Model predictive control of phthalic anhydride synthesis in a fixed-bed catalytic reactor via machine learning modeling

Wu, Zhe; Tran, Anh; Ren, Yi; Barnes, Cory S.; Chen, Scarlett; Christofides, Panagiotis D.

doi:10.1016/j.cherd.2019.02.016

Cited by 25 publications

(16 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Specifically, homogeneous ensemble regression models are derived from the ensemble learning method if a single base learning algorithm is used, while heterogeneous models are produced in the case of multiple learning algorithms. The reasons that ensemble regression models are able to improve the prediction performance are as follows . First, a single RNN model that achieves a desired training accuracy may perform poorly in the region that lacks sufficient training data, while ensemble methods can reduce the risk of relying on a single flawed model by aggregating all candidate models.…”

Section: Recurrent Neural Networkmentioning

confidence: 99%

“…Compared to a single model prediction, ensemble learning has demonstrated benefits in robustness and accuracy in solving classification or regression problems . In Reference ensemble learning‐based MPC has proven to be successful in regulating product yield for an industrial‐scale fixed‐bed reactor with a highly exothermic reaction. Additionally, in References and different ensemble learning methods were introduced to improve model prediction performance of neural network models for a batch polymerization reactor.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Machine learning‐based predictive control of nonlinear processes. Part I: Theory

et al. 2019

Self Cite

View full text Add to dashboard Cite

This article focuses on the design of model predictive control (MPC) systems for nonlinear processes that utilize an ensemble of recurrent neural network (RNN) models to predict nonlinear dynamics. Specifically, RNN models are initially developed based on a data set generated from extensive open-loop simulations within a desired process operation region to capture process dynamics with a sufficiently small modeling error between the RNN model and the actual nonlinear process model. Subsequently, Lyapunov-based MPC (LMPC) that utilizes RNN models as the prediction model is developed to achieve closed-loop state boundedness and convergence to the origin. Additionally, machine learning ensemble regression modeling tools are employed in the formulation of LMPC to improve prediction accuracy of RNN models and overall closed-loop performance while parallel computing is utilized to reduce computation time. Computational implementation of the method and application to a chemical reactor example is discussed in the second article of this series.

show abstract

Section: Recurrent Neural Networkmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Machine learning‐based predictive control of nonlinear processes. Part I: Theory

et al. 2019

Self Cite

View full text Add to dashboard Cite

show abstract

“…The training process of this model can be either adaptive or based on offline learning. Adaptive learning improves accuracy by deploying ML in real-time, and MPC model coefficients change online in real-time [197][198][199] while, in offline modeling, the coefficient and structure of the model are developed ahead of time using offline statistical data [192,[200][201][202].…”

Section: Ai and Mpc Integrationmentioning

confidence: 99%

Model Predictive Control of Internal Combustion Engines: A Review and Future Directions

et al. 2021

View full text Add to dashboard Cite

An internal combustion engine (ICE) is a highly nonlinear dynamic and complex engineering system whose operation is constrained by operational limits, including emissions, noise, peak in-cylinder pressure, combustion stability, and actuator constraints. To optimize today’s ICEs, seven to ten control actuators and 10–20 feedback sensors are often used, depending on the engine applications and target emission regulations. This requires extensive engine experimentation to calibrate the engine control module (ECM), which is both cumbersome and costly. Despite these efforts, optimal operation, particularly during engine transients and to meet real driving emission (RDE) targets for broad engine speed and load conditions, has still not been obtained. Methods of model predictive control (MPC) have shown promising results for real-time multi-objective optimal control of constrained multi-variable nonlinear systems, including ICEs. This paper reviews the application of MPC for ICEs and analyzes the recent developments in MPC that can be utilized in ECMs. ICE control and calibration can be enhanced by taking advantage of the recent developments in the field of Artificial Intelligence (AI) in applying Machine Learning (ML) to large-scale engine data. Recent developments in the field of ML-MPC are investigated, and promising methods for ICE control applications are identified in this paper.

show abstract

“…Ensemble learning is a machine learning process that combines multiple learning algorithms to obtain better prediction performance ( [26]) in terms of reduced variability and improved generalization performance. The reasons that ensemble regression models are able to improve prediction performance are given in [10,26]. In this work, homogeneous ensemble regression models are constructed following the RNN learning algorithm in Section 3.1 and a k-fold cross validation.…”

Section: Ensemble Learningmentioning

confidence: 99%

“…Additionally, a well-conditioned polynomial nonlinear state-space model has been used in [8] to approximate the actual (first-principles) nonlinear process model (normally substituting of the actual nonlinear process) for EMPC. Given the recent interest in machine learning modeling and particularly its usage in solving nonlinear regression problems ( [9,10]), in this work, we will derive a nonlinear data-driven process model for EMPC by taking advantage of machine learning techniques.…”

Section: Introductionmentioning

confidence: 99%

Economic Machine-Learning-Based Predictive Control of Nonlinear Systems

Christofides

2019

Mathematics

Self Cite

View full text Add to dashboard Cite

In this work, a Lyapunov-based economic model predictive control (LEMPC) method is developed to address economic optimality and closed-loop stability of nonlinear systems using machine learning-based models to make predictions. Specifically, an ensemble of recurrent neural network (RNN) models via a k-fold cross validation is first developed to capture process dynamics in an operating region. Then, the LEMPC using an RNN ensemble is designed to maintain the closed-loop state in a stability region and optimize process economic benefits simultaneously. Parallel computing is employed to improve computational efficiency of real-time implementation of LEMPC with an RNN ensemble. The proposed machine-learning-based LEMPC method is demonstrated using a nonlinear chemical process example.

show abstract

Model predictive control of phthalic anhydride synthesis in a fixed-bed catalytic reactor via machine learning modeling

Cited by 25 publications

References 14 publications

Machine learning‐based predictive control of nonlinear processes. Part I: Theory

Machine learning‐based predictive control of nonlinear processes. Part I: Theory

Model Predictive Control of Internal Combustion Engines: A Review and Future Directions

Economic Machine-Learning-Based Predictive Control of Nonlinear Systems

Contact Info

Product

Resources

About