Transcript Fuzzy Logic Based Supervision of a DC Link PI Control in a

FUZZY LOGIC BASED SUPERVISION
OF DC LINK PI CONTROL IN A DSTATCOM
Harish Suryanarayana
Doctoral Student
Energy Sources and Systems
Purdue University
OVERVIEW OF THE PRESENTATION
 Custom
Power
 The Distribution Static Compensator
 Symmetrical Component Theory
 Fuzzy Logic
 Fuzzy Logic Based Supervision
 Simulation Results
CUSTOM POWER
It is a concept based on the use of Power Electronic
controllers in the distribution system to supply valueadded, reliable, high-quality power to its customers.
 Power Electronic Controller – DSTATCOM, DVR
 Distribution level : 1kV to 38kV
 High Quality Power: No sags/swells/harmonics

THE DISTRIBUTION STATIC COMPENSATOR

An SSG or a Static Synchronous Generator is defined by the IEEE as a
self-commutated switching power converter supplied from an
appropriate electric energy source and operated to produce a set of
adjustable multiphase voltages, which may be coupled to an AC
power system for the purpose of exchanging independently
controllable real and reactive Power.

STATCOM : An SSG with a capacitor as the energy source is known as
a STATCOM or a Static Compensator.

DSTATCOM : When a STATCOM is used at the distribution level or the
load end, it is known as a DSTATCOM or Distribution Static
Compensator.
DSTATCOM
Courtesy : PSERC 2003 Seminar
DSTATCOM – MAIN GOALS
To cancel the effect of harmonics due to load so
that the current drawn from the source is nearly
sinusoidal .
 To help maintain near unity power factor by
canceling the effect of poor load power factor
 To help offset the effect of unbalanced loads,
such that the current drawn from the source is
balanced.

DSTATCOM SCHEMATIC
vsa
Ls
Rs
isa
PCC
vsb
isb
ilb
vsc
isc
ilc
Rf
Lf
N
A phas e
H-Bridge
S1a
Za
Zb
Zc
ifa ifb ifc
S3a
S1b
S3b S1c
S3c
if
S2a
S4b
S2b S4c
i Cd c
vdc
L nl
S2c
n'
i0
n
Unbalanced
load
DC Link
C dc
S4a
ila
R nl
Nonlinear load
SYMMETRICAL COMPONENT THEORY
Any set of ‘n’ unbalanced polyphase quantities
could be expressed as the sum of ‘n’
symmetrical sets of balanced phasors.
 Three Phase: Positive Sequence, Negative
Sequence and Zero Sequence

ia 0  1 1
 i   1 a
 a1  
ia 2  1 a 2
1  ia 
2  
a  ib 
a  ic 
SYMMETRICAL COMPONENT THEORY

Source Currents are balanced.
isa  isb  isc  0

Only the average load power is supplied by the source.
vsa isa  vsb isb  vsc isc  Plavg

Relation between Source Currents and Source Voltages.
(vsb  vsc  3 vsa )isa  (vsc  vsa  3 vsb )isb  (vsa  vsb  3 vsc )isc  0

All equations in matrix form.
1
isa  
i   v  v  3 v
sc
sa
 sb   sb
isc  
vsa
1
vsc  vsa  3 vsb
vsb
1
1
  0 


vsa  vsb  3 vsc   0 
  plavg 
vsc
SYMMETRICAL COMPONENT THEORY

Reference Compensator Currents.
i*
fa  ila  isa  ila 
i*
fb  ilb  isb  ilb 
i*
fc  ilc  isc  ilc 

vsa  
 vsb
 vsc

 Plavg 
vsb  

 vsc  vsa

 Plavg 
vsc  

 vsa  vsb

 Plavg 

Reference Compensator Currents with loss.
i*fa  ila  isa  ila 
vsa  
 vsb
 vsc

 Plavg
 Ploss


 Plavg
 Ploss


 Plavg
 Ploss
i*fb  ilb  isb  ilb 
vsb

   vsc  vsa
i*fc  ilc  isc  ilc 
vsc

   vsa  vsb


j a ,b ,c
vsj2

  tan( ) 3

FUZZY LOGIC
Concept introduced in 1965 by Lotfi. A. Zadeh
 Crisp set and Fuzzy set. Ex. Set of tall people.

Diagram of a crisp set and a fuzzy set.
FUZZY LOGIC CONTROLLER
The four main components of a Fuzzy Controller.
1) The Fuzzification Interface
2) The Inference Mechanism
3) The Rule Base
4) The Defuzzification Interface
Input
Fuzzification
Inference
Mechanism
Rule Base
Defuzzification
Output
FUZZY CONTROLLER

Inputs:
ref
err(i) = v dc
- v dc (i)
derr(i) = err(i) - err(i-1)

Outputs:
K p  K pref  K p
K i  K iref  K i

Calculation of Ploss
ref
ref
Ploss  K p (vdc
 vdc )  K i  (vdc
 vdc ) dt
FUZZIFICATION

Inputs to the Fuzzy controller: Error and change in error
of the capacitor voltage.

NL
-15
N
M
-10
NS
Z
-5
PS
1
0
PM
5
10
PS
PM
5
10
PL
15
err(t) in Volts

NL
NM
NS
Z
PL
1
-15
-10
-5
0
15
derr(t)in Volts
INFERENCE MECHANISM

The two main functions of the inference mechanism are:
a) Based on the active membership functions in error and the change in error
inputs, the rules which apply for the current situation are determined.
b) Once the rules which are on are determined, the certainty of the control action
is ascertained from the membership values. This is known as premise
quantification. ( Minimum Operation used )
IF
"error" is  PL (positive large)
"change in error" is  PM (positive medium)
THEN
"Kp"is L(LargeKp)
"Ki"isSKi(SmallKi)
RULE BASE
Capacitor Voltage Waveform during a load change
Volts
NL
err < 0
NM
NS
Z
time(s)
PS
PM
PL
err > 0
derr>0
derr
err
derr<0
derr>0
derr<0
derr
err
NL
NM
NS
Z
PS
PM
PL
Z
NL
SKi
SKi
SKi
Z
Z
Z
Z
Z
S
NM
SKi
SKi
SKi
Z
Z
Z
Z
Z
Z
Z
NS
LKi
LKi
LKi
Z
Z
Z
Z
Z
Z
Z
M
Z
LKi
LKi
LKi
Z
LKi
LKi
LKi
Z
S
S
M
L
PS
Z
Z
Z
Z
LKi
LKi
LKi
Z
S
S
M
L
L
PM
Z
Z
Z
Z
SKi
SKi
SKi
S
S
M
L
L
L
PL
Z
Z
Z
Z
SKi
SKi
SKi
NL
NM
NS
Z
PS
PM
PL
NL
L
L
L
M
S
S
NM
L
L
M
S
S
NS
L
M
S
S
Z
M
Z
Z
PS
Z
Z
PM
S
PL
Z
Rule base for Kp
Rule base for Ki
FUZZY CONTROLLER

Inputs:
ref
err(i) = v dc
- v dc (i)
derr(i) = err(i) - err(i-1)

Outputs:
K p  K pref  K p
K i  K iref  K i

Calculation of Ploss
ref
ref
Ploss  K p (vdc
 vdc )  K i  (vdc
 vdc ) dt
DSTATCOM SCHEMATIC
vsa
Ls
Rs
isa
PCC
vsb
isb
ilb
vsc
isc
ilc
Rf
Lf
N
A phas e
H-Bridge
S1a
Za
Zb
Zc
ifa ifb ifc
S3a
S1b
S3b S1c
S3c
if
S2a
S4b
S2b S4c
i Cd c
vdc
L nl
S2c
n'
i0
n
Unbalanced
load
DC Link
C dc
S4a
ila
R nl
Nonlinear load
SIMULATION VALUES
System Parameters
Values
Supply voltage
220V (phase-rms), 50 Hz
Unbalanced load
Rla = 50 , Lla = 20 mH
Rlb = 35 , Llb = 40 mH
Rla = 70 , Lla = 20 mH
Non-linear load
Three-phase full wave rectifier drawing a dc current of 5 A
DC capacitor
2200 µF
Interface inductor
Lf = 20 mH, R f = 5 
Reference dc link voltage
500 V
Hysteresis band
0.6 A
Gains tuned using the Energy concept
Kp= 110, Ki= 55
SIMULATION RESULTS
Nonlinear Unbalanced Load Currents
15
phas e-a
phas e-b
phas e-c
Current in Amperes
10
5
0
-5
-10
-15
0
0.01
0.02
0.03
0.04
0.05
0.06
Time in seconds
0.07
0.08
0.09
0.1
SIMULATION RESULTS
Filter Current - Reference and Actual
Ref erence and Actual Filter Currents in Phase A
5
Referenc e c ompens ator c urrent
ac tual c urrent
4
3
Current in Amperes
2
1
0
-1
-2
-3
-4
0.1
0.102
0.104
0.106
0.108
0.11
0.112
Time in seconds
0.114
0.116
0.118
0.12
SIMULATION RESULTS
Source Currents – Balanced and Sinusoidal
Actual Source Current Wavef orms
25
phas e-a
phas e-b
phas e-c
20
15
Current in Amperes
10
5
0
-5
-10
-15
-20
-25
0.01
0.015
0.02
0.025
0.03
0.035
Time in seconds
0.04
0.045
0.05
SIMULATION RESULTS
Normally tuned PI and Fuzzy Supervised waveforms
DC link Voltages - Normally tuned and Fuzzy tuned
540
Fuzzy supervised PI
530
Normally tuned PI
DC Link Voltage in
Volts
520
510
500
490
480
470
460
0
0.1
0.2
0.3
Time in
seconds
0.4
0.5
0.6
SUMMARY – TAKE HOME POINTS
The DSTATCOM can be used to ensure
balanced and sinusoidal source currents even
if the load is unbalanced and non-linear.
 Fuzzy supervision of the DC link PI controller
can be used to reduce the error in DC Capacitor
voltage during load change.
