CH 5
CH 5
CH 5
F (θ , t ) =
34 N
e
I m cos(θ ae − ωet )
2π P
a
[Tqd 0 (θ )] = 23 sin θ sinθ − 23π sinθ + 23π [Tqd 0 (θ )]−1 = cosθ − 23π sinθ − 23π
1
1 2π 2π
cos θ + 3 sin θ + 3
1 1 1
2
2 2
(
p [Tqd 0 (θ )]−1[λqd
s ]
0
)
- sinθ cosθ 0
2π 2π dθ qd 0
= − sin θ − cosθ − 0 [λs ] + [Tqd 0 (θ )]−1[ pλqd
s ]
0
3 3 dt
2π 2π
− sin θ + cosθ + 0
3 3
Energy Conversion Lab
CIRCUIT MODEL OF 3φ INDUCTION MACHINE
0 0 0
3 3
L +
ls 2 ssL 0 0 2 Lsr 0 0
= 0 Lls + Lss 0 isqd 0 + 0 0 irqd 0
3 3
Lsr
2 2
0
0 L
ls
0
0 0
3 3
2 Lsr 0 0 Llr + Lrr 0 0
2
3 3
= 0 Lsr 0 isqd 0 + 0 Llr + Lrr 0 irqd 0
2 2
0 0 0 0 0 Llr
Energy Conversion Lab
qd0 EQUIVALENT CIRCUIT IN ARBITRARY FRAME
ωλds L (ω-ωr)λdr
L
ωλqs L L (ω-ωr)λqr
Torque equation
resistive element is with copper loss
reactance is with magnetic field energy
speed voltage term is with mechanical work
we can express torque with speed voltage terms
definition of speed voltage
Eqs = ωλds, Eds = -ωλqs
E’qr = (ω-ωr)λdr, E’dr = - (ω-ωr) λ’qr
steady-state torque is the real power absorbed by these
four speed voltage sources
Tem =
3 P
2 2ωr
[
Re (Eqs − jEds )(iqs − jids ) + (E ' qr − jE 'dr )(iqr' − jidr' )
* *
]
Energy Conversion Lab
Project 6-1 Arbitrary Frame Transformation
3vsg = (vas+vbs+vcs)-(vag+vbg+vcg)
psiqr
vag Term
Vm*cos(u[1]) Qaxis
Fcn vqs Tem ias
vbg Product
Vm*cos(u[1]-2*pi/3) vds wr/wb ibs
Fcn1
v0s ics
vcg Product1 Rotor
???
Vm*cos(u[1]+2*pi/3) abc2qds qds2abc Sum
psids Tmech
Fcn2 ids
psidr
Term1
Daxis i0s
Induction Machine Simulation
in Stationary Reference Frame
Zero_seq
2 Mux (u[3]-u[2])/sqrt(3) 2
in_vbg Fcn1 out_vds
3 (u[1]+u[2]+u[3])/3
in_vcg 3
Fcn2
Mux
out_v0s
1 vsg
4 50*Zb*wb Sum
s
ias+ibs+ics Integrator
1/Csg
2 1 1
vqss = vag − vbg − vcg
3 3 3
vdss =
1
(vcg − vbg )
3
v0 s = (vag + vbg + vcg ) − vsg
1
3
d
vsg = 3( Rsg + Lsg )i0 s
Energy Conversion Lab dt
SIMULATION OF INDUCTION MACHINE
Inside the qds2abc block
(6.109)
u[1] + u[3] 1
1
Fcn out_ias
in_iqs
3
-(u[1]-sqrt(3)*u[2])/2 + u[3] 3
in_i0s
Fcn2 out_ics
Mux
(6.112) (6.115)
psiqs Mux iqs
1 (u[1]-u[2])/xls 2
1 Mux wb*(u[2]+(rs/xls)*(u[1]-u[3]))
s out_iqs
Fcn4
in_vqs psiqs_ Mux4
Fcn
Mux
(6.117) psiqm
Mux xM*(u[1]/xls+u[2]/xplr)
Fcn3
(6.113)
Mux3 4
1
in_psids 1
2 (6.118) out_Tem
startup
Rotor frame
( ) ( ) ( )
v s r = vqrr − jvdrr e jθ r ( t ) = Vmr e j ( sωet −θ r ( 0 ) −δ ) e jθ r (t ) , i s r = iqrr − jidrr = I mr e j ( sωet −θ r ( 0 ) −δ −φr ) e jθ r (t )
rotor voltage equations, current equations
(
v r = v − jv s
qr
s
dr )=V mr e j (ωe t −δ )
, i
s
r = iqr
s
− jidr
s
= I mr e − j (δ +φr ) e jωet
the rms time phasor
~ V ~ I at steady state : θ r (t ) = ωe (1 − s )t + θ r (0)
Vas = ms e j 0 , Ias = ms e − jφs
2 2
~ V ~ I
Var = mr e − jδ , Iar = mr e − j (δ +φr )
2 2
rms space vector is rms time phaser times ejwt
vqss − jvdss
~ jωet s iqss − jidss ~ jωet
V − jV =
s
qs
s
ds = Vase , I qs − jI ds =
s
= Iase
2 2
vqrs − jvdrs ~ jωet s iqrs − jidrs ~ jωet
Vqr − jVdr =
s s
= Var e , I qr − jI dr =
s
= Iar e
2 2
Energy Conversion Lab
DERIVATION OF STEADY STATE MODEL
Stationery qd voltages/currents expressed in rms qd0
voltages and currents
v s = vqss − jvdss = Vms e jωet , i s = iqss − jidss = I ms e − jφs e jωet
v r = (vqrr − jvdrr )e jθ r ( t ) = Vmr e j ( sωet −θ r ( 0 )−δ ) e jθ r ( t ) , i r = iqrr − jidrr = I mr e j ( sωet −θ r ( 0 )−δ −φr ) e jθ r ( t )
Phasor voltage equations (see pg. 185)
Vas = (rs + jωe Lls ) Ias + jωe Lm ( Ias + Iar' )
~ ~ ~ ~
developed
mechanical
power
developed
mechanical
power
Vth2 (rr' / s )
rr'
3P when = rth2 + ( xth + xlr' ) 2
Tem = smax
2ωe (rth + rr' / s )2 + (xth + xlr' )2
high torque
Zin
lower current
lower torque
Zin
f ( x0 , u0 , y0 ) = 0
Linear system analysis
neglect higher order Δ terms, we can rewrite the governing
equation into a standard state-variable form
∆x = A∆x + B∆u
∆y = C∆x + D∆u
with state-variable form, we can obtain the transfer function,
analyze zero and pole of the system by changing system
parameters
Procedure to obtain state-variable matrices of [A,B,C,D]
using steady state analysis to initialize x0, u0, and y0
complete the steady-state of the SIMULINK system at some desired
operating point using trim function
use linmod function to determine [A,B,C,D] matrices of the small-
signal model about the chosen steady-state operating point
input and output port numbers are required by trim and linmod to
obtain the matrices
once [A,B,C,D] matrices are available, use ss2tf.m and tf2zp.m
functions to obtain transfer function, zeros and poles of the system
Linearized system analysis (example)
% M-file for Project 4 on linearized analysis in Chapter 6
% To be used in the same directory containing the SIMULINK
% file, s4eig, of an induction machine in the synchronous
% reference frame
% After initializing the parameters of the imwe in the Matlab workspace, make initial guess of all values in the state (x), input (u), and output (y) vectors.
From the diagram of imwe, we see that u = [vqe; vde; Tmech], y = [iqse; idse; Tem; wr/wb]
Psipdro = 0;
wrbywbo = 1;
use this function to find actual ordering of the states x
[sizes,x0,xstr] = s4eig([], [], [], 0)
% which yields xstr =[psids,psidr‘,psiqs,psiqr‘,s] or x = [ psids; psipdr; psiqs; psipqr; wr/wb ]
u = [Vm; 0; Tmech];
x=xg;
y=yg;
% use index variables to specify which of the above input in the initial guess should be held fixed
% Use Matlab linmod function to determine the state-space representation at the chosen operating point
% dx/dt = [A] x + [B] u
% y = [C] x + [D] u
[A, B, C, D] = linmod('s4eig', x, u);
% Use Matlab ss2tf to determine transfer function of the system at the chosen operating point If desired, transfer function can be printed using
[numt(index,:),dent(index,:)] = ss2tf(A,bt,ct,dt,1);
printsys(numt(index,:),dent(index,:),'s')
% Use Matlab tf2zp to determine the poles and zeros of system transfer function
[zt(:,index),pt(:,index),kt(index)] = tf2zp(numt(index,:),dent(index,:));
% Use Matlab ss2tf to determine transfer function of the system at the chosen operating point.
[numv(index,:),denv(index,:)] = ss2tf(A,bv,cv,dv,1);
printsys(numv(index,:),denv(index,:),'s')
% Use Matlab tf2zp to determine the poles and zeros of system transfer function
[zv(:,index),pv(:,index),kv(index)] = tf2zp(numv(index,:),denv(index,:));
end % end of for Tmech loop
Linearized system analysis (example)
y(1)
u(1) iqse
y(3)
u(2)
y(2)
y(4)
idse
u(3)
Linearized system analysis (example)
% Print loading level and corresponding gain,zeros, and poles
% of (Dwr/wb)/DTmech
index = 0;
for Tmech = T
index = index + 1;
(Dwr/wb)/DTmech
Tmech loading is -7.92e+001
Gain is 1.50e-001
Zeros are:
-9.639e+001 3.598e+002i
-9.639e+001 -3.598e+002i
-6.697e+001 2.765e+001i
-6.697e+001 -2.765e+001i
Poles are:
-1.084e+002 3.636e+002i
-1.084e+002 -3.636e+002i
-2.108e+001 1.672e+002i
-2.108e+001 -1.672e+002i
-6.770e+001 0.000e+000i
Linearized system analysis (example)
% Print loading level and corresponding gain,zeros, and poles
% of (Dwr/wb)/Dvqse
index = 0;
for Tmech = T
index = index + 1;
Zeros are:
-2.972e+015 0.000e+000i
-1.865e+002 0.000e+000i
-1.236e+001 5.388e+001i
-1.236e+001 -5.388e+001i
Poles are:
-1.084e+002 3.636e+002i
-1.084e+002 -3.636e+002i
-2.108e+001 1.672e+002i
-2.108e+001 -1.672e+002i
Project. 6-4 Linear Analysis of Induction Motor
Use the simulation model parameters as page 224 to run the
simulation in synchronous frame as follow:
1. (iqse/vqse) when machine is running at steady state with mechanical
torque Tmech=-Tb (numerically negative-motoring), the source voltage
suddenly drop from 1pu to 0.8pu at 0.8 sec, plot iqse , vqse, iqse/vqse with
respect to time, observe and discuss your results
2. (Tem/vqse) when machine is running at steady state with Tmech=-1Tb, at
0.8 sec, the source voltage suddenly drop from 1pu to 0.8pu, plot vqse,
Tem, Tem/vqse with respect to time, observe your results
3. (linear analysis) use linearized model to analyze the transfer functions
of Δiqse/Δvqse, ΔTem/Δvqse.
a) Plot the zeros and poles of the transfer functions and draw the
root-locus plot, analyze your plots
b) Using the transfer functions you obtained in 3. to plot (iqse/vqse ,
Tem/vqse) with respect to time. Compare your results with the
ones you got from 1 and 2.
Linearized system analysis (optional)
Investigation on effect of (B1*Kagc) in isolated area’s frequency bias supplementary
control performance
Background:
The block diagram of frequency bias supplementary control for one area is shown as
Fig.1. The assumption is made under the following condition: a large disturbance, DPL,
(step function) occurs and induce a generation response at output, DPG.
1. using Mason’s formula to obtain the transfer function T(s) = DPG/DPL.
2. using the derived transfer function to draw root-locus plot.
3. using MATLAB/Simulink tools to “estimate” transfer function T(s) = DPG/DPL.
4. using the T(S) obtained from “3.” to draw root-locus plot
5. compare the T(S) you obtained from “4.” with the one in “1.”
DPL DPG
Linearized system analysis (optional)
Background:
6. vary “Bias Coeff” B1 from 29.4 with step increase 10 until the system
become unstable, draw the zero and pole plots, show the pole trend, and
determine the value of B1 that starts to become unstable
7. system parameters are as follow Tch=6; Tg=0.2; T=7.52; R1=0.05;
M=39; D=4.41; B1=29.4
DPL DPG