We hear the term "this is so non-linear" or the term "these are really
nonlinear effects" progressively more and more frequently now-a-days. While,
this is often a statement used to mean "I give up" or "I simply don't know what
is going on", more often than not, it is really an expression to suggest that
the system in question is behaving in a complicated fashion.
This class is a
modern introduction to the analysis, stability theory and control of nonlinear
systems. The class starts with an introduction to the qualitative theory of
nonlinear systems, with special attention being paid to the development of
intuition on 2-D systems. This is followed by the development of the
mathematical background necessary to study nonlinear systems.
We develop in
detail the stability theory for nonlinear systems, so called Lyapunov theory
with its application to many problems including stabilization, the Lur'e
problem, adaptive control problems.
We finally develop the geometric theory
of nonlinear control, including linearization by state feedback for SISO and
MIMO systems, zero dynamics of nonlinear systems, decoupling by state feedback.
We discuss observers for nonlinear systems and rudiments of constructive
controllability and observability.
This is a busy and sophisticated course,
but how much one gets out of the course is primarily a function of what one puts
into the course both in terms of supplementary reading and contemplation.
Enjoy!
Office: 284 HMMB
Office Hours: Monday Wednesday 3:00pm-4:00pm
Email: mailto:sastry@eecs.berkeley.edu
URL: http://robotics.eecs.berkeley.edu/~sastry
| Homework | 30% |
| Midterm (1 Day Take Home) | 30% |
| Final Exam (1 Day Take Home) | 40% |
Prerequisites
Some introduction to linear systems theory and mathematical analysis is desirable. Thus, for example a course like EECS 221A and Math 104A is desirable. However, sophistication of thought is more important than formal pre-requisites
S. S. Sastry, "Nonlinear Systems: Analysis, Stability and Control," Springer Verlag, 1999.
2007 Course Outline, Additional References.
Problem
Set 1 , Issued January 18th, due February 1st, 2007. Typo on Line 1 of
Problem 4, the invariant set should have radius sqrt lambda rather than 1.
Solution Set will be posted on February 4th 2007.
Problem
Set 2 , Issued February 1st, 2007, due February 15th, 2007.
Problem
Set 3 , Issued February 15th, 2007, originally due March 1, 2007: now due
March 8th.
Problem
Set 4 , Issued March 6th, 2007, due March 22nd, 2007.