Adaptive and Predictive Control
Responsible Instructor:dr.ir. S. Baldi (Simone)
Contact Hours / Week x/x/x/x:0/0/3/0
Course Contents:Adaptive Control covers a set of techniques which provide a systematic approach for automatic adjustment of the controllers in real time, in order to achieve or to maintain a desired level of performance of the control system when the parameters of the plant dynamic model are unknown and/or change in time. Predictive control focuses on the optimization of a performance index with respect to some future control sequence, using predictions of the output signal based on a process model, coping with amplitude constraints on inputs, outputs and states. One-step ahead predictive controllers enjoy the nice feature that they can cope with the effect of parameter uncertainty upon the performance of the control system. The course presents a basic ground for analysis and design of adaptive and predictive control systems. After an introductory part and an initiation to parameter adaptation algorithms, Model Reference Adaptive Control (MRAC) and Self-Tuning Control (STC) schemes constitute the core part of the course. These techniques are based on one-step ahead predictive strategies, namely model reference and single-step control. Stability analysis in a deterministic environment and convergence analysis in a stochastic environment are both dealt with. Due to historical reasons, the Model Reference Adaptive control will be formulated in a deterministic setting, while the Self-tuning Control in a discrete-time stochastic setting. Multi-step ahead predictive strategies are finally introduced, with finite/infinite horizon predictive control, stability and robustness of predictive control. Hands-on experience is obtained by MATLAB exercises.
Study Goals:At the end of the course the student should be able to:
- Design, simulate, and implement parameter adaptation schemes;
- Design, simulate, and implement single-step ahead adaptive control schemes;
- Solve the finite and infinite horizon predictive control problem;
- Master the main analytical details in stability proofs of adaptive and predictive control schemes;
- Simulate adaptive and predictive control methodologies in Matlab;
- Discuss simulation results.
Education Method:Lectures and flippled classroom. Classic lectures will be accompanied by a couple of lectures devoted to solution of exercises and assignements. Here students will be invited to show their solutions to their peers.
Literature and Study Materials:1. Landau, Lozano, M;Saad, and Karimi, Adaptive Control: Algorithms, Analysis and Applications,
2nd edition, Springer-Verlag, 2011.
2. Ioannou and Fidan, Adaptive Control Tutorial, SIAM, 2006.
3. Mosca, Optimal, Predictive, and Adaptive Control, Prentice Hall, 1995.
Additional course material distributed during the course.
Assessment:Three homework assignments, each one accounting for 20% of the final grade. Final written exam, accounting for 40% of the final grade. The homework assignments will consist of both problems and programming exercises.
Final exam is individual. Assignments can be solved in groups of 2/3 persons. Students are invited to form the groups in the first week of the course (possibly first lecture). The composition of some groups can be revised to accommodate for different prior knowledge.
Remarks:Old course code: SC4060
Basic notions of linear systems theory, Lyapunov stability and stochastic processes. Basic MATLAB programming skills.