Jet time

signal
f Ct co 0 or an
consefentorl
EX Unit stop function 71 Ct 1 y
0 It
O

l a
Act 2
7 t
2

Linearity LTI what is a

Tine invariance system
U4
Causality yct

system mapping of input

signal
to output signal
5150 input
singe single output

Linearity
1 scale invariance

Uft yet
suppose
V Ct
G t
men If 8 multiplying constant

for any

superposition
u U
suppose TED y
17h y ya
vz y U2
time invariant

L 442 Inflict
tren V Ct Zyrs Y Ct 2ms
17h

causality
can't respond to
dy system signals in tire
future

unit set
impulse function I
limit i 3
impossibly
thin f
def 3 0
I impossibly
tall T
3
Area I E t
o

Tilts Sct

Sct
II
unit impulse
response fine

1µm
t

not

j U JJ Udt dtz

.ru

dynamical systems integrate her input

Uct
4 from scaled Sct s is the same as
from
superposition
impulse response
convo tutor

yet
ft CT n t I DT

Laplace Transform L
s
possibly complex
L gets Jet e stat Fcs

Another Embarrassment

Y u
III y ten in 1h1 ii
pass through

leprace Transform
gets e state
LC gets
Iff Fcs

inverse f Fcs T c
get T2 E

weanty L cg q g F Cs t
q Fels
L sets l L Ceos wt It
w
w

L Ict T L Cs in wt's S2 t w

Ias
at
L e Lce at
smut we

at Sta
L e cos wt y Sta Z t WZ
Anti Convolution Thins

yct ffgc.cz hit E at

a
7 s HCS VCs Cin S domain

pointwise ins

7 me Dame in
Laplace Domain
he y u 117 7

HCS L htt
system transfer functions
I gfCT DT
Ffg Ys is the time
integral operator

Fief III EEP's 977 7 1

I pure a

integrator o
ji
o o

I gets SFI FL E

Iffy
tnm
1St unit conefence
pure impulse Ys s 1
differentiator

partial Fracten Expansion

MATLAB can do partial
symbolic toolbox
fractions
EI FCS est jest y poles of system 5 l se 3
 I
Gcs poles s I zj
52 4
l
FCS GCS
Ste Sts Sty

lust 1 165
j 263 521 4

Feedback
e
rots T IFT s ye s emer
n T ects e ret
yet
Ks
H L
Rcs Ecs
n
1HI Yes yes Hcs EG
His pts Vs

Ks It Hcs I Hod RLS

I Ya
III L fast
closed loop
EX not Ict
e Hcs L
rct 3 Ict RLS Yes

47 1 s

3 E E

I yet 3 se te 3u E
 e
ref lcatn.ve s
actvaIs plantT
p IsensorT
Ym y

sensors
sensors ideally a
gain
y a sensor would have y
good
close to y s
t
low pass behavior 4 s
high bandwidth
IX Ym
low bandwidth
know um 8 00
w Full model
u
g lo
l O Om 40ns _xo Ifm i k get

cow pass to Sle snot 4mi

8
x
One co Om
Yo Ijm
actual
On measured
X bandwidth IRC far 12C filter
frequency above which can
go thru

a
tie unit
given ramp response
ramp Ct 2 t 2 Sint

I
net vs Hls

L Zant 2 1
 Fpa E 2.15

DC motor

It I v n o

GE EE TE E
f 1
x
Eg
I E t I

t.EE 0 I 0
ffEttFI7i
Saturation actual sat C commend Vmat

V actual
may

command
i
rat

HAT CAD Code

abs CV command
actual min Vinay signCVcommaner
Ss TF
control Uct u tr order LTI
fare
2 undamped resonant modes
K X u
Pi
M corresponds veo
MYZf
O 0
pz o w

P2
equilibrium
p
define it z Kit Kz

physics based EON s

Mi Pil t Ki z p KzPz U
Mz z th z p t Kzpz O
 choose
f I Piz P Pz JT

g g IF i

l tt O
O l
O O
O O ti tn o

I A X t B U

07
1515
choose Pa ee r
y O O G I T putty
Piggy sensor on Pz
C

I Axt Bu
u h
y c x Y

M S2 t k Pi S
iz Kz Pz S u Cs
mz 5 t Kz P2 CS k Pics 0

h Stk iz Mz SH K 2
p y K pus VCD substitute to
Kz
remore p i

Tf S4 Cm t m 2 12
s t K PzG us

S4 azs a o
pz s bo Us

Ks
F sb azs transfer furuta's
hi
14
5smztT fc s y cruet is yes
5
HLS

if HIE't
 Vs
by.si asI p szIggL yes Hls Ucs

y tajy aoy bow

choose X I Z Z TT
L

t tl O
ai
l O O H p O
0
O l O

4
0000.431
lbIH s 7
Control canonical form T
Taz5ta_ IBI

esgn
R t ii
y 9
closed loop
transfer fine

R
it 71 TEI
g

Y GDceCR i
5
4

taxes c
vet Is ideal s

L r o y stays small
STABILITY 1 yet
t really
usually fast
dangerous
44 3 Yet
L r mores
y follows it really slow
liraculHGI wtf'sTheFERGIE
etc Mirin soapresponses
 HCS KCS z Cs zz P poles roots of denominator
2 Zeros croots
S p 2C s p y numerator

Ex STU z 6
Cst 2 St 13 I 2
p 13

decaff exponential
I
52 16 undamped harmonic

sztzzwn.tw zmderdanpedoseiuatam
1 023
2h05 scale to modal participation
Ex 3.25 It 251 1 2 Cst 12 wren polls have
51 357 2 ti Cst 2 real
negative
parts stable
PFE t
2 7 yet e t z e at

1
Hz
5 1 5 2
of
shapes
n

µ staronty is
governed by
L I f pots but zeros determine

µ
Stl 572
t
Shape
Yet e e
Im
Hr only O zero
X X O the poles
z I

IX Ls it stable GD KP ICTs b
t D i T it D
7
01 71 b
KP
Tsxb

g Jst Cbt Kp
stable

D i Kp 70
D2 T Kds cysts
Dz i Kd S a o t Kd ICTs tb
Dz Kits Cki 70 Kds
Dt ka Stb stable
 D Kits Tests
y
X 3.26 2 Iz I t kits stb

Hls 2 1 P z Ki
52 25 5
L
4 zj stable
Sta e Ts T DS t ki
9
as long as his
FE Stl
z
Tty E Stl 2 4
Vale es pas stance

tilt e e
f0 zcuszt
1
2 sunt
I 52 25 5 as It 23Wh Stunt
X z

i 0
i
Wn TE 2.24
boys
R
i I i L
x 2 TE
e est
Wh Wn Tf 3T 2m04s
L JI
o E s wn I
determines scaling
exponenton y
seems.io
uncover.no
Iii
r tr rise time ieo
t
l Is time
I i Settling
tr ts

approximations zpoles typical

tr l
91Wh Mp
w
s 4 610
T 31h37 206
Mp e
445 3
Fig 3.24
i
i want
tr ur il poles
I damply X I
close
X intros
Overshoot
l l tooigns
1 to makes s Sto
X i as
k Hp 3 t
possible so
X don't break
I any tug or

busamore
poles moststay out
setts x
of grey zones to meet
time
o specifications