(147m) Control Advances for Dynamic Cyber-Physical Systems Incorporating Structural Simulation Methods and Quantum Computation

Manufacturing in the future will involve increasingly complex and diverse processes that will require implementation of current technologies in new ways, as well as the application of novel technologies, to improve safety, optimize performance, and operate processes in innovative and useful ways. My work investigates several of the considerations introduced by future manufacturing relating to the use of improved modeling, computational, and theoretic methods, with respect to process control systems.

Specifically, there are two related objectives in this research. The first topic studies improved modeling methods for use in control applications, as advanced model-based controllers may be able to utilize techniques like computational fluid dynamics (CFD) and finite element analysis (FEA) for improved operation from a cyber-physical perspective. This is studied through the simulation of two processes. The first is a steam methane reforming (SMR) reactor, which consists of catalyst-packed tubes which produce hydrogen gas from methane and water vapor. A second process called powder bed fusion (PBF), which creates objects from layers of melted powder, is also studied. In both simulations, the consequences of control actions are considered on the structure of the system (the tube wall for the SMR and the created object in PBF), and it is hypothesized that controllers with this explicit knowledge will be better equipped to handle safety, cybersecurity, and product quality concerns.

Secondly, given that computational complexity is always an issue with advanced control and modeling, this work also investigates the potential application of control algorithms on quantum computers. While significant challenges still exist for practical implementation of quantum computers, it is appealing because quantum computers have been theoretically shown to outperform classical computers in certain cases. Quantum computers introduce new phenomena such as superposition, entanglement, and probabilistic operation that will need to be considered for control applications. In this work, a control algorithm is designed for a circuit-based quantum computer framework and, while the algorithm is neither efficient nor practical, is used to demonstrate how certain phenomena will affect the operation of a controller. Also considered is a theoretical analysis of these phenomena, including the effects of rounding and nondeterminism. It is hypothesized that a study of control algorithms on quantum computers may help direct future research towards practical implementation of quantum computers in engineering applications, by guiding algorithm development and problem formulation.

Research Interests:

Control and Optimization

Cyber-physical Systems

Simulation Methods including Computational Fluid Dynamics and Finite Element Analysis

Quantum Computing Algorithms