Back to category

ARF-003
Automation, robotics, sensors - 2019

''Robust'' system control

OBJECTIVE :

This course will introduce learners to a variety of modern, robust control techniques. The techniques covered will include:

  • H∞, µ and LMI techniques
  • The LQ, LQG and H2 controls and their variants in the presence of uncertainties
  • Recent non-smooth optimization techniques
  • Control analysis and validation techniques.

The course focuses on resolution strategies and methodologies, the practical application of tools and the comparative assessment of techniques. Case studies from aeronautics and space systems help to improve understanding of methods and tools.

COURSE DURATION AND TIMETABLE :

The course lasts 5 days (33 hours).

It combines lectures and numerous workstation-based tutorials.

GENERAL APPROACH :

Understanding the concepts of robustness and model uncertainties, formulating complex specifications, translating frequency or time performance and solving difficult practical cases.

This course will introduce learners to a variety of robust control techniques. These techniques combine the "standard" and "modern" approaches and offer the engineer a flexible and efficient framework for dealing with all kinds of specifications.

In particular, the internship focuses on the latest techniques of non-smooth optimization that have met with significant success at the industrial level. Also discussed are the choice and complexity of the control, multiple specifications and dealing with several models.

PREREQUISITE :

Course level: Specialization

Knowledge of standard frequency methods.

COURSE DIRECTOR(S) :

Pierre APKARIAN:

Research Director at ONERA, Toulouse

Daniel ALAZARD:

Professor at ISAE-SUPAERO (the Institut Supérieur de l’Aéronautique et de l’Espace or National Higher French Institute of Aeronautics and Space).

CONTENT :

  • Problems and motivations
    • Concept of robustness of a corrector, dynamic and parametric uncertainty, examples
    • Performance/robustness compromise
  • Linear automatic reminders
    • Transfer function, state, standards
    • Internal stability and performance
    • Initiation to C.A.O. Matlab and Robust Control modules
  • Standard H∞ problem (focus)
    • Generic standard form of the control
    • H∞ formulation, robust stability, low gain
    • Shaping the frequency response (loop shaping)
    • Resolution algorithms
    • Methodology, specifications, choice of weighting filters
    • Difficulties, traps and remedies
    • Examples of H∞ synthesis of a corrector
  • LQG methods and variants (focus)
    • LQ, LQG, LQG/LTR controls
    • LQG synthesis with parametric model uncertainties
  • Implementation/reduction questions (overview)
    • Implantation of correctors, structures
    • Observer structure/state return
    • Interpolation of compensators
  • LMI methods (overview)
  • Non-smooth optimization techniques (focus)
    • Non-smooth optimization techniques and the HINFSTRUCT function
    • Structured control/compensators (loops and multiple components, interconnections)
    • Synthesis of fixed architecture controls
    • Multiple specifications and models
    • Practical advantages, applications
  • Future plans, conclusions and discussion

 

English version programmable on demand

WHEN AND WHERE :

Scheduled in French:

TOULOUSE: 20 to 24 May 2019

 

For the English realization, please, consult us.

COURSE FEES :

€2,160 excluding tax (20% VAT)

See general terms

Dowload pdf
Print
Download registration form
Contact us