06/03/2017

UWE Bristol Today’s Lecture

Industrial Control • System modelling – determining
UFMF6W-20-2

Control Systems Engineering differential equations

UFMEUY-20-3
• Using the Laplace operator
Lecture 2: Modelling Systems • Determining the transfer function
• Standard forms

© 2017 University of the West of © 2017 University of the West of

England, Bristol. England, Bristol.

System Modelling Mechanical Systems

• Determine differential equation that • Force (F)-displacement (x) relationships
represents the system • Spring: coefficient k
• Mechanical or Electrical FS = kx
• Standard equations for both • Damper: coefficient c
dx
• Analogies (covered next week) FD = cx! = c
• Mass (inertia): NSL: ma dt
d 2x
FI = m!x! = m 2
dt
• Also apply to rotational systems
© 2017 University of the West of
England, Bristol.
– Force à Torque; x à θ; mass à J (MoI)
© 2017 University of the West of
England, Bristol.

Mechanical Example Electronic Systems

• Mass/spring/damper system • Voltage (V) - Charge (Q) - Current (i)
– Relate F and x d x 2 relationships
Inertia : FI = m!x! = m
dt 2 • Capacitor
Q 1
dx VC = = ò idt
k Damping : FD = cx! = c C C
dt
• Resistor
m F
Spring : FS = kx dQ
VR = R = iR
c
d 2x dt
å F = m dt
x
2
= F - FD - FS • Inductor
d 2Q di
d 2x dx
VL = L 2 = L
F =m + c + kx dt dt
dt 2
© 2017 University of the West of © 2017 University of the West of
England, Bristol. dt England, Bristol.

© 2017 University of the West of

England, Bristol. 1
 06/03/2017

Electronic Example s-domain

• LCR Circuit i • Differential equations can be awkward to
– Relate Vin and i deal with
di 1 • Using Laplace operator, ‘s’, can help in
Cò
VL = L VC = idt VR = iR L
dt Vin
manipulation – using Laplace transforms,
C LDE are converted to simple algebraic
Vin = VL + VR + VC
equations
di 1 R
Vin = L + iR + ò idt • Transient and Steady state components of
dt C the solution can be obtained

© 2017 University of the West of © 2017 University of the West of

England, Bristol. England, Bristol.

Laplace Transforms Laplace transforms

• Definition of Laplace Transform of a • Derivatives
function f(t):
é dx (t ) ù Multiplication by ‘s’ in the Laplace

ú = sX ( s )
¥
Lê domain corresponds to differentiation in
L[ f (t )] = F ( s) = ò f (t )e dt - st
ë dt û the time domain

0 Replace occurrence of d/dt with s:

é d 2 x(t ) ù
• Complex variable, ‘s’, is Laplace operator Lê 2 ú
= s 2 X (s)
ë dt û é d n x(t ) ù n
• More details in your Maths notes Lê n ú
= s X ( s)
é d x(t ) ù
3
ë dt û
Lê 3 ú
= s 3 X ( s)
ë dt û
© 2017 University of the West of © 2017 University of the West of
England, Bristol. England, Bristol.

Laplace Transforms Mechanical Example

• Integrals • Mass/spring/damper system
k
• Integral in the time domain is converted to
division in the ‘s’ domain. m f(t)

é¥ ù F ( s) 1 0+ c
x(t)
L êò f (t )dt ú = + ò f (t )dt
ë0 û s s 0- • When initial conditions are zero (x(0) = x!(0) = 0)
• Second term almost always zero, hence: 2
d x(t ) dx (t )
m +c + kx (t ) = f (t )
é
¥
ù F ( s) dt 2 dt
L êò f (t )dt ú =
s ms 2 X (s) + csX (s) + kX (s) = F (s)
ë0 û
© 2017 University of the West of © 2017 University of the West of
England, Bristol. England, Bristol.

© 2017 University of the West of

England, Bristol. 2
 06/03/2017

Electronic Example Transfer Function

• LCR Circuit i(t) • Represent the input-output relationship for a
di (t ) 1 component
vL (t ) = L vC (t ) = ò i (t )dt • Expressed in terms of ‘s’
dt C L
R(s) C(s)
vR (t ) = i(t )R v(t) G(s)
Note: C ( s) = G ( s) R( s)
C
di (t ) 1
v(t ) = L + i (t ) R + ò i (t )dt • Transfer Function:
dt C R output C ( s)
= = G(s)
1 input R( s)
V ( s) = LsI ( s) + RI ( s) + I (s)
Cs
© 2017 University of the West of © 2017 University of the West of
England, Bristol. England, Bristol.

Transfer Function Transfer Function

Qin ( s ) = AsH ( s ) + kH ( s ) • Mass/spring damper
• Single order Qin ( s ) = ( As + k )H ( s )
ms 2 X (s) + csX (s) + kX (s) = F (s)
Flow in
Qin ( s )
qin(t)

H (s)
= As + k (ms 2
)
+ cs + k X ( s ) = F ( s )
h(t)
F (s)
Flow out H (s) 1 = ms 2 + cs + k
kh(t)
= X (s)
Qin ( s ) k + As
dh(t ) X (s) 1
qin (t ) = A + kh(t ) H (s) 1 1 =
dt = F ( s ) ms 2 + cs + k
Qin ( s ) k 1 + A s
k X (s) 1
=
F (s) æ c m ö
k ç1 + s + s 2 ÷
© 2017 University of the West of
England, Bristol. è Bristol. k ø
k
© 2017 University of the West of
England,

Transfer Function Standard Forms

1
• LCR Circuit V ( s) = LsI ( s) + RI ( s) + I (s) • First order Gain
Cs
æ 1 ö Cs C ( s) 1
V ( s ) = ç Ls + R + ÷ I ( s ) Multiply by =g
Cs ø R( s ) 1 + ts
è Cs
V (s) 1 I ( s) æ Cs ö
1 • e.g. Tank: Time constant
= Ls + R + = ´ç ÷
I (s) Cs V ( s ) Ls + R + 1 è Cs ø
I (s) 1 Cs H (s) 1 1
= = Gain = 1/k
I ( s) Cs Qin ( s ) k 1 + A s
V ( s ) Ls + R + 1 =
Time constant = A/k

V ( s ) 1 + RCs + LCs 2 k
Cs
© 2017 University of the West of © 2017 University of the West of
England, Bristol. England, Bristol.

© 2017 University of the West of

England, Bristol. 3
 06/03/2017

Standard Forms Standard Forms

• Second order • Second order
C (s) 1 • e.g. Mass/Spring/Damper
=g
R( s) 2z s2 k
1+ s+ wn =
wn wn2 æ ö m
wn = undamped natural frequency [rad/s] X ( s) 1 ç 1 ÷
= ç ÷
ζ=
c k
ζ = damping ratio [] c m
F (s) k ç 1 + s + s 2 ÷
ç ÷ 2k m
g = gain è k k ø
1
g=
k

© 2017 University of the West of © 2017 University of the West of

England, Bristol. England, Bristol.

Today’s lecture
• Standard equations for Mechanical and
Electrical systems – these will be supplied
• Use the Laplace operator, ‘s’, to simplify dealing
with LDEs
• Transfer functions represent input-output
relationship (output/input)
• Standard forms help predict performance
• Tutorial Sheet 2: On Blackboard. Determining
transfer functions for systems.

© 2017 University of the West of

England, Bristol.

© 2017 University of the West of

England, Bristol. 4