Wiener-Neural-Network-Based Modeling and Validation of Generalized Predictive Control on a Laboratory-Scale Batch...

Area of Science:

• Process Control and Automation
• Artificial Intelligence in Chemical Engineering
• Nonlinear System Modeling

Background:

• Batch reactors are critical in chemical industries but present significant modeling and control challenges due to their inherent nonlinearity.
• Accurate prediction and tracking of temperature profiles are essential for optimizing processes like reactant mixing and waste treatment.
• Traditional control methods often struggle with the complex dynamics of these systems.

Purpose of the Study:

• To develop a Wiener neural network (WNN) model for predicting and tracking the temperature profile of a batch reactor.
• To design and validate a generalized predictive controller (GPC) utilizing the WNN model for arbitrary temperature profile tracking.
• To demonstrate the efficacy of WNN-based control for nonlinear batch reactor systems.

Main Methods:

• Experimental modeling of a batch reactor using input-output data.
• Training a Wiener neural network (WNN) to learn the nonlinear temperature dynamics from the dataset.
• Designing a generalized predictive controller (GPC) based on the trained WNN model for robust temperature tracking.

Main Results:

• The Wiener neural network (WNN) successfully modeled the nonlinear temperature profile of the batch reactor.
• The developed generalized predictive controller (GPC) demonstrated effective tracking of arbitrary temperature profiles.
• The WNN-based GPC provided a robust solution for controlling complex batch reactor systems.

Conclusions:

• Wiener neural networks offer a powerful tool for modeling and controlling nonlinear batch reactor systems.
• The WNN-based generalized predictive controller (GPC) is validated as an effective method for precise temperature profile tracking.
• This approach enhances process control capabilities in industries utilizing batch reactors.