A systematic framework for the development and analysis of signed digraphs for chemical processes. 2. Control loops and flowsheet analysis
30 Aug 2003-Industrial & Engineering Chemistry Research (American Chemical Society)-Vol. 42, Iss: 20, pp 4811-4827
TL;DR: In this paper, the authors present a signed digraph (SDG) model for control loops and discuss a framework for application of graph-based approaches at a flowsheet level.
Abstract: The objectives of this part of the two part paper are (i) development of signed digraph (SDG) models for control loops and (ii) discussion of a framework for application of graph-based approaches at a flowsheet level. Further, two case studies are used to explain the methods developed in part 11 (Ind. Eng. Chem. Res. 2003, 42, in press) and this paper. The first case study (continuous stirred tank reactor case study) explains the basic concepts of the generate and test method for SDG analysis, generation of redundant equations using algebraic manipulation, and analysis of systems with a single control loop. Case study 2 (flash vaporizer case study) deals with different methods of generating redundant equations and the analysis of systems with multiple interacting control loops.
Citations
More filters
TL;DR: A data-driven method for identifying the direction of propagation of disturbances using historical process data using the application of transfer entropy, a method based on the conditional probability density functions that measures directionality of variation.
Abstract: In continuous chemical processes, variations of process variables usually travel along propagation paths in the direction of the control path and process flow. This paper describes a data-driven method for identifying the direction of propagation of disturbances using historical process data. The novel concept is the application of transfer entropy, a method based on the conditional probability density functions that measures directionality of variation. It is sensitive to directionality even in the absence of an observable time delay. Its performance is studied in detail and default settings for the parameters in the algorithm are derived so that it can be applied in a large scale setting. Two industrial case studies demonstrate the method
247 citations
Cites background from "A systematic framework for the deve..."
...A requirement for the identification, however, was the existence of a causal map in the form of a digraph that had to be derived from expert knowledge of the process....
[...]
TL;DR: This review aims at equipping researchers and industrial practitioners with structured knowledge on the state of machine learning applications in mineral processing with suggestions on data collection, technique comparison, industrial participation, cost-benefit analyses and the future of mineral engineering training.
155 citations
TL;DR: A direct causality detection approach suitable for both linear and nonlinear connections is described, based on an extension of the transfer entropy approach, and a direct transfer entropy (DTE) concept is proposed to detect whether there is a direct information flow pathway from one variable to another.
Abstract: The detection of direct causality, as opposed to indirect causality, is an important and challenging problem in root cause and hazard propagation analysis. Several methods provide effective solutions to this problem when linear relationships between variables are involved. For nonlinear relationships, currently only overall causality analysis can be conducted, but direct causality cannot be identified for such processes. In this paper, we describe a direct causality detection approach suitable for both linear and nonlinear connections. Based on an extension of the transfer entropy approach, a direct transfer entropy (DTE) concept is proposed to detect whether there is a direct information flow pathway from one variable to another. Especially, a differential direct transfer entropy concept is defined for continuous random variables, and a normalization method for the differential direct transfer entropy is presented to determine the connectivity strength of direct causality. The effectiveness of the proposed method is illustrated by several examples, including one experimental case study and one industrial case study.
155 citations
Additional excerpts
...TABLE II...
[...]
TL;DR: Two case studies are presented to illustrate SDG-based analysis of process flowsheets containing many units and control loops and it is shown that digraph-based steady-state analysis results in good diagnostic resolution.
135 citations
TL;DR: In this paper, the authors present a thematic review of the methods for integration of process design and control, and the evolution paths of these methods are described and the advantages and disadvantages of each method are explained.
Abstract: There is a large variety of methods in literature for process design and control, which can be classified into two main categories. The methods in the first category have a sequential approach in which, the control system is designed, only after the details of process design are decided. However, when process design is fixed, there is little room left for improving the control performance. Recognizing the interactions between process design and control, the methods in the second category integrate some control aspects into process design. With the aim of providing an exploration map and identifying the potential areas of further contributions, this paper presents a thematic review of the methods for integration of process design and control. The evolution paths of these methods are described and the advantages and disadvantages of each method are explained. The paper concludes with suggestions for future research activities.
122 citations
References
More filters
TL;DR: This three part series of papers is to provide a systematic and comparative study of various diagnostic methods from different perspectives and broadly classify fault diagnosis methods into three general categories and review them in three parts.
2,263 citations
TL;DR: In this paper, the mauvais fonctionnement and l'effet des erreurs se propagent par causalite d'une variable a l'autre, en satisfaisant finalement les contraintes du procede en regime permanent.
Abstract: Dans une analyse qualitative le mauvais fonctionnement et l'effet des erreurs se propagent par causalite d'une variable a l'autre, en satisfaisant finalement les contraintes du procede en regime permanent. Les contraintes des series d'equations algebriques, sont representees par des equations de confluence stationnaires. La causalite est representee par le graphe oriente signe etendu (Extended Signed Directed Graph: ESDG)
141 citations
TL;DR: A potential methodology to construct fault trees automatically is proposed in this paper, based on a component- by-component basis instead of a loop-by-loop or node-By-node basis, which avoids the tedious work of generating digraphs, transition tables, decision tables, and knowledge-based rules.
Abstract: Fault tree analysis (FTA) has been used in the chemical process industry (CPI) for systematic safety and reliability analysis during the past decades. Conventional manual construction of fault trees can be extremely time consuming and vulnerable to human error. A computer-aided fault tree synthesis methodology can be an initial step, or as an independent check to assist or supplement manual FTA. However, no entirely satisfactory algorithm has been published for fault tree synthesis, especially when control loops are encountered. A potential methodology to construct fault trees automatically is proposed in this paper. This algorithm works directly form the system block diagram, thus avoids the tedious work of generating digraphs, transition tables, decision tables, and knowledge-based rules. Mini cause-and-effect trees are used to model the cause and effect logics around each item of equipment. Control loops are treated by special cause-and-effect unit models — logical combinations of the unit models of their constituent components. Multiple or complex control loops can be easily taken into account by providing their corresponding cause-and-effect unit models. In particular, the fault tree construction algorithm presented here is based on a component-by-component basis instead of a loop-by-loop or node-by-node basis. The tree structure is much more concise and easier to read. An example is embedded in the description of the methodology for better understanding. Analysis shows that the fault tree generated here is equivalent to the published result.
33 citations
TL;DR: In this paper, an approach is proposed in which fault detection and diagnosis (FDD) tasks are distributed to separate FDD modules associated with each control system located throughout a plant.
31 citations
TL;DR: In this article, a distributed hybrid strategy for the isolation of faults and disturbances in the Tennessee Eastman process, which would build on existing structures for distributed control systems, so should be easy to implement, be cheap and be widely applicable.
14 citations