Real-Time Industrial Process Fault Diagnosis Based on Time Delayed Mutual Information Analysis

Abstract

Causal relations among variables may change significantly due to different control strategies and fault types. Off line-based knowledge is not adequate for fault diagnosis, and existing causal models obtained from data driven methods are mostly based on historical data only. However, variable correlation would not remain identical, and could be very different under certain industrial operation conditions. To deal with this problem, a fault diagnosis framework is proposed based on information solely extracted from process data. By this method, mutual information (MI) between each pair of variables is first calculated to obtain thresholds using historical data, as variable correlation under normal conditions is mostly contributed by random noises, which is often neglected in existing causal analysis models. Once a process deviation is detected, each pair of variables with mutual information beyond these thresholds are further investigated by time delayed mutual information (TDMI) analysis using current data, so as to determine the causal logic between them, which is represented as fault propagation paths, can be tracked all the way back to the root cause. The proposed method is first applied to a simulated process and the Tennessee Eastman process. The results show that the difference in variable correlation under diverse operation or control response conditions can be captured in real time, and fault propagation path can be objectively identified, together with the root cause. Then, the method has been successfully applied to a whole year data in an industrial process, which proves the feasibility of industrial application.

Transfer Entropy-Based Automated Fault Traversal and Root Cause Identification in Complex Nonlinear Industrial Processes

“…This leads to manipulation of the valve position to prevent the fault from cascading any further. From the fault path, X51 was correctly identified as the root cause variable, and this is in agreement with fault description in the literature . The causal map obtained using the Gaussian estimator is shown in Figure a.…”

“…From the fault path, X51 was correctly identified as the root cause variable, and this is in agreement with fault description in the literature. 45 The causal map obtained using the Gaussian estimator is shown in Figure 13a. Figure 13b shows the transfer entropy obtained along causal map edges.…”

A Review on Data-Driven Process Monitoring Methods: Characterization and Mining of Industrial Data

“…Considering the fault information in real time data, Ji et al proposed a real time fault diagnosis method based on time delayed mutual information. In their work, both historical data under normal operating conditions and real time data were employed for fault propagation analysis, therefore, the changes in variable correlation when a fault occured could be captured to obtain a more objective root cause diagnosis [213]. More research on fault propagation analysis using several other methods can be found, such as k nearest neighbor for a nonlinear multi-input, single-output process [214], and convergent cross-mapping for capturing nonlinear relationships with a low computational cost [215].…”

Industrial Process Diagnosis based on Information-based Features

“…This can be described directlyin an independent way -based on specific elements of knowledge but can be also described in terms of probabilities distributions (in discrete case or probability density functions (in the continuous case), which are provided by information models of the source of the signals. An introduction and more explained examples and case studies are available in [14], [15], and [16].…”