Scribd
Upload
60 views8 pages

Linear Quadratic Control: Derivation: Indian Institute of Technology Roorkee

This document describes the derivation of the optimal linear quadratic control law. It shows that: 1) The optimal control law that minimizes a performance index is given by u = -Kx, where K is the feedback gain matrix. 2) By substituting the control law into the system model and performance index, an equation relating K and system matrices A, B, Q, R can be derived. 3) Solving this equation for K results in the optimal feedback gain being K = R-1B*P, where P satisfies the continuous-time Riccati equation A*P+PA-PBR-1B*P+Q=0.

Uploaded by

Shivam
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
60 views8 pages

Linear Quadratic Control: Derivation: Indian Institute of Technology Roorkee

This document describes the derivation of the optimal linear quadratic control law. It shows that: 1) The optimal control law that minimizes a performance index is given by u = -Kx, where K is the feedback gain matrix. 2) By substituting the control law into the system model and performance index, an equation relating K and system matrices A, B, Q, R can be derived. 3) Solving this equation for K results in the optimal feedback gain being K = R-1B*P, where P satisfies the continuous-time Riccati equation A*P+PA-PBR-1B*P+Q=0.

Uploaded by

Shivam
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 8

INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

LINEAR QUADRATIC CONTROL:


DERIVATION
OPTIMAL CONTROL PROBLEM

 Consider a system represented in state-space form as


x&= Ax + Bu (1)

 Problem: Determination of matrix K of the optimal control law

u =-Kx (2)

 Such that the performance index given below is minimized


�
J = (
� x T
Qx + u T
Ru ) dt (3)
0
where, Q is positive semi-definite Hermitian and R is positive definite
Hermitian matrix

2
 Substitute the control law from (2) in (1), we get

x&= ( A - BK ) x (4)

 Replace the value of u from (2) in (3). Hence,


�
J =� ( )
xT ( Q + K T RK ) x dt (5)

0
 For ease of convenience, set
-d T
( x ( Q + K RK ) x =
T T
) dt
( x Px ) (6)

where, P is positive definite Hermitian matrix

3
 Simplifying (6), we get

( )
xT ( Q + K T RK ) x = - x&T Px - xT Px&

 From (4), we get


xT ( Q + KT RK ) x = -xT �
(� ) P + P( A - Bk ) �
T
A - BK x
�
 Hence
-( Q + KT RK ) = ( A - BK ) P + P ( A - Bk )
T
(7)
 From (5) � �
-d
J =� ( )
xT ( Q + K T RK ) x dt = � ( xT Px )
dt
0 0

= xT (0) Px(0) - xT (�) Px(�)


 An assumption here is that (A-BK) is stable. Hence x(�) = 0 and
J = xT (0)Px(0)
4
 To obtain the optimal control law, assume
R = T *T
 From (7), we get
( A - K B ) P + P( A - BK ) + Q + K T TK = 0
* * * * *

 Add and subtract P * BR -1 B * P to both sides of equation, we get

A P + PA- PBR B P +Q- PBK - K B P + K T TK + P B( T T ) B*P = 0


* -1 * * * * * * * -1

 On rearranging the terms by using property (AB)* = B*A* , we obtain

AP+PA-PBR B P+Q-PBK -K T ( T )
* -1 * *
( * * -1
) B P+(TK) TK+P B( T T)
* * * * -1 *
B P=0

5
 On further simplification, we get

AP+PA-PBR B P+Q-PBT TK +(TK) TK -( T ) B P +P B( T T) B P = 0


* -1 * -1 *
( * -1 *
) * * -1 *

 Thus, we get

( )( )
*
AP+PA-PBR B P+Q+ TK -( T ) B P TK -( T ) B P = 0
* -1 * * -1 * * -1 *

 The minimization of the performance index J with respect to K


requires the minimization of

( ) B P ) ( TK - ( T ) B P ) x
*
x TK - ( T
* * -1 * * -1 *

 The minima of the above non-negative expression occurs when it


is equal to 0, i.e.,
TK = ( T )
* -1
B* P

6
 Hence,
K =T -1
(T )
* -1
B* P
= R -1 B * P

 Therefore, the optimal control law is given by


-1 *
u =-Kx =-R BPx
where, P is the solution of the following Riccati equation

A*P+ PA- PBR-1B*P+Q = 0

7
8

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy