Integrated Data-Driven Process Monitoring and Explicit Fault-Tolerant Multiparametric Control

Onel, Melis; Burnak, Baris; Pistikopoulos, Efstratios N.

doi:10.1021/acs.iecr.9b04226

Cited by 11 publications

(9 citation statements)

References 60 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Apart from that, multiparametric programming has been coupled with machine learning techniques for process monitoring, in a contribution by Onel et al (2019). In particular, support vector machine feature-based selection and modeling is used to identify potential faults that exist in a given process.…”

Section: Integration Of Machine Learning and Multiparametric Programmingmentioning

confidence: 99%

Multiparametric Programming in Process Systems Engineering: Recent Developments and Path Forward

et al. 2021

Self Cite

View full text Add to dashboard Cite

The inevitable presence of uncertain parameters in critical applications of process optimization can lead to undesirable or infeasible solutions. For this reason, optimization under parametric uncertainty was, and continues to be a core area of research within Process Systems Engineering. Multiparametric programming is a strategy that offers a holistic perspective for the solution of this class of mathematical programming problems. Specifically, multiparametric programming theory enables the derivation of the optimal solution as a function of the uncertain parameters, explicitly revealing the impact of uncertainty in optimal decision-making. By taking advantage of such a relationship, new breakthroughs in the solution of challenging formulations with uncertainty have been created. Apart from that, researchers have utilized multiparametric programming techniques to solve deterministic classes of problems, by treating specific elements of the optimization program as uncertain parameters. In the past years, there has been a significant number of publications in the literature involving multiparametric programming. The present review article covers recent theoretical, algorithmic, and application developments in multiparametric programming. Additionally, several areas for potential contributions in this field are discussed, highlighting the benefits of multiparametric programming in future research efforts.

show abstract

Section: Integration Of Machine Learning and Multiparametric Programmingmentioning

confidence: 99%

Multiparametric Programming in Process Systems Engineering: Recent Developments and Path Forward

et al. 2021

Self Cite

View full text Add to dashboard Cite

show abstract

“…A process network involving species i ∈ I, equipment j ∈ J, process streams s ∈ S, utilities u ∈ U, scheduling time intervals t S ∈ T S , and control time intervals t C ∈ T C . Onel et al [9][10][11] Simultaneous fault detection and diagnosis using support vector machines Process and maintenance scheduling Dedopoulos and Shah 12 Overall profitability maximization of multipurpose equipment Vassiliadis and Pistikopoulos 13 System effectiveness optimization considering environmental impact Cheung et al 14 Short-term preventive maintenance scheduling using mixed-integer linear programming Wiebe et al 15 Robust mixed-integer linear programming involving equipment condition degradation modeling Gordon et al 16 Multiobjective profit and safety optimization and ensemble support vector classification failure prediction Fault-aware control Modeling and control of mixed logical dynamical systems Axehill et al 22 Branch and bound algorithm for multiparametric mixed-integer quadratic programming problems…”

Section: Generalized Systemmentioning

confidence: 99%

Data‐driven prescriptive maintenance toward fault‐tolerant multiparametric control

Gordon

Pistikopoulos

2021

AIChE Journal

Self Cite

View full text Add to dashboard Cite

Prescriptive maintenance can improve system effectiveness and system safety via integrated production and maintenance optimization. However due to system disruptions there is potential for abnormal operations and an undesirable increased occurrence of process safety incidents. This research provides a multiparametric‐based framework for safety‐aware, maintenance‐aware, and disruption‐aware process control. It leverages ensemble classification via machine learning classifiers for fault detection, mixed‐integer nonlinear programming for integrated safety‐aware production and maintenance scheduling, as well as hybrid multiparametric model predictive control for fault‐tolerant setpoint tracking. The results show that the ensemble classifier outperforms the individual classifiers in terms of fault detection accuracy, sensitivity, and specificity. Furthermore, it is seen that the developed controllers are able to reconfigure the control actions based on process disruption information. The framework is illustrated with a chemical complex system, and a cooling water system. The approach can be used to help improve the safety and productivity of industrial processes.

show abstract

“…What is worse, those process monitoring methods with a single model may cause frequent false alarms since they may not model the diversified steady states and the rich state switching information. 42,43,44 Many machine learning-based methods have been proposed to tackle the nonstationary problem in industrial process monitoring, which can be mainly divided into three categories, 43 including adaptive approaches, 20 multimode modeling approaches, 21−23 and stationary analysis approaches. 17,34,35 The main idea of adaptive approaches is to capture the frequent changes along with time directions by continuously updating the model.…”

Section: Introductionmentioning

confidence: 99%

Multiple Condition Indicator Decoupling and Hierarchical Monitoring for Large-Scale Nonstationary Processes

Zhao

2022

Ind. Eng. Chem. Res.

View full text Add to dashboard Cite

Frequent changes of operation conditions widely exist in industrial processes, resulting in typical nonstationary characteristics. Especially for large-scale plant-wide industrial processes, a large number of facilities behave differently with the change of multiple operating conditions and present diverse nonstationarity. This leads to severe coupling characteristics and brings considerable challenges for process monitoring. In this work, a multiple condition indicator decoupling and hierarchical monitoring strategy is proposed to solve this problem. First, a stationary subspace analysis and elastic network (SSA-EN) method is proposed to extract the key variables corresponding to the local condition indicator. In this way, the large-scale plant-wide processes can be decoupled into multiple variable subsets for fine-grained descriptions. Then, the stationary and nonstationary features in each variable subset are divided by the SSA-based method, which provides a comprehensive process status identification. A stationary feature statistic is designed to reflect the global characteristics throughout the whole process. For the nonstationary features, a condition mode division method is developed to divide the process into several modes. Each mode can present the local characteristics described by the designed two local statistics using equilibrium relationships and the Gaussian mixture model. A real industrial process is adopted to verify the performance of the proposed method.

show abstract

Integrated Data-Driven Process Monitoring and Explicit Fault-Tolerant Multiparametric Control

Cited by 11 publications

References 60 publications

Multiparametric Programming in Process Systems Engineering: Recent Developments and Path Forward

Multiparametric Programming in Process Systems Engineering: Recent Developments and Path Forward

Data‐driven prescriptive maintenance toward fault‐tolerant multiparametric control

Multiple Condition Indicator Decoupling and Hierarchical Monitoring for Large-Scale Nonstationary Processes

Contact Info

Product

Resources

About