Scribd
Upload
63 views2 pages

Pointwise Min-Norm Control: March 14, 2018

This document discusses pointwise min-norm control (PWMC) to provide exponential convergence of system states to zero while minimizing the control signal. PWMC is designed by choosing matrices Q and R to solve the continuous algebraic Riccati equation and define a positive definite Lyapunov function. The control input is chosen to satisfy the control Lyapunov function constraint while minimizing the control effort. This results in a PWMC law that exponentially stabilizes the system states at a rate determined by the parameter . PWMC can also be applied to affine nonlinear systems by first feedback linearizing the system.

Uploaded by

Muhammad Waqas
Copyright
© © All Rights Reserved
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
63 views2 pages

Pointwise Min-Norm Control: March 14, 2018

This document discusses pointwise min-norm control (PWMC) to provide exponential convergence of system states to zero while minimizing the control signal. PWMC is designed by choosing matrices Q and R to solve the continuous algebraic Riccati equation and define a positive definite Lyapunov function. The control input is chosen to satisfy the control Lyapunov function constraint while minimizing the control effort. This results in a PWMC law that exponentially stabilizes the system states at a rate determined by the parameter . PWMC can also be applied to affine nonlinear systems by first feedback linearizing the system.

Uploaded by

Muhammad Waqas
Copyright
© © All Rights Reserved
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/ 2

Pointwise Min-Norm Control

March 14, 2018

Objective: In this lecture, we aim to design a controller to provide exponentially convergence


of the system states to zero while the control signal is minimized.
Consider the following affine linear system:
ẋ = Ax + Bu (1)
where the matrix pair (A, B) is controllable.
Choose symmetric positive definite matrices Q and R, and let P be the symmetric positive
definite solution to the continuous algebraic Riccati equations (CARE)
AT P + P A − P BR−1 B T P + Q = 0 (2)
Define the following positive-definite Lyapunov function
V (x) = xT P x (3)
Taking derivative of Eq. (3) and substituting Eq. (1) into it yields
V̇ (x) = 2xT P ẋ = 2xT P Ax + 2xT P Bu = xT (AT P + P A)x + (2xT P B)u (4)
Then, to achieve the exponentially convergence of the solution, the following condition is desired

V̇ (x) ≤ −V (x) (5)


where  > 0 determines the convergence rate. Tuning of  can make a tradeoff between the
control effort and the rate of convergence of the states to zero.
Substituting Eq. (4) into Eq. (5) gives the following condition
xT (AT P + P A)x + (2xT P B)u + V (x) ≤ 0 (6)
The above condition is called control Lyapunov function (CLF) constraint. Now, the minimal
selection of an admissible control u is desired to be designed subject to the CLF constraint of
Eq. (6). Such admissible control is called pointwise min-norm control (PWMC) and should be
defined as
u = {argminkuk s.t. CLF constraint satisfied} (7)
The following PWMC law is a solution to the minimization problem
( φ (x)φ (x)
− φT0 (x)φ11 (x) if φ0 (x) > 0
u= 1 , (8)
0 if φ0 (x) ≤ 0
where φ0 and φ1 are defined as
φ0 (x) = xT (AT P + P A)x + V (x) (9)
φ1 (x) = 2xT P B
Remark 1: Note that the Lyapunov function defined in Eq. (3) satisfies the following condition:
λmin (P )kxk2 ≤ V (x) ≤ λmax (P )kxk2 (10)
λmin (P )kx(0)k2 ≤ V (x(0)) ≤ λmax (P )kx(0)k2
and solution of Eq. (5) is
V (x(t)) ≤ e−t V (x(0)) (11)
Combining Eq. (11) with Eq. (10) gives

kxk ≤ βe− 2 t kx(0)k (12)
q
where β = λλmax (P )
min (P )
. Eq. (12) shows that using the designed PWMC that satisfies the inequality
of Eq. (11), the system states exponentially converge to zero with the converge rate 2 .
Remark 2: Consider the general affine nonlinear systems,
ẋ = f (x) + Bu (13)
To use the PWMC law for the nonlinear system of Eq. (13), feedback linearization should be
first applied to the above system to achieve an affine linear system in the form of Eq. (1) (if
system of Eq. (13) is a feedback linearizable system).
Remark 3: Further details on PWMC and CLF can be found in a book entitled Robust
Nonlinear Control Design by Randy A. Freeman et al., 1996.

You might also like