Course Title: Applied Optimal Control

EE6412
Lectures: 3 hr/week
Recitation: 2 hr/week
Evaluation: Quiz (20), Assignments (30), EndSem (25), Mini project (25)

Content of the course

1. Introduction to Optimization
2. Calculus of variations: Weak and strong extrema; First-order necessary
conditions for weak extrema; Hamiltonian formalism and mechanics; Vari-
ational problems with constraints; Second-order conditions
3. Optimal control problem formulations; Variational approach to the fixed-
time, free-endpoint problem
4. Pontryagin maximum principle
5. The Hamiltonian-Jacobi Bellman equation
6. Numerical Methods for Optimal Control Problems: Evaluation of Parameter-
Dependent Functionals and their Gradients; Indirect Methods; Direct
Methods; Dynamic programming
7. Applications: Time-optimal control of linear systems; Singular Control;
Optimal control to target curves; Linear quadratic regulator

Prerequisite
Optimization, Calculus of one variable, ordinary differential equations

Reference Books
1. Pinch, Enid R. Optimal control and the calculus of variations, Oxford
University Press
2. Daniel Liberzon, Calculus of variations and optimal control theory A
concise introduction, Princeton University Press
3. Kirk, Donald, E. Optimal control theory: An introduction, Dover Pub-
lications

1
 Analysis and control of Switched Systems EE6417
Slot E
Course Instructor: Dr. Ramkrishna Pasumarthy

Course contents:

Mathamatical Foundations:
1. Matrices, Matrix operations, Eigen values, Eigen vectors, Matrix transformations like
SVD, Cholesky factorization etc
2. Matrix norms: H_2 and H_infinity
2. Basics of linear programming and Introduction to Yalmip and Sedumi toolboxes in
MATLAB
2. Introduction to Convex sets and Convex funtions
3. Linear Matrix inequalities in Control
LMI as convex constraint and Geometry of LMIs
Numerical and analytical solutions to LMIs
Kronecker product related theorems
Lyapunov LMI
Methods to reformulate hard problems into LMI's: Change of Variables, Schur
Complement
Standard LMI problems: Feasibility problem, linear optimization problem and generalized
eigenvalue
minimization problem
D-Stability using LMI's
4. Formulation of Control problems using LMI's: Stability and Controller Design
Stabilizability
Simultanous Stabilizability
Positive real lemma and Bounded real lemma
Extended H_{2} and H_{infinity} norm characterizations
Multi-objective control using LMI's: like H_{2} and H_{infinity}
5. Eigen Structure Assignment in Control
6. Pole placement: Regional and Exact using LMIs and Eigen Structure Assignment

Switched systems:
1. Basic concepts: State and Time dependent Switching
2 Necessary and Sufficiant Conditions for Arbitrary switching stability
3 Stability under Arbitrary Switching: Existence of Common and Switched Lyapunov
functions
4. Stabilizing Switched control design with pole placement for SISO and MIMO systems
 Nonlinear Control Design-EE 573
Course outline

The course deals with nonlinear analysis for the most part and the remaining
is devoted to control design techniques.

1. Mathematical preliminaries: Open and closed sets, compact set, dense

set, Continuity of functions, Lipschitz condition, smooth functions, Vector
space, norm of a vector, normed linear space, inner product space.
2. We will begin with an introduction to simple mechanical systems wherein
the notion of degree-of-freedom, configuration space, configuration vari-
ables will be brought out. The state-space models of a few benchmark
examples in nonlinear control will be derived using Euler-Lagrange for-
mulation. The notion of equilibrium points and operating points will help
us to extract linearized models based on Jacobian linearization.
3. Second-order nonlinear systems occupy a special place in the study of non-
linear systems since they are easy to interpret geometrically in the plane.
Here, we will touch upon the concept of a vector field, trajectories, vector
field plot, phase-plane portrait and positively invariant sets. The classi-
fication of equilibrium points based on the eigenvalues of the linearized
system will also be introduced and we will see why the analysis based on
linearization fails in some cases. Periodic solutions and the notion of limit
cycles will lead us to the Bendixsons theorem and Poincare-Bendixson
criteria that provide sufficient conditions to rule-out and rule-in the exis-
tence of limit cycles respectively for a second-order system. We will end
this discussion with two methods for obtaining an approximate solutions
of periodic solutions.

4. Stability notions: Stability is central to control system design and here we

will study various notions of stability such as Lagrange stability, Lyapunov
stability, asymptotic stability, global asymptotic stability, exponential sta-
bility, relative stability and instability. The tools that we will use to infer
the stability properties include Lyapunovs direct and indirect method, La
Salles invariance property and singular perturbations.

5. Design methods: Finally, we will see the design of control laws based on
Lyapunov function and Sliding mode control and illustrate the methodol-
ogy on a few benchmark examples.

References

1. Nonlinear System Analysis: M. Vidyasagar

2. Nonlinear Systems: H. K. Khalil