ISTATE SpACE ANALYSIS
OF CONTROL SYSTEM
The Representation of a wel ordec clifferential eqn
based on state Space opproach is olotatyed below —
2x Ct) da(ty =
0 ae +e SE rext= ult) —o
if tnittal condition Given as —
dz 26 and xy =0 at t=o
lt
Pe) LoL ax
cle Fora t a ee ee)
let 2xQ@) = x, and dxlw _ x
—— = SO, block
%, 0, Cc! chegran
eax
>
TRE dynamic €q7 can be wathe ag —
X= % —@®
feb i)
go ZIT Be Rah aq ult) —&
ta Metrix form [] =
tL = /0 ue o °
a 1
iG hsb [-]+ 4 [ue
a a a Eq: is Known as Atte &pace Representatioy of
Syslem Given by eqn @-
¥otate Space Re presentattoy of nt oyder different,
equatioy —p
“I
| a Me tiy op amy = uly
aen bent o
S Enittel » lyr dry S
Conditons en We aarti ee
Given.
Also “Terms yo, Gl, <= yn Ut) can be
Considered as Q set of nm stale Variables.
Now = Z
ee d= % ey Sey
Y= Can Wai ten ae xX
' ag a
i AY
ge = Xp Sn-— = ey,
A =
Iso 2%, = An, mea — 9) 2n-) =o
A General Ge ee of Above eqn
Can expressed tn -form of- state equatoy as_
m= An Bu
Here» = a And Az 0 1--0d _
A © o-~Lol? B=] 0
: 9 e-04 dag
in uo
vee KA} 9 "
Xe Qt Ah ain
coh vo +) wees)
U— Conto| (scalar) bY The output eqn is Given by —
x
ieee oro oJ %
ial ‘ ——
©
Cx
(Len Matrix Scalar J
gesiate space Reprecentatioy -for electrical NI» —>
>i —¥4 Ye
BY The eqn of dy naaic behaviowr of Syslem ae -
Vigo e, avn +t =o —O
R abt
}
s dvs War a0 ®
at Ra
a o4y-y =O ®
ae
By Reanraugirf atl eqn we finds —
vi ¢ €
dv) - -s= - = RG
ree icy
ee
at Ro Cn —®
oe ieee
at =
+ let zr = My
X25 Vo
X = t
se yo tt xy —-+—x% + e
| — Racy CL Rich
° _— 4 .
' % = iT 2 x~3
ee aucs Co
“re = tee he
2% Fae
Tet vector-mertrix form gtate space Repyesentatioy
x pee o -t |)
i
EC = Rig ' xy
2
: ° = + a
4, 2a Se
nee ei 0 oe |
L [E, |
|
+ ft j
Rac. —t
aca} 6 (
° tt fs fa relatiou to tronsfx of a sys lem -from one
| state +0 another by appropriate input Conte) in
a finite time.
Now taxing a block diagram Representatoy os —
OG) eS 2oe i ee ta —@
= 2%
diagram it Ls noted that Stocte x, is
operate x fom bs initiqs state x2 (to) +o final
fe x2 (+5)
w ler, Consider Follows state equations as ~
% = AL+ Bu —-@
¢ = ct + dU ——@® we velocity vector (nx/)
2% = State vectoy (x1)
i = inpar Viecho? (x43 |
¢ = owipur veetor (px4)
A, Bi Ci D = Marefy (nny 1X. PX, Pxkn
| a Ae spectively 5
| ele) x(n)
| x
xu Couhel] of State Vector.
| oe *O) = Arey + BUlo)
~@) = Ax + Buly = ala x0) + BA(o)]
+BuUL) |
x1) = AxXQ) + Bua) = Ara 2 FBUu(I)
: + Bu)
xin) = AY x(0) + A! BUlo) FAT ?BU (I) + +84]
. Ulm)
"7 = : 1A2g:--- Ane,
This x(n) - Px) = [8 06:Ade: ~
Tae ours Con cezned ane the obsecvaritifg
| ey eee
Equatioy x, = —2X, FU ——O
% = 2H rk, ——O
Here fate x, not connecia| to ouput =b g
meastured » %)\ is Known but x
Cann't measures,
So only Lt Con be Observed.
“$M = AX+BU —@
YS pce. —@
Where 2. = velocitG veetor (nx1)
input vector (x1)
"
stot veetor (nXx))
outpet vector (px)
Pp ack &
ul
,B,C = matasx (nen, nxm, Pxn)
By definitioy Yo) = Cx(0)
ya) = Cxa) eg Bye 4x00)
YQ) = cA? (Xo)
j —®
S-1) = can! w(o) r ¥(o) c
ya) a
vi) | car | % 2)
¥(n-1), car-]
y unique dleterminactioy of state. x00) tue NaTEX
Eni st have jnverse, —
c oo. ele have Q vouk=n
so
cae fe or fs a square Mabex
A
; so rt chould be non singub
carl for ovservabitily *
> que Mataix is colled observabitity tet alsin:
ane it con sewaite as —
Tr atcl: i tat!
chia’ : (AD) oT, (A) OT
oh =
r+ obecsva itity cond tHoy fs satisfied v
be of 6ank7 ay fe sald thot ca ts
observable palas { .
Ex» check observatitiy tr
X= Mr
X = —2x,—- 3% +u
it
Y= xy tee,
2
{3
=
grok
eo
"
CC; 4
boo
ae te
—
RoR
Cae
eas
7
—
ae)
a,
J
I
r Co
oa |
—- Oo 1
ee
1 oe
be
uy 3
. a.
= ics a!
in
Cc. -
qe
csi al ti
<
"
(eal
°
a
DR,
8
= 1 a
i 1
b tu Aowk of Matix Is
not equal to n
So Ur is Unobservalle sys Transfer Mactsin —>
E
Re Motels eleting laplace +aausform of owlput +o
laplace traneform rtof input of State space Represeatatio,
of 4 Control System Called transfee Matrix
1 Tle State eq of a syslem is given as —
wile Aut Bu = 70
yo= Ger ou ——& |
i
state Vector (X))
= inpat'; y= output :
Where
a
‘matrix (nxn)
Matrix (1¥n)
= ¢calan Matrix
Ci 2 ok
a
ow taking lap lace 4ransformatioy of Both sidle of
suotions @ and @ —
Sx(s) xo) = ARG) TF avis) —@
yis) = Cx + BUCS) —©
AS per definitioy of trousfer #7 %(0)=0
Sx(s) = AX(s) + BULS)
a + DU
and Als ui) = (st-Ay' BU(S) puce |
yids Value tu eq” @ We have- y(s) =[¢ (sT-ay! B+ DJ U(s)
4 ° _ Ys) ade Nal]
Taansfer functong GCs) = TOW) Neb.
ats) = MS = e(st-aylg —@
v(s)
| Te afer £9 in
3
AW = AN-GA fo,
& . -
The state model of SI¢o (single -input- single —cutpur)
is. Given by -
x= Ax + BU
i Y= cx
Choracteristics polynomias of cys le is
Jazt-aA] =o
Tae eisay values of System Mats, A ate Aoots Of
Cchanacieuist’s eqn of closed [oop Poles of Systea-
here system matrices A&B should form 4
| contobabie paix for applying Stare variable ‘feed back:
Ww
8
Ko
i
w= Axt
Ww ed
e@ = Ax t+ B(4-kKX)
= (A-BK)x + BA
\
system :
Motux Tex cigan values oF Moth (A-BK) Cau ea
ad?u sted by ey tee feedback Gaim Mate k
Now on Opplicetioy of State variable feedback +H
Characteristics Polynornial is -
[st - CA-8x)| —o
Here feedback, Gain mebtly K Cau be determined
&) Dekrmine Her“ Chancicterst os pelmorial aL
| pet desi rec Closed —loop pole tocattey.
Le Petermine characteristics _polynomia| in
Wot j
terms: ‘OF “Feedlloack os Halal k.
\— Compare cofficieuts of Oia G
polynomiat we sRe DM aol @ rte
feedback Gain Maliiy k lofted.
Maiti.
