worse than existing techniques.     

Locally Linear Back-propagation Based Contribution for Nonlinear Process Fault Diagnosis

This paper proposes a novel locally linear back-propagation based contribution (LLBBC) for nonlinear process fault diagnosis. As a method based on the deep learning model of auto-encoder (AE), LLBBC can deal with the fault diagnosis problem through extracting nonlinear features. When the on-line fault diagnosis task is in progress, a locally linear model is firstly built at the current fault sample. According to the basic idea of reconstruction based contribution (RBC), the propagation of fault information is described by using back-propagation (BP) algorithm. Then, a contribution index is established to measure the correlation between the variable and the fault, and the final diagnosis result is obtained by searching variables with large contributions. The smearing effect, which is an important factor affecting the performance of fault diagnosis, can be suppressed as well, and the theoretical analysis reveals that the correct diagnosis can be guaranteed by LLBBC. Finally, the feasibility and effectiveness of the proposed method are verified through a nonlinear numerical example and the Tennessee Eastman benchmark process.