Scribd
Upload
91 views5 pages

Optimal Control: Optimal Control For Continuous Time Dynamical Systems: LQR, Minimum Time Control

This document discusses optimal control for continuous time dynamical systems, including linear quadratic regulator (LQR) problems and minimum time control problems with inequality state and control constraints. It outlines topics like inequality state constraints, inequality control constraints, and examples of applying optimal control to problems like Zermelo's problem and the brachistochrone problem. Methods covered include the penalty function approach and Hamiltonian approach for handling inequality constraints. Minimum time control problems subject to control constraints are also discussed.

Uploaded by

7sen
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
91 views5 pages

Optimal Control: Optimal Control For Continuous Time Dynamical Systems: LQR, Minimum Time Control

This document discusses optimal control for continuous time dynamical systems, including linear quadratic regulator (LQR) problems and minimum time control problems with inequality state and control constraints. It outlines topics like inequality state constraints, inequality control constraints, and examples of applying optimal control to problems like Zermelo's problem and the brachistochrone problem. Methods covered include the penalty function approach and Hamiltonian approach for handling inequality constraints. Minimum time control problems subject to control constraints are also discussed.

Uploaded by

7sen
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

Optimal Control

Lecture 11

Optimal Control for Continuous Time Dynamical Systems: LQR, Minimum Time
Control

John T. Wen

February 22, 2002


JTW-OC11 RPI ECSE 6440 Optimal Control.

Outline

 Inequality state and control constraints


– Inequality state constraint
– Inequality control constraint

 examples (we’ll finish off the Zermelo’s problem and Brachistochrone problem)

Ref: Bryson & Ho, Sec. 3.8-3.9, Ch.5, Stengel 3.4,3.5,3.7


Last Time

 LQR with free terminal state or fixed terminal state (let’s revisit HW#3 problem 2)

February 22, 2002Copyrighted by John T. Wen Page 1


JTW-OC11 RPI ECSE 6440 Optimal Control.

Inequality Constraint

Suppose we impose a set of hard inequality constraints:

c(x(t ); u(t ); t )  0:

Example:
Control constraint: (u u max )(u umin )  0.
State constraint: Ax(t )  0.
Penalty function approach: convert hard constraint to soft constraint and solve the uncon-
strained problem as before, i.e., modify J to
ZT
J = Jo + p(x(t ); u(t ); t ) dt
0

where p is the exterior penalty function:


8
< kc(x; u; t )2 c(x; u; t )  0
p(x u t ) =
; ;
: 0 c(x; u; t ) < 0:

Increase k until the constraints are satisfied

February 22, 2002Copyrighted by John T. Wen Page 2


JTW-OC11 RPI ECSE 6440 Optimal Control.

Inequality Constraint (cont.)

Hamiltonian approach: Modify H to

H = L + λ T f + µT c

where 8
< µi = 0 if ci < 0
: µi  0 if ci = 0:
Costate equation (unchanged if c only depends on u)
 ∂L T  ∂ f T  ∂c T
λ̇ = λ µ:
∂x ∂x ∂x
Optimal control (unchanged if c only depends on x)
∂H ∂L T ∂f T ∂c
0= = +λ +µ :
∂u ∂u ∂u ∂u

Example: Unicycle with steering constraint

February 22, 2002Copyrighted by John T. Wen Page 3


JTW-OC11 RPI ECSE 6440 Optimal Control.

Minimum Time Control subject to Control Constraint

If u is restricted to be in a set U (e.g., C(u; t )  0 for a given function C), then the optimal
control needs to solve
u(t ) = min H (x; u; t ; λ):
u2U

Consider a single-input LTI system: ẋ = Ax + Bu, 1  u  1.


Hamiltonian: H = 1 + λT (Ax + Bu). Optimal control must then satisfy:
8
< 1 if BT λ < 0
u=
: 1 if BT λ > 0 :

BT λ is called the switching function.


Example: Minimum time control of double integrator ẍ = u with specified initial condi-
tion x0 and final condition x f = 0, and control constraint juj  1.

February 22, 2002Copyrighted by John T. Wen Page 4

You might also like