Optimal Control and Estimation
MAE 546

Robert Stengel

Princeton University
School of Engineering and Applied Science
Department of Mechanical and Aerospace Engineering
Spring 2018
Tuesday and Thursday, 3-4:20 pm
Engineering Quadrangle

Designing control logic that commands a dynamic system to maximize performance while minimizing cost is a common objective in such diverse fields as economics, medicine, robotics, chemical process control, and vehicle engineering. This course presents mathematical foundations and numerical methods for optimal control of these systems. The course explores conditions for deterministic optimality of nonlinear systems, effects of state and control constraints, singular control, parametric and gradient-based optimization, and linear, neighboring-optimal feedback control. Linear-quadratic (LQ) control is examined via time-domain and frequency-domain analyses.

Disturbances and measurement error can affect the quality of optimal control. These factors are accounted for by including optimal linear and nonlinear state estimation in the design process. The Kalman filter, Extended Kalman Filter, particle filters, and linear-quadratic-Gaussian (LQG) regulator present underlying structures for this discussion. Model-referenced control systems that adapt to changing systems parameters are introduced, and robustness of control in the presence of system parameter uncertainty is investigated.

MAE 546 Syllabus

2017 Seminar Slides

Titles of Previous Term Papers


The primary reference for this course is OPTIMAL CONTROL AND ESTIMATION, R. F. Stengel, Dover Publications, 1994. Additional books that may be useful for reference include:


http://www.princeton.edu/~stengel/MAE546.html
key words: optimization, optimal control, probability theory, statistics, optimal state estimation, control systems, nonlinear control, adaptive control.
Last updated February 4, 2018.
Copyright 2018 by Robert F. Stengel. All rights reserved.