This course offers a comprehensive, advanced, and rigorous coverage of Model Predictive Control (MPC) and realtime dynamic optimization for complex, constrained engineered systems. Designed for postgraduate students, the curriculum bridges fundamental control theory with real-world computational implementation. Students will systematically master the mathematical foundations of constrained discrete-time dynamic optimization, receding horizon strategies, closed-loop nominal stability guarantees, and recursive feasibility. The course thoroughly addresses key practical deployment challenges, including numerical optimization formulations (Quadratic Programming and Sequential Quadratic Programming), real-time embedded solvers, state/output feedback estimation, state observer design, disturbance observer design, offset-free tracking, and robust tubebased/nonlinear MPC formulations. Highlighting multi-domain engineering applications across autonomous vehicles, robotic manipulators, aerial robots, and smart manufacturing systems, the course also introduces emerging frontiers such as data-driven control, Control Barrier Functions for safety-critical systems, and Physical AI integration. Through structured coding assignments, computational laboratories, and an integrated miniproject, students will acquire both deep theoretical insights and hands-on algorithm implementation skills required to design, tune, and deploy high-performance real-time predictive controllers.
Unlock more information by signing in with your HKUST account.