Process Control

Data-driven stabilizing model predictive control of nonlinear systems

This work will focus on computational aspects of using data-driven approaches to obtain stabilizing MPC formulations. The approaches will include use of physics inspired NNs, recurrent NNs, as well as nonlinear extensions of purely trajectory based descriptions.

Integrated design and control of batch heat exchanger networks

Energy integration in batch processes is tricker, as compared to their continuous counterparts, due to additional time constraints. Previous work has shown that sequential design and control can lead to suboptimal performance during operation. This project aims at developing a simultaneous design and control framework for such networks. 

The project is simulation-based and experience with Matlab/Python will be beneficial.

Dual Adaptive and Predictive Control of Nonlinear and Distributed Systems

Model predictive control (MPC) is the most widely used multivariable control scheme in industrial control. Success of any MPC scheme critically depends on prediction models. The closed loop performance of an MPC scheme can deteriorate over a period of time if the prediction model is not updated to account for the changing operating conditions. The adaptive (self-learning) control schemes that solve both identification and control problems simultaneously provide an attractive option to alleviate this problem.

Online Optimizing Control of Nonlinear Processes using Machine Learning Techniques

Process industry is moving towards use of nonlinear dynamic models in advanced process control solutions such as fault tolerant control, process monitoring and online real time optimization. Development of control relevant models that capture system behavior over a wide range is always a challenge in chemical processes because of tight mass and energy integrations through recycle

Distributed control architecture synthesis

Control of integrated networks is challenging due to strong interactions between variables (limiting performance of decentralized controllers) and large system size (difficult design of a centralized controller). In this context, distributed controllers pose an optimal architecture with reduced system size and inclusion of key interactions. A key question is how to decompose an integrated system into distributed architecture. We address this problem via structural analysis. Specifically, we abstract the control system into an equivalent graph.

Design aspects of energy-integrated batch distillation

Distillation is one of the most commonly used as well as the most significant contributor of energy in chemical processing complex. Energy integration can improve the sustainability of the process by reducing utility requirement in batch distillation. However, operation of such columns is challenging. Traditionally, design of such systems is pursued without giving any consideration for operation. In the light of this, this project aims as developing a design framework for such distillation columns to address operational challenges.