Professional Documents
Culture Documents
NASA
.
TP
3
1118 c.1
.
Eigenvalue/Eigenvector Assignment
S. Srinathkumar
P E B ~ U A R Y197 8
NASA
IlllH0134489IllIIll 1ulll111ill I Il l
NASA Technical Paper 1118
S. Srinathkumar
1978
SM AY U MR The problem o f e i g e n v a l u e assignment i n a l i n e a r t i m e - i n v a r i a n t system u s i n g o u t p u t feedback i s c o n s i d e r e d . New s u f f i c i e n t c o n d i t i o n s are d e r i v e d t o 1) d i s t i n c t e i g e n v a l u e s a s s i g n an almost a r b i t r a r y s e t o f minimum (n,m + r where n , m, and r are t h e number o f s t a t e s , i n p u t s , and o u t p u t s , r e s p e c t i v e l y . These c o n d i t i o n s p r e c i s e l y i d e n t i f y , t h e class o f systems where such a n assignment is i m p o s s i b l e . The s y n t h e s i s t e c h n i q u e a l s o h i g h l i g h t s t h e freedom i n s e l e c t i o n of closed-loop e i g e n v e c t o r s under o u t p u t feedback. The u t i l i t y o f e i g e n v a l u e / e i g e n v e c t o r assignment i n t r a n s i e n t r e s p o n s e shaping is i l l u s t r a t e d by t h e d e s i g n o f a c o n t r o l l e r f o r t h e l a t e r a l dynamics o f an a i r c r a f t .
INTRODUCTION
Control system d e s i g n b a s e d on e i g e n v a l u e o r p o l e assignment h a s r e c e i v e d a g r e a t d e a l o f a t t e n t i o n i n t h e l i t e r a t u r e . It is w e l l known t h a t f o r a cont r o l l a b l e system, i f s t a t e v a r i a b l e feedback i s employed, a l l e i g e n v a l u e s can b e a s s i g n e d ( r e f . 1 ) . Also i t i s known t h a t f o r m u l t i - i n p u t systems, t h e feedback l a w a s s i g n i n g a g i v e n s e t o f e i g e n v a l u e s i s n o t unique and t h a t d i f f e r e n t cont r o l laws can y i e l d i d e n t i c a l e i g e n v a l u e s w h i l e y i e l d i n g r a d i c a l l y d i f f e r e n t e i g e n v e c t o r s . S i n c e t h e e i g e n v e c t o r s determine t h e i n f l u e n c e of each e i g e n v a l u e on each s t a t e v a r i a b l e r e s p o n s e , f a i l u r e t o use t h e m u l t i - i n p u t d e s i g n freedom f u l l y may r e s u l t i n u n d e s i r a b l e mode c o u p l i n g and o t h e r poor t r a n s i e n t b e h a v i o r . For n - s t a t e feedback s y s t e m s , it h a s been shown ( r e f . 2 ) t h a t w i t h m i n p u t s , i n a d d i t i o n t o t h e assignment o f a l l n e i g e n v a l u e s , up t o m e n t r i e s i n each e i g e n v e c t o r can be a r b i t r a r i l y a s s i g n e d . However, t h e problem o f e i g e n v a l u e assignment u s i n g o u t p u t feedback i n s t e a d o f s t a t e feedback h a s n o t y e t been comp l e t e l y r e s o l v e d . The problem o f d e t e r m i n i n g c o n d i t i o n s under which a l l e i g e n v a l u e s o f a s y s t e m can be a r b i t r a r i l y a s s i g n e d t o a system under o u t p u t feedback h a s been i n v e s t i g a t e d i n r e f e r e n c e s 3 and 4 . Bounds on t h e number o f s t a t e s , i n terms o f number of i n p u t s , o u t p u t s , c o n t r o l l a b i l i t y , and o b s e r v a b i l i t y i n d i c e s are e s t a b l i s h e d f o r complete p o l e a s s i g n a b i l i t y . Reference 5 shows t h a t f o r a system w i t h r o u t p u t s , i f m r 2 n t h e n t h e system i s p o l e - a s s i g n a b l e provided t h e feedback g a i n elements are allowed t o be complex numbers. References 6 and 7 a d d r e s s t h e converse problem: g i v e n a c o n t r o l l a b l e , o b s e r v a b l e system, how many e i g e n v a l u e s can be a r b i t r a r i l y a s s i g n e d t o t h e system. I n g e n e r a l , it is concluded ( r e f . 7 ) t h a t minimum (n,m + r 1) e i g e n v a l u e s can 'Ialmost1? always be a s s i g n e d t o t h e system u s i n g o u t p u t feedback. The q u a l i f i c a t i o n llalmost" was i n t r o d u c e d t o cover c l a s s e s o f systems where such a n assignment i s i m p o s s i b l e . I n e f f e c t , t h e a n a l y s i s i n r e f e r e n c e 7 does n o t p r e c i s e l y determine t h e condit i o n s under which ( m + r 1) e i g e n v a l u e s cannot b e a s s i g n e d t o t h e system.
T h i s r e p o r t c o n s i d e r s t h e problem of d e t e r m i n i n g t h e number o f e i g e n v a l u e s a s s i g n a b l e t o a g i v e n system. By f o r m u l a t i n g a n e i g e n v a l u e / e i g e n v e c t o r a s s i g n ment problem, s u f f i c i e n t c o n d i t i o n s r e q u i r e d f o r t h e assignment o f minimum 1 ) e i g e n v a l u e s are d e r i v e d . These c o n d i t i o n s p r e c i s e l y i d e n t i f y (n,m + r
t h e class o f systems which can be a s s i g n e d only d e i g e n v a l u e s , where maximum ( m , r ) 5 d < (m + r 1 ) . The new f o r m u l a t i o n p e r m i t s t h e development o f a n algorichm t o a s s i g n (m + r 1) e i g e n v a l u e s . I n addition, (r 1) e i g e n v e c t o r s can be p a r t i a l l y a s s i g n e d w i t h , a t most, m e n t r i e s i n each v e c t o r a r b i t r a r i l y chosen. I n t h e e v e n t n > ( m + r I ) , various synthesis a l t e r n a t i v e s t o stabil i z e t h e system are a l s o i n v e s t i g a t e d s i n c e i n t h i s case a l l system e i g e n v a l u e s cannot be a s s i g n e d . The c o u n t e r example o f r e f e r e n c e 7 is used t o demonstrate t h e u t i l i t y of t h e new s u f f i c i e n t c o n d i t i o n s i n i d e n t i f y i n g systems which c a n n o t be a s s i g n e d (m + r 1) e i g e n v a l u e s . F i n a l l y , t h e advantage o f both e i g e n v a l u e and e i g e n v e c t o r a s s i g n m e n t s i n r e s p o n s e s h a p i n g i s i l l u s t r a t e d by d e s i g n i n g a c o n t r o l l e r t o meet the l a t e r a l h a n d l i n g q u a l i t i e s s p e c i f i c a t i o n s f o r an a i r c r a f t .
SYMBOLS
U.S.
A A E Rnxn
C a l c u l a t i o n s were made i n
real matrix
aY B
C
C
l a t e r a l a c c e l e r a t i o n , m/sec2 ( f t / s e c 2 )
input matrix
mea s u r emen t m a t r i x
measurement v e c t o r m a t r i x used i n e q u a t i o n ( 4 5 )
E
D(k)
d,i,j,k,t
matrices d e f i n e d by e q u a t i o n (B4)
indices v e c t o r d e f i n e d i n n o t a t i o n ( 1 ) i n appendix B m a t r i x d e f i n e d by e q u a t i o n (9) v e c t o r defined by s t e p 3 ( a ) o f appendix B m a t r i x d e f i n e d by e q u a t i o n (IO) v e c t o r d e f i n e d by s t e p 3 ( a > o f appendix B
j E {~(k> K
is an element o f i n d e x set
A(k)
feedback m a t r i x
Li
R
M(k) ,M(k-l
matrices d e f i n e d by n o t a t i o n ( 6 ) i n appendix B
m
max m in
number o f i n p u t s maximum v a l u e minimum v a l u e m a t r i x d e f i n e d by e q u a t i o n (13) number o f s t a t e s m a t r i x used i n e q u a t i o n ( 4 6 ) and d e f i n e d immediately afterward roll r a t e , deg/sec matrices d e f i n e d by e q u a t i o n ( B 3 )
N
n
P
P
Q(k-1) , Q ( i )
Q(0) 9
m a t r i x d e f i n e d by n o t a t i o n ( 3 ) i n appendix B feedback v e c t o r d e f i n e d by e q u a t i o n ( 1 8 ) number o f o u t p u t s m a t r i x d e f i n e d by e q u a t i o n ( 8 ) v e c t o r d e f i n e d immediately p r i o r t o e q u a t i o n ( 2 0 ) t r a n s f o r m a t i o n matrices d e f i n e d where u s e d i n p u t v e c t o r d e f i n e d by e q u a t i o n ( 2 ) modal m a t r i x ( m a t r i x o f e i g e n v e c t o r s ) m a t r i x used i n e q u a t i o n ( 1 2 ) eigenvectors v e c t o r used i n e q u a t i o n ( 6 ) ; a p a r t i t i o n o f v
r
S
S
TO,*1
U
v(r>
V,Vi W
state v e c t o r d e f i n e d by e q u a t i o n s ( 1 )
n
x
x E Rn
real v e c t o r
Y
Z(r>,z(t)
o u t p u t v e c t o r d e f i n e d by e q u a t i o n s (1
matrices c o n t a i n i n g t h e v e c t o r s
respectively
zi;
= 1 t o r and 1 t o t ,
v
v e c t o r used i n e q u a t i o n ( 6 ) ; a p a r t i t i o n o f s i d e s l i p , deg
B
A( 1
,At
(1
,Ar ( 1 ,A( k)
set o f i n d i c e s d e f i n e d i n appendix B
&a
&k
a i l e r o n a n g u l a r d e f l e c t i o n , deg d e t e r m i n a n t d e f i n e d i n s t e p 1 o f appendix B rudder a n g u l a r d e f l e c t i o n , deg p e r t u r b e d q u a n t i t i e s d e f i n e d i n s t e p 1 and s t e p 4 o f appendix B, respectively v e c t o r d e f i n e d by n o t a t i o n ( 6 ) i n appendix B d i a g o n a l e i g e n v a l u e matrices used i n e q u a t i o n s ( 1 2 ) and ( B 6 ) , respectively eigenvalues v e c t o r s d e f i n e d by e q u a t i o n s ( 2 1 )
&r
&Xk,&Zk
e
A r 9%
X,Xi,Xk C , W I 91-12 k
scalar d e f i n e d by n o t a t i o n ( 4 ) i n appendix B
bank a n g l e , deg yaw r a t e , deg/sec
Superscripts:
-1
A
matrix inverse
-
, ,-,*
transformed q u a n t i t y
= AX + BU
y = cx
(1)
u = K y
i n o r d e r t o a s s i g n a r b i t r a r y e i g e n v a l u e s f o r t h e closed-loop system. To i n d i cate c l e a r l y t h e freedom a v a i l a b l e i n t h e s e l e c t i o n o f closed-loop e i g e n v a l u e s and e i g e n v e c t o r s under o u t p u t feedback, t h e measurement m a t r i x C i s assumed t o be i n a s p e c i a l c a n o n i c a l form:
= [Cl : CZ]
c1 = [I .;
7 .I
c2 =
[. 1
.I
11
c = p
I . . .
(3)
BKC) a f t e r a p p l y i n g feedback l a w ( 2 )
(i = 1 , 2,
sat i s f i e s
(A
+ BKC)vi = X i v i
. . .,
n)
(4)
where X i is t h e i t h e i g e n v a l u e and v i i s t h e c o r r e s p o n d i n g e i g e n v e c t o r . The e i g e n v a l u e / e i g e n v e c t o r assignment problem i s t o determine t h e number o f e i g e n v a l u e s i n e q u a t i o n ( 4 ) t h a t can b e a r b i t r a r i l y a s s i g n e d and t o determine t h e freedom a v a i l a b l e i n t h e s e l e c t i o n o f t h e a s s o c i a t e d e i g e n v e c t o r s .
where A11,Bl E Rmxm and B1 is n o n s i n g u l a r . S i n c e B i s f u l l r a n k , t h e nons i n g u l a r i t y o f B1 can be a s s u r e d , i f n e c e s s a r y , by r e o r d e r i n g t h e s t a t e v a r i a b l e s i n e q u a t i o n s ( 1 ) . Completing t h e m u l t i p l i c a t i o n o f the p a r t i t i o n e d matrices and some algebraic o p e r a t i o n s (ref. 8) p e r m i t s e q u a t i o n ( 5 ) t o be e x p r e s s e d as a s e t o f c o n s t r a i n t s on t h e s e l e c t i o n o f e i g e n v e c t o r s . For c l a r i t y o f p r e s e n t a t i o n , t h e s e r e l a t i o n s are d e t a i l e d o n l y f o r r e a l e i g e n v a l u e s . Extens i o n t o complex c o n j u g a t e p a i r s i n q u a s i - d i a g o n a l form y i e l d i n g real e i g e n v e c t o r p a i r s is s t r a i g h t f o r w a r d ( r e f . 8 ) .
For real e i g e n v a l u e s
[XI,_,
[AI
F]w = [ + G
(6) (7)
v ' = [z'
: w'];
+ B I K C I V = XZ
v
is t h e e i g e n v e c t o r with
A1 = [All
: A121
(11)
Equation ( 6 ) r e p r e s e n t s an underdetermined system o f n m equations i n n unknowns. Thus m e i g e n v e c t o r e n t r i e s c o r r e s p o n d i n g t o t h e z-vector can be chosen a r b i t r a r i l y provided does n o t c o i n c i d e w i t h t h e spectrum o f F. Examination o f e q u a t i o n ( 7 ) r e v e a l s t h a t a t l e a s t r e i g e n v a l u e s and r eigenv e c t o r s s a t i s f y i n g e q u a t i o n ( 6 ) can be a s s i g n e d t o t h e s y s t e m i n e q u a t i o n s ( 1 ) by t h e feedback m a t r i x
where A, is t h e d i a g o n a l matrix o f r e i g e n v a l u e s and Z ( r ) and V ( r ) have t h e form T ( r ) = [ti : t 2 : : tr] (where t i are v e c t o r s ) . The s o l u t i o n t o e q u a t i o n ( 1 2 ) is .guaranteed provided t h e eigenvalues/eigenvectors are chosen t o i n s u r e t h e n o n s i n g u l a r i t y o f [dr)]. should be noted t h a t It i n t h e case o f s t a t e v a r i a b l e feedback ( C = I ) a l l n e i g e n v a l u e s can be a s s i g n e d t o t h e system provided t h e modal m a tn . x ( m a t r i x o f e i g e n v e c t o r s ) ri ' V = [VI : v2 : . : vn] i s n o n s i n g u l a r . An a l g o r i t h m which c o n s t r u c t s such a n o n s i n g u l a r V is d e t a i l e d i n r e f e r e n c e 8 . A pendix B e x t e n d s t h i s algor i t h m t o g u a r a n t e e t h e n o n s i n g u l a r i t y o f [CV(r)f.
...
..
e i g e n v a l u e s can be
The a n a l y s i s so far i n d i c a t e s t h a t only max(m,r) e i g e n v a l u e s can be a s s i g n e d t o t h e system u s i n g o u t p u t feedback. However, by s a c r i f i c i n g some degree of freedom i n t h e s e l e c t i o n o f t h e a s s o c i a t e d e i g e n v e c t o r s it is p o s s i b l e 1) t o extend t h e number o f e i g e n v a l u e s t h a t can be a s s i g n e d t o min(n,m + r as is shown i n t h e f o l l o w i n g s e c t i o n .
ALGORITHM TO A S S I G N M I N I M U M (n,m
+ r
1 ) EIGENVALUES
The basic approach i n t h e development o f t h i s a l g o r i t h m is t o 2 o n s t r u c t t h e o u t p u t feedback l a w i n e q u a t i o n ( 2 ) as a sum o f two feedbacks (K + K). The first feedback (E) a s s i g n s t e i g e n v a l u e s , and t h e second feedback a s s i g n s t ) eigenvalues while ensuring t h e p r o t e c t i o n of the t a d d i t i o n a l min(m,n e i g e n v a l u e s a l r e a d y a s s i g n e d . The c o n s t r u c t i o n procedure y i e l d s a s e t o f s u f 1 ) e i g e n v a l u e s . These condif i c i e n t c o n d i t i o n s f o r a s s i g n i n g min(n,m + r t i o n s a l s o h e l p c h a r a c t e r i z e the class o f systems which cannot be a s s i g n e d 1 ) e i g e n v a l u e s . F i n a l l y , some d e s i g n freedom s t i l l e x i s t s t o min(n,m + r p a r t i a l l y assign (r - 1 ) eigenvectors.
[.N!
N2
[VI
: v2 :
...
Vt]
(13)
N1 E R t x t and n o n s i n g u l a r . Appendix B d e t a i l s a procedure t o c o n s t r u c t N. Let K" be t h e nonunique feedback ( e q . ( B 6 ) ) corresponding t o t h i s a s s i g n ment. Then t h e closed-loop m a t r i x is
& = A + BffC
S t e p 2: Apply a c o o r d i n a t e t r a n s f o r m a t i o n
(C,$,B)
-+
(14)
( C T ?, T ~ - I ~ T T ~ - ~ B > ,,
where
_I
(16)
It
I n o r d e r t o a s s i g n a d d i t i o n a l e i g e n v a l u e s t o t h e system o f e q u a t i o n s (162 w h i l e p r o t e c t i n g t e i g e n v a l u e s (A,) a l r e a d y a s s i g n e d , t h e second feedback (K) is r e s t r i c t e d t o b e o f u n i t y r a n k o f t h e form K = qRc w i t h q E Rm, R E R r , and is chosen so t h a t
R'P1
Now q
c2]
= [o : c]
(17)
must be chosen s o t h a t
is assigned
min(m,n
t ) eigenvalues.
S i n c e ( i 2 2 ,g2) is c o n t r o l l a b l e , the f o l l o w i n g r e s u l t h o l d s : Theorem 1: The s i n g l e o u t p u t subsystem (c,A22,B2) can be a s s i g n e d min(m,n t ) e i g e n v a l u e s i f and o n l y i f ( a ) (c,A22) is o b s e r v a b l e and ( b ) 62 is f u l l r a n k .
/ . 4
Theorem 1 f o l l o w s d i r e c t l y from lemma 1 . F u r t h e r , c o n d i t i o n s ( a ) and ( b ) restrict t h e a d m i s s i b l e s e t o f e i g e n v a l u e / e i g e n v e c t o r assignments i n s t e p 1 . For c o n c e p t u a l convenience, these p a r a m e t r i c r e s t r i c t i o n s are formulated i n terms o f c o n t r o l l a b i l i t y c o n d i t i o n s o f a f i c t i t i o u s dynamic system i n t h e s t a t e variable representation. After matr;x o p e r a t i o n s i n e q u a t i o n s (14) and (16) have been performed, t h e submatrix A22 can be w r i t t e n as
= Ai25
Ai21-11
c'1-12
n = $5
w i t h 6 E Rn-t, N2 = -N2N1'1.
Bi1-11
1-11
E
q E Rm,
Rt,
(21 1
1-12 E R ' ,
1-11 =
fi&,
1-12 = s l q , and
&
rank; ( d )
[[A22
+ A{~G;],C~
is c o n t r o l l a b l e .
Conditions ( c ) and ( d ) c l e a r l y in
2:
+ r
N1 i s n o n s i n g u l a r
I1
62 = B2 + N2B1
nl
I11
{ [A;2
+ Ai&]
is controllable
F (appendix B ) .
Condition I i s r e q u i r e d t o g u a r a n t e e t h e e x i s t e n c e o f t h e t r a n s f o r m a t i o n T I i n s t e p 2 . T h i s c o n d i t i o n can be e x p l i c i t l y i n c l u d e d i n t h e s y n t h e s i s procedure as d e t a i l e d i n appendix B. Condition I11 is o b t a i n e d from t h e p r o p e r t y t h a t f o r system ( e q s . (2111, t h e class of feedback from i n p u t 1-11 should be r e s t r i c t e d s o t h a t t h e c o n t r o l l a b i l i t y o f t h e feedback system w i t h r e s p e c t t o t h e i n p u t 1-12 i s p r e s e r v e d . Conditions I1 and I11 y i e l d nonlinear_ a l g e b r a i c c o n s t r a i n t s for t h e e l e m e n t s o f t h e e i g e n v e c t o r parameter matrix N2 and t h u s , i n g e n e r a l , can only be used as t e s t c o n d i t i o n s f o r each assignment i n s t e p 1 . However, example 1 o f t h e s e c t i o n e n t i t l e d llNumerical Examples1! shows how t h e s e c o n d i t i o n s can b e e x p l i c i t l y checked.
S t e p 4:
(22)
1)
+ r
e i g e n v a l u e s t o t h e system. ASSIGNING n
EIGENVALUES
The development so far has r e v e a l e d t h a t f o r systems where n > ( m + r - l ) , a l l system e i g e n v a l u e s cannot u s u a l l y be a s s i g n e d . However, by u s i n g equat i o n s ( 6 ) and ( 7 ) , it is p o s s i b l e t o d e r i v e c o n d i t i o n s f o r a s s i g n i n g a l l e i g e n v a l u e s o f t h e system as f o l l o w s .
If t h e e i g e n v a l u e s t o be a s s i g n e d are n o n c o i n c i d e n t w i t h t h e spectrum o f F ( i f n e c e s s a r y , by a p e r t u r b a t i o n i n s p e c i f i c a t i o n ) , t h e n e q u a t i o n ( 6 ) can b e e x p l i c i t l y solved f o r t h e w-vector and s u b s t i t u t e d i n t o e q u a t i o n ( 7 ) . T h i s subs t i t u t i o n y i e l d s a s e t o f homogeneous e q u a t i o n s o f t h e form
[ H i + B KL- 2-i = 0 1 1 1
(i
= 1 , 2,
. . ., n )
(23)
Then e q u a t i o n (23) h a s a n o n t r i v i a l
+ BlKLi] < m f o r a l l i
(24)
A set of n nonlinear equations i n the m , r parameters of the gain m a t r i x K can be d e r i v e d by s e t t i n g t h e a p p r o p r i a t e d e t e r m i n a n t i n e q u a t i o n (24) e q u a l t o z e r o . However, no g e n e r a l c o n c l u s i o n s can b e drawn r e g a r d i n g t h e e x i s tence of t h e s o l u t i o n .
A a l t e r n a t i v e approach n o t i n v o l v i n g s o l u t i o n of n o n l i n e a r e q u a t i o n s is t o n a s s i g n t e i g e n v a l u e s as i n e q u a t i o n (14) and t o a t t e m p t t o a s s i g n t h e remaining (n t ) e i g e n v a l u e s approximately t o t h e subsystem i n e q u a t i o n (18) u s i n g t h e c o n d i t i o n i n e q u a t i o n (24). I n t h i s case, a s e t o f ( n t) l i n e a r equations i n m unknowns r e s u l t s , and a l e a s t - s q u a r e s s o l u t i o n can be o b t a i n e d . The f o r e g o i n g d i s c u s s i o n s assume t h a t r 2 m i f n e c e s s a r y by c o n s i d e r i n g t h e d u a l system.
10
NUMERICAL EXAMPLES
0
0 0
0-
A =
0 O
1L
0 ,
B =
0 0
1-
c =
1
0
0 0
O 1
O01
1
l Reduce t h e system (25) t o t h e s p e c i a- form o f e q u a t i o n s ( 1 ) by o r d e r i n g s t a t e v a r i a b l e s as ( X J , X ~ , X ~ , X t o make ~ ) n o n s i n g u l a r (appendix A) and by applyi n g t h e c o o r d i n a t e t r a n s f o r m a t i o n (C,A,B) (ET0 ,To-lZTo ,To-IG) where
sa-
::::
0 0 1 0 0 0 0 1
(26)
11
t o yield
0 0 0 0 0 0 1 0 0 0
A =
0 0
B =
c =
1
0
O1 1
From e q u a t i o n ( 6 )
where t h e e i g e n v a l u e i s
X and t h e e i g e n v e c t o r is
From theorem 2 , c o n d i t i o n I , n1 # 0 ; t h e r e f o r e , choose n1 = 1 w i t h o u t l o s s -9 -n4) I . Condition I1 i s met f o r all o f g e n e r a l i t x . Then, N2 = ( 4 2 c h o i c e s o f N2 since B i = 0. Condition I11 i m p l i e s t h a t
/
11-12
-n3
0
-n4
0
l !
n2n3
n4
n32 + n3n4 #
( 3 11
12
0
0
_1
0
:l
0
1-
OI. 0
It i s r e q u i r e d t o a s s i g n e i g e n v a l u e s c l o s e t o - 1 , -2, and -5. The system ( 3 2 ) can be reduced t o t h e form of e q u a t i o n s ( 1 ) by f o l l o w i n g t h e procedure i n append i x A as Step 1: S t e p 2: Measurement m a t r i x i s i n t h e d e s i r e d form w i t h Apply c o o r d i n a t e t r a n s f o r m a t i o n
+
ea
nonsingular.
(C,A,B)
where
1
0
~CT~,T~-~~T~,T~-~B~
0
1
0
1
13
yielding
1
A =
0
1
1
B =
"1
1
c =
1
0
"1
1
B1 =
[;
(1
-1
s =
-1)
F = O
G
= (0
-1)
and e q u a t i o n (6) y i e l d s
x w = [A
-(1
A)]
1.": .1 1
Lz2
(35)
X # 0.
Equation (35) i n d i c a t e s t h a t z1 and > can be r b i t r a r i l y chosen, provided Now, a p p l y i n g t h e a l g o r i t h m t o a s s i g n ( m + r 1 ) eigenvalues y i e l d s t h e f o l l o w i n g s y n t h e s i s sequence.
Step 1: Assign A = -1. From appendix B , case 11, A t ( 1 ) = {l}, and t h i s i m p l i e s z1 # 0 i n e q u a t i o n (35). ( C o n d i t i o n I , theorem 2.) One a c c e p t a b l e assignment
is
14
N =
(1
1)'
(36)
and a nonunique feedback gain corresponding to this assignment from equation (B6) is
Step 2: Transform the system to the canonical form of equations (16) using
1 . 0
....
TI 1 . 1
1 . 0
to yield
-1
-1
0
1
OI
. ... . ..
0 . - 1 0 . 0
0 0
j L
....
-1 -1
- 0
e=[;
; ;j
15
with
S i n c e feedback from I-rl does n o t affect t h e c o n t r o l l a b i l i t y of system ( 4 1 ) w i t h r e s p e c t t o ~ 2 t, h e subsystem is p o l e a s s i g n a b l e from c o n d i t i o n I11 o f theorem 2 . F u r t h e r , t h e assignment i n s t e p 1 s a t i s f i e s c o n d i t i o n I1 s i n c e 82 is n o n s i n g u l a r . Thus theorem 2 h o l d s . S t e p 3: Choose R' = (-2 1). It now remains t o choose q and A = -5. 3 S t e p 4: Assign where A2 = -2, A3 = -5 t o t h e s i n g l e o u t p u t subsystem i n e q u a t i o n (18)
T h i s choi2e p r o t e c t s
so t h a t
K = qR'
X = -1 a s s i g n e d i n s t e p 1 . a s s i g n s t h e e i g e n v a l u e s X2 = -2
c = E1
13
16
with feedback
and a s s i g n s t h e e i g e n v a l u e s -1,
The advantage o f combined c o n t r o l o f closed-loop e i g e n v a l u e s and eigenvect o r s u s i n g s t a t e v a r i a b l e feedback h a s been i n v e s t i g a t e d i n r e f e r e n c e IO. The u t i l i t y o f t h e o u t p u t feedback e x t e n s i o n s developed i n t h i s r e p o r t w i l l now b e i l l u s t r a t e d through t h e d e s i g n o f a l a t e r a l c o n t r o l l e r f o r a n a i r c r a f t . The l i n e a r p e r t u r b a t i o n model f o r t h e l a t e r a l motions o f an a i r c r a f t can
b e modeled as
j:
= Ax + Bu
y = Ex + Du
where x i s t h e s t a t e v e c t o r o f r o l l r a t e p , yaw r a t e 9, s i d e s l i p 6, and and r u d d e r bank a n g l e 0, r e s p e c t i v e l y . The c o n t r o l v e c t o r o f a i l e r o n 6, 6, a n g u l a r d e f l e c t i o n s is u . R o l l rate p , yaw r a t e Q, and l a t e r a l accelerat i o n ay c o n s t i t u t e t h e o u t p u t v e c t o r y . A l l a n g l e s are i n degrees, r a t e s i n deg/sec, and a c c e l e r a t i o n i n m/sec2 ( f t / s e c 2 ) . The r e s p e c t i v e matrices i n e q u a t i o n s (45) f o r a f i g h t e r a i r c r a f t a t a n a l t i t u d e of 6096 m (20 000 f t ) , a Mach number of 0.67, and an a n g l e o f a t t a c k of 3.45O are g i v e n by
(45)
17
- II
A =
r-3*79
-0.14
0.04 -0.36
0.06
-1
0.06
0.05
0
25
0.01
0
0.05
0
0 0 -3.42
0 0 0
1
L-0.13 -0.06
11.03
where t h e elements o f t h e matr-ces are approx-aated t o two s i g n i f i c a n t d i g ts. S i n c e t h e o u t p u t v e c t o r i n e q u a t i o n s (45) is d e r i v e d as a l i n e a r combinat i o n of both s t a t e v a r i a b l e s and c o n t r o l i n p u t s , t h e closed-loop system after a p p l y i n g feedback u = K*y t a k e s t h e form
'
(e +
-
~ K ( ? ) x + DPu
where P = [Im K*D]-' and K = PK* is t h e e q u i v a l e n t o u t p u t feedback matrix o b t a i n e d by s e t t i n g = 0. Thus, t h e a l g o r i t h m s developed e a r l i e r f o r systems w i t h 5 = 0 are a p p l i c a b l e t o systems o f t h e form o f e q u a t i o n s (451, provided P exists. Then t h e feedback g a i n
K*
i s computed by u s i n g t h e r e l a t i o n
(47)
K* =
K[Ik
+ fiK1-l
18
The h a n d l i n g q u a l i t i e s s p e c i f i c a t i o n s ( r e f . 1 1 ) imply t h a t t h e l a t e r a l a i r c r a f t dynamics should be composed o f two weakly coupled subsystems. R o l l r a t e and bank a n g l e c o n s t i t u t e t h e first subsystem and d i s p l a y predominantly t h e roll s u b s i d e n c e and s p i r a l modes. The second subsystem i s c h a r a c t e r i z e d by a w e l l damped Dutch r o l l mode d e f i n i n g t h e yaw rate and s i d e s l i p motions. These s p e c i f i c a t i o n s can now be f o r m u l a t e d as a n e i g e n v a l u e / e i g e n v e c t o r assignment problem.
Table I summarizes t h e modal c h a r a c t e r i s t i c s o f t h e free a i r c r a f t . From t h e table it is seen t h a t t h e Dutch roll mode is v e r y l i g h t l y damped and a p p e a r s dominantly i n the r e s p o n s e o f t h e roll v a r i a b l e s p and 9 as evidenced by dominant e n t r i e s i n t h e c o r r e s p o n d i n g e i g e n v e c t o r p a i r . Thus, t h e e i g e n v a l u e / e i g e n v e c t o r m o d i f i c a t i o n r e q u i r e s t h a t t h e closed-loop system have t h e modes and mode-variable a s s o c i a t i o n s o f t a b l e 11. ( I n t a b l e s I t o 111, j =
.
n.
I
iJ
B
-. 002
.261
.002 .998
i
1
Eigenvalue
-6
-1.0 - j0.2 +
-0.01
19
The o u t p u t feedback a n a l y s i s i n t h e main t e x t shows t h a t a l l system eigenv a l u e s can be a s s i g n e d s i n c e n = m + r 1 and o n l y two e i g e n v e c t o r s ( t ) can be a s s i g n e d w i t h a t most two (m) e n t r i e s i n each v e c t o r a r b i t r a r i l y chosen. The e i g e n v e c t o r freedom a v a i l a b l e was used t o c o n t r o l t h e s t r u c t u r e o f t h e eigenvect o r p a i r c o r r e s p o n d i n g t o t h e Dutch r o l l mode t o e f f e c t t h e d e s i r e d yaw rate and s i d e s l i p dominance. The modal c o u p l i n g matrices D ( k ) (eq. (B4)) a i d i n t h e s e l e c t i o n o f t h e Dutch roll mode t o y i e l d t h e a p p r o p r i a t e e i g e n v e c t o r forms. The f o u r e i g e n v e c t o r e n t r i e s t h a t were f r e e l y chosen corresponded t o t h e roll rate and yaw rate components o f t h e real e i g e n v e c t o r p a i r a s s o c i a t e d w i t h t h e Dutch roll mode. S i n c e g a i n magnitude c o n s t r a i n t s cannot be e x p l i c i t l y i n c l u d e d i n t o t h e s y n t h e s i s a l g o r i t h m , t h e d e s i g n parameters have t o be i t e r a t i v e l y modif i e d t o meet g a i n l i m i t r e q u i r e m e n t s . After some d e s i g n i t e r a t i o n s , a compromise d e s i g n y i e l d e d t h e modal c h a r a c t e r i s t i c s summarized i n t a b l e 111. For example, Z t was noted t h a t t h e Dutch roll mode damping could n o t be reduced ( t o improve $ and B r e s p o n s e s ) w i t h o u t v i o l a t i n g feedback g a i n l i m i t s which were s e t a t u n i t y f o r t h i s a n a l y s i s . The d e s i g n y i e l d e d a feedback g a i n m a t r i x K* (eq. ( 4 7 ) ) as
1-0.16
0.19
-0.6
7 Eigenvector
components
Eigenvalue of
-1.0
-0.01 1 ( S p i r a l)
-0.013
-6 (Roll s u b s i d e n c e )
0.986
.009
0.02
35
.25
-.05
- 53
.72
.05
-.031
-.007 -.165
I.
- .082
-.996
- .
T a b l e I 1 i l l u s t r a t e s t h a t t h e closed-loop e i g e n v e c t o r s have approached 1 t h e d e s i r e d mode-decoupled s t r u c t u r e . I n p a r t i c u l a r , t h e d e s i r e d m o d i f i c a t i o n achieved i n t h e e i g e n v e c t o r p a i r corresponding t o t h e Dutch roll mode should b e noted. The improvement i n t r a n s i e n t response c h a r a c t e r i s t i c s u s i n g t h e feedback c o n t r o l l e r i s i l l u s t r a t e d i n f i g u r e 1 . The r e s p o n s e c u r v e s demonstrate t h a t t h e c r o s s c o u p l i n g between t h e roll a x i s ( p , @ ) and yaw a x i s (Q,B) has been s i g n i f i c a n t l y reduced.
20
CONCLUDING REMARKS
New s u f f i c i e n t c o n d i t i o n s t o a s s i g n minimum (n,m + r 1 ) e i g e n v a l u e s by means o f o u t p u t feedback have been d e r i v e d . I n g e n e r a l , i n a d d i t i o n t o t h e assignment o f minimum (n,m + r 1) e i g e n v a l u e s , ( r 1) e i g e n v e c t o r s can b e p a r t i a l l y a s s i g n e d with a t most m e n t r i e s i n each v e c t o r a r b i t r a r i l y chosen. The u t i l i t y o f a s s i g n i n g b o t h e i g e n v a l u e s and e i g e n v e c t o r s f o r r e s p o n s e modific a t i o n is i l l u s t r a t e d by d e s i g n i n g a feedback c o n t r o l l e r f o r t h e l a t e r a l dynami c s o f an a i r c r a f t . The s y n t h e s i s a l g o r i t h m i s c o m p u t a t i o n a l l y s i m p l e and i n v o l v e s only t h e s o l u t i o n o f a l i n e a r system o f e q u a t i o n s .
Langley Research Center N a t i o n a l A e r o n a u t i c s and Space A d m i n i s t r a t i o n Hampton, VA 23665 December 19, 1977
21
APPENDIX A
REDUCTION O (C,A,B) TO CANONICAL F R F OM
. -
y = Step 1:
ez
(AI
Reorder t h e s t a t e v a r i a b l e s ( i f n e c e s s a r y ) s o t h a t t h e measurement m a t r i x
S t e p 2: Apply a c o o r d i n a t e t r a n s f o r m a t i o n
x = TOX
where
l o
and
1n-r
J
with
e,
E Rrxn-r
and
Es
as d e f i n e d i n e q u a t i o n s ( 3 ) .
The
transformed system
i
has t h e d e s i r e d form o f e q u a t i o n s ( I ) .
22
For completeness o f p r e s e n t a t i o n , an e x i s t i n g a l g o r i t h m ( r e f . 8 ) developed f o r e i g e n v a l u e / e i g e n v e c t o r assignment u s i n g s t a t e v a r i a b l e feedback i s d e t a i l e d i n t h i s appendix. E x t e n s i o n s are made t o a d o p t t h e a l g o r i t h m for t h e o u t p u t feedback cases d i s c u s s e d i n t h e main t e x t .
SPECTRAL SYNTHESIS ALGORITHM
A d i r e c t way o f c o n s t r u c t i n g a n o n s i n g u l a r modal m a t r i x i s t o g e n e r a t e t h e e i g e n v e c t o r s which s a t i s f y e q u a t i o n ( 6 ) s e q u e n t i a l l y and i n s u r e t h a t t h e y do n o t l i e i n t h e e i g e n s p a c e g e n e r a t e d by t h e v e c t o r s a l r e a d y s y n t h e s i z e d . The a l g o r i t h m p r e s e n t e d accomplishes t h i s c o n s t r u c t i o n w h i l e c o n s t a n t l y t e s t i n g t o i n s u r e t h a t t h e s e t of e i g e n v e c t o r s i s a l i n e a r l y independent set t o a degree determined by a numerical t o l e r a n c e parameter set by t h e d e s i g n e r . For c l a r i t y o f p r e s e n t a t i o n t h e a l g o r i t h m is d e t a i l e d f o r r e a l e i g e n v a l u e assignments. The f o l l o w i n g n o t a t i o n s are used throughout t h e a l g o r i t h m p r e s e n t a t i o n :
( 1 ) e j is an n-vector w i t h j t h e n t r y e q u a l t o 1 and a l l o t h e r e n t r i e s equal t o zero.
(2) v ; specified.
[z;
: w;]
is t h e k t h e i g e n v e c t o r , where
Q ( O ) = In and
Zk
is designer
i # 0 , is defined i n
Q(i),
(4)
V - (k-1)
Ok.
is the j t h e n t r y of
Vk(k-l)
A(1)
is the set of i n d i c e s ( 1 , 2 ,
. . ., n ) .
n and index 1 ,
an elementary upper t r i a n g u l a r m a t r i x o f o r d e r
...
(7)
Q(k-1) = M(k-1) , ~ ( k - 2 , )
- .
- 9
M(1)
(B3)
23
APPENDIX B k The a l g o r i t h m is now accomplished by c o m p l e t i n g s t e p s 1 t o 4 f o r 1 , 2, ., n and by computing t h e feedback g a i n s from e q u a t r o n (12) using C = In.
..
S t e p 1:
??Or
A = Ak, COmpvte
6, = d e t
A k
[Akin-, -
F]
6k = 0, p e r t u r b 6k
0 , proceed t o s t e p 2 .
S t e p 3:
For some
j E {A(k)},
and
( i i ) if
(iii) i f
a k = 0, s e l e c t a n o t h e r ak =
j E { ~ ( k > and r e t u r n t o s t e p
3(a).
for all
j E { ~ ( k > , go t o s t e p 4 .
S t e p 4:
For some
W k
{A(k>,
#
0, compute
and r e t u r n t o s t e p 4 ( a ) . A k t o (Xk
( c ) if , j ( k ) return t o s t e p 1.
for a l l
j E { ~ ( k i ) , perturb
+ EAk) and
24
APPENDIX B
(iii) The i t e r a t i v e procedure i n p a r t ii of s t e p 3 ( b ) a t t e m p t s t o meet exac.t e i g e n v a l u e / e i g e n v e c t o r s p e c i f i c a t i o n s . I n s t e p 4 ( b ) a n a t t e m p t is made t o meet e x a c t xk s p e c i f i c a t i o n s w i t h s l i g h t l y r e l a x e d Zk s p e c i f i c a t i o n s . The t e s t i n s t e p 4 ( c ) i n d i c a t e s t h a t t h e xk s p e c i f i c a t i o n implied t h a t t h e corresponding V k l i e s i n t h e e i g e n s p a c e a l r e a d y g e n e r a t e d . The t e s t t h u s demands an e i g e n v a l u e p e r t u r b a t i o n .
( i v ) The a k p r o v i d e a good measure o f t h e l i n e a r independence between e i g e n v e c t o r s , i f a l l v e c t o r s are normalized t o a s t a n d a r d b a s i s . S i n c e t h e d e t e r m i n a n t o f t h e modal m a t r i x V i s g i v e n by t h e p r o d u c t o f t h e ak, (k 1 , 2 , . . ., n ) , t h e numerical ill c o n d i t i o n i n g o f V f o r i n v e r s i o n can be e f f e c t i v e l y c o n t r o l l e d by s e t t i n g a t o l e r a n c e on t h e a k .
OUTPUT FEEDBACK EXTENSIONS
S i n c e t h e d i s t r i b u t i o n m a t r i x C is a l r e a d y i n t h e s p e c i a l c a n o n i c a l form, g e n e r a t i n g l i n e a r l y independent v e c t o r s t o g u a r a n t e e e x i s t e n c e o f feedback m a t r i c e s i n e q u a t i o n s ( 1 2 ) and ( 1 4 ) i s simply achieved by r e s t r i c t i n g t h e admiss i b l e pivotal indices A(k) ( e q . ( B 2 ) ) i n t h e s p e c t r a l s y n t h e s i s a l g o r i t h m as follows.
Case I:
Computing
E,
In equation ( 5 ) , l e t A r ( l ) be t h e s e t o f r column i n d i c e s Then apply i n g t o t h e r l i n e a r l y independent columns o f C ( I r ) . t r a l synthesis algorithm using instead of A ( 1 ) (notation k 1 , 2 , . . ., r . T h i s a p p l i c a t i o n g u a r a n t e e s t h e i n v e r t i b i l i t y i n equation (12).
Case 11:
Computing
E,
Equation ( 1 4 )
I n e q u a t i o n ( 5 1 , l e t A t ( 1 ) be t h e s e t of t column i n d i c e s n o t c o n t a i n i y l ; h e v e c t o r c ( e q s . ( 3 ) ) . Then a p p l y t h e s p e c t r a l s y n t h e s i s a l g o r i t h m u s i n g At. instead of A(1) ( n o t a t i o n ( 5 ) ) f o r k = 1, 2 , . t . This applicat i o n g u a r a n t e e s t h e n o n s i n g u l a r i t y of N1 l e q . ( 1 3 ) ) . F u r t h e r , s i n c e i n t h i s case CN i s rank t , t h e feedback m a t r i x K ( e q . ( 1 4 1 1 , computed u s i n g t h e re l a t ion
. .,
ECN
= Bl-I[Z(t)At
.- AlN]
25
APPENDIX B
to assign t s o l u t i o n for
26
REFERENCES
1 . Wonham, W. M.: On P o l e Assignment i n Multi-Input C o n t r o l l a b l e Linear Systems. IEEE Trans. Autom. C o n t r o l , v o l . AC-12, no. 6 , Dec. 1967, pp. 660-665.
2 . Srinathkumar, S.; and Rhoten, R. P.: Eigenvalue/Eigenvector Assignment f o r M u l t i v a r i a b l e Systems. E l e c t r o n i c s L e t t . , v o l . 11, no. 6 , Mar. 1975, pp. 124-125.
3 . Kimura, Hidenori:
4. Kimura, Hidenori:
P o l e Assignment by Gain Output Feedback. Autom. C o n t r o l , v o l . AC-20, no. 4 , Aug. 1975, pp. 509-516.
IEEE Trans.
A F u r t h e r R e s u l t on t h e Problem o f P o l e Assignment by Output Feedback. IEEE Trans. Autom. C o n t r o l , v o l . AC-22, no. 3, June 1977, pp. 458-463.
5. Hermann, Robert; and Martin, Clyde F.: A p p l i c a t i o n s of A l g e b r a i c Geometry t o Systems Theory - Part I. I E E E Trans. Autom. C o n t r o l , v o l . AC-22,
no. 1 , Feb. 1977, pp. 19-25.
A Algorithm for Pole Assignment n U s i n g Output Feedback. Proceedings: 1974: J o i n t Automatic C o n t r o l Conference, American I n s t . Chem. Eng., c.1974, pp. 309-312.
8. Srinathkumar, S.:
Ph. D. D i s s . ,
9. Davison, E. J . ; and C h a t t e r j e e , R . : A Note on P o l e Assignment i n L i n e a r Systems With Incomplete S t a t e Feedback. IEEE Trans. Autom. C o n t r o l , v o l . AC-16, no. 1 , Feb. 1971, pp. 98-99.
I O . Srinathkumar, S . ; and Rhoten, R . P.: Eigenvalue/Eigenvector C o n t r o l V i a S p e c t r a l C h a r a c t e r i z a t i o n : A A p p l i c a t i o n t o H e l i c o p t e r Hover Dynamics. n Ninth Annual Asilomar Conference on C i r c u i t s , Systems, and Computers, Shu-Park Chan, e d . , Western P e r i o d i c a l Co., ( N . Hollywood, C a l i f . ) , Nov. 1975, pp 605-609.
1 1 . Hartmann, Gary L . ; Hauge, James A . ; and Hendrick, R u s s e l l C.: CCV F l i g h t C o n t r o l Laws. NASA CR-2629, 1976.
F-8C D i g i t a l
27
III Il1 Il 1 I
10 I
-10I .5
A -n -
-.5
-.25
id
-3
t
Time, s e c
( a > p ( 0 ) = 10 deg/sec.
F i g u r e 1.- Comparison o f f r e e a i r c r a f t and augmented a i r c r a f t r e s p o n s e t o roll rate and s i d e s l i p s t e p d i s t u r b a n c e s . A i n d i c a t e s free a i r c r a f t r e s p o n s e and B corresponds t o augmented a i r c r a f t r e s p o n s e .
28
,-
100 I
-100I
A
B
50
r
I
1
10
= IOo.
F i g u r e 1.- Concluded.
29
1. Report No.
NASA TP-1118
4. Title and Subtitle
S. Srinathkumar
9. Performing Organization Name and Address
I
I
T e c h n i c a l Paper
5. Supplementary Notes
6. Abstract
The problem o f e i g e n v a l u e assignment i n a l i n e a r t i m e - i n v a r i a n t system u s i n g o u t p u t feedback is c o n s i d e r e d . New s u f f i c i e n t c o n d i t i o n s are d e r i v e d t o a s s i g n a n a l m o s t a r b i t r a r y set o f minimum (n,m + r - 1 ) d i s t i n c t e i g e n v a l u e s where n , m , and r are the number of s t a t e s , i n p u t s , and o u t p u t s , r e s p e c t i v e l y . These condit i o n s p r e c i s e l y i d e n t i f y t h e class of systems where such a n assignment i s impossib l e ; The s y n t h e s i s t e c h n i q u e a l s o h i g h l i g h t s t h e freedom i n s e l e c t i o n of c l o s e d l o o p e i g e n v e c t o r s under o u t p u t feedback. The u t i l i t y o f e i g e n v a l u e / e i g e n v e c t o r assignment i n t r a n s i e n t r e s p o n s e s h a p i n g i s i l l u s t r a t e d by t h e d e s i g n o f a c o n t r o l l e r f o r the l a t e r a l dynamics of an a i r c r a f t .
.
r.
Key Words (Suggested by Author(s))
Unclassified
Unlimited
S u b j e c t Category 6 3
-
Unclassified
Unclassified -~
29
~ ~
*For sale by the National Technical Information Service. Springfield, Virginia 22161
NASA-Langley, 1978
-I
.National Aeronautics and Space Ad m inis t rat ion Washington, D.C. 20546
Official Business Penalty for Private Use, $300
Postage and Fees Paid National Aeronautics and Space Administration NASA451
15
1 lU,G,
01 2 3 7 8 S O d 3 d j d S
i
(SUL)
I9R:
If Undeliverable (Section 1 5 8 Postal Manual) Do Not Return I ,