Intelligent Control for Flexible Manufacturing Systems: Data-Driven Modeling, Control, and Energy Management

Waseem, Muhammad

Intelligent Control for Flexible Manufacturing Systems: Data-Driven Modeling, Control, and Energy Management 11 views

Author

Waseem, Muhammad, Mechanical and Aerospace Engineering - School of Engineering and Applied Science, University of Virginia 0000-0002-7609-3739

Advisors

Chang, Qing , EN-Mech & Aero Engr Dept , University of Virginia

Abstract

This dissertation advances smart manufacturing by developing intelligent, data driven solutions for mobile robot-assisted multiproduct flexible manufacturing systems (FMS), structured around three core pillars: modeling, control, and energy management. First, the research addresses system modeling by developing a foundational dynamic mathematical model for a multi-product FMS that captures inherent uncertainties. To overcome the computational expense of high-fidelity Digital Twins (DTs), this dissertation proposes efficient Digital Twin Surrogates. These data-driven approximations enable rapid performance prediction and decision support without the overhead of traditional simulation. Second, leveraging these models, the dissertation explores reinforcement learning (RL)-based control strategies, progressing from centralized control to a sophisticated decentralized multi-agent RL (MARL) framework that improves decision-making by embedding system properties like permanent production loss (PPL). Additionally, a novel Nash-Multi-Agent Deep Deterministic Policy Gradient (Nash-MADDPG) method is developed to handle complex systems. Finally, to address sustainable operations, a joint model of the FMS and a microgrid is developed to co-optimize production with energy storage system (ESS) degradation. Furthermore, a MARL-based control framework is designed for an energy-intensive conveyor belt dryer, enhancing product quality while significantly reducing energy consumption. By synergistically combining advanced modeling, intelligent control, and holistic energy management, this dissertation delivers scalable and resilient solutions for the dynamic challenges of modern smart manufacturing.

Degree

PHD (Doctor of Philosophy)

Keywords

Intelligent control; flexible manufacturing; energy management; Data driven modeling; Digital twin; Smart manufacturing

Language

English

Rights

Issued Date

2025-11-29

Suggested Citation

Waseem, Muhammad. Intelligent Control for Flexible Manufacturing Systems: Data-Driven Modeling, Control, and Energy Management. University of Virginia, Mechanical and Aerospace Engineering - School of Engineering and Applied Science, PHD (Doctor of Philosophy), 2025-11-29, https://doi.org/10.18130/dfw7-5968.