A hierarchical approach for causal modeling of process systems

Suresh, Resmi; Sivaram, Abhishek; Venkatasubramanian, Venkat

doi:10.1016/j.compchemeng.2018.12.017

Cited by 23 publications

(6 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The TE process model is a realistic simulation program of a chemical plant, which is widely accepted as a benchmark for control and monitoring studies [26]. The flow diagram of the process is described in [27,28], and the FORTRAN code of the process is available on the Internet. The process has two products from four reactants as shown in Equation ( 14):…”

Section: Tennessee Eastman Processmentioning

confidence: 99%

A Novel Dynamic Process Monitoring Algorithm: Dynamic Orthonormal Subspace Analysis

Hao

Lou

et al. 2023

Processes

View full text Add to dashboard Cite

Orthonormal subspace analysis (OSA) is proposed for handling the subspace decomposition issue and the principal component selection issue in traditional key performance indicator (KPI)-related process monitoring methods such as partial least squares (PLS) and canonical correlation analysis (CCA). However, it is not appropriate to apply the static OSA algorithm to a dynamic process since OSA pays no attention to the auto-correlation relationships in variables. Therefore, a novel dynamic OSA (DOSA) algorithm is proposed to capture the auto-correlative behavior of process variables on the basis of monitoring KPIs accurately. This study also discusses whether it is necessary to expand the dimension of both the process variables matrix and the KPI matrix in DOSA. The test results in a mathematical model and the Tennessee Eastman (TE) process show that DOSA can address the dynamic issue and retain the advantages of OSA.

show abstract

Section: Tennessee Eastman Processmentioning

confidence: 99%

A Novel Dynamic Process Monitoring Algorithm: Dynamic Orthonormal Subspace Analysis

Hao

Lou

et al. 2023

Processes

View full text Add to dashboard Cite

show abstract

“…Different approaches to describe control systems in causal models have been presented in e.g. [3] and [12], but typically involve additional modeling effort. The propagation strategy presented in this paper in Algorithm 1 correctly derives the steady-state development of variables in closed-loop systems given knowledge of the active control-loops.…”

Section: E Effect Matrixmentioning

confidence: 99%

“…Causal process models have been considered for rootcause analysis and process supervision for a long time [2], and nowadays, improved automated generation and validation techniques enable their application for a wide range of processes. [3] presents the generation of causal models in form of hierarchically structured directed graphs based on transfer entropy, while [4] uses a Bayesian model to identify optimal causal network structures for root-cause diagnosis. A method for semi-automated generation of causal functional models from existing documentation of plant topology in the form of P&ID diagrams is presented in [5].…”

Section: Introductionmentioning

confidence: 99%

A Causal Model-based Planner for the Reconfiguration of Continuous Processes

Reinartz

Enevoldsen

Galeazzi

et al. 2021

2021 European Control Conference (ECC)

View full text Add to dashboard Cite

This paper presents a planning framework for discrete planning of human operations in highly automated industrial plants, where manual and automatic control coexist. Based solely on qualitative knowledge of the controlled continuous process represented through signed directed graphs, the planner exploits a greedy algorithm to determine the optimal action for the human operator upon changes in the process' operating conditions or reconfiguration of the control system due to faulty conditions. The planner results in a decision support tool that instructs the human operator on the best course of action. The planning framework and the resulting planner are demonstrated on a quadruple-tank system.

show abstract

“…Causal models are good tools to create explainable systems that show which variables affect the final output variables, thus providing a cause‐and‐effect chain of transparency and mechanistic insights. Considerable work has gone into creating graphical causal models using data‐driven techniques 4,5 and deriving causal models from equations 6,7 . While such models analyze the data or the underlying model equations without the inference of the parameters, it often requires the use of a known model form to estimate the parameters from data 8 .…”

Section: Introductionmentioning

confidence: 99%

XAI‐MEG: Combining symbolic AI and machine learning to generate first‐principles models and causal explanations

Sivaram

Venkatasubramanian

2022

AIChE Journal

Self Cite

View full text Add to dashboard Cite

Current machine learning methods generally do not reveal any mechanistic insights or provide causal explanations for their decisions. While this may not be a big concern in typical computer vision, game playing, and recommendation systems, this is important for many problems in chemical engineering such as fault diagnosis, process control, and process safety analysis. To address these drawbacks, one needs to go beyond purely data‐driven machine learning techniques and incorporate the lessons learned from the expert systems era of artificial intelligence (AI), in the 1970s and 1980s. In this article, we present such a hybrid‐AI framework that demonstrates how symbolic AI techniques can be integrated with numeric AI‐based machine learning methods.

show abstract

A hierarchical approach for causal modeling of process systems

Cited by 23 publications

References 25 publications

A Novel Dynamic Process Monitoring Algorithm: Dynamic Orthonormal Subspace Analysis

A Novel Dynamic Process Monitoring Algorithm: Dynamic Orthonormal Subspace Analysis

A Causal Model-based Planner for the Reconfiguration of Continuous Processes

XAI‐MEG: Combining symbolic AI and machine learning to generate first‐principles models and causal explanations

Contact Info

Product

Resources

About