Learning-based Adaptive Control: An Extremum Seeking Approach - Theory and Applications

Adaptive Control seeks for optimal action based on the current characteristics of a system and learning how the characteristics change along the way, opposed to a plain Robust Control approach that establishes a range of variation in the parameters and ensures the stability of the systems within those values. Two are the main approaches to the learning process: Model Based and Model Free being so that in the first case you have to precisely model the characteristics of a system and they have to remain stable. For Model Free Adaptive Control you start knowing nothing about the system and learning from action (inputs) and reaction (outputs). This will be the first book that offers a blended approach to Adaptive Control having some physical modeling but allowing the characteristics to evolve naturally over time maintaining the stability of the system. In the cases presented by the author there is remarkable gain in performance and adaptability of the systems by applying his findings. Includes a good number of Mechatronics Examples of the techniquesCompares and blends Model-free and Model-based learning algorithmsCoverage of fundamental concepts, state of the art research, necessary tools for modeling, and applications INDICE: 1. Survey 2. Mathematical tools 3. Results about stability and convergence of multi-variable extremum seeking (MES) theory (MES time invariant, MES time-variant) 4. Extremum Seeking-based Indirect Adaptive Nonlinear Control (might be split into two chapters) 5. Extremum Seeking-based System Identification and Model Reduction theory 6. Extremum Seeking -based MPC theory 7. Conclusion and open problems

• ISBN: 978-0-12-803136-0
• Editorial: Butterworth-Heinemann
• Encuadernacion: Rústica
• Páginas: 312
• Fecha Publicación: 01/08/2016
• Nº Volúmenes: 1
• Idioma: Inglés