Embed Size (px)
Citation preview
Dynamic Wind Power Control for System Integration
Review of PAC Controller
The Generator-Response Following concept
Feedforward Controller for stable use of the Generator-Response Following concept
Part 2: Experimental DFIG prototype
Questions
David Campos-Gaona, Olimpo Anaya-Lara, William LeitheadUniversity of Strathclyde, Glasgow, UK.
SUPERGEN Wind General Assembly November 2017
Power adjusting controller (PAC) applied to each rotor
PAC adjusts the output from rotor i byDPi
Acts on information regarding state of each turbine to adjust operating point
WPP controller determines the adjustments DPi as required to meet objectives.
PAC and the wind power plant controller (WPPC)
Source: Adam Stock, Supergen Wind
The WPP controller enables a wind power plant to provide the full range of ancillary services including synthetic inertia at the wind far level rather than a single turbine level.
• Connect at the Point of Connection of the wind farm a fully instrumented
small/medium synchronous generator to which the wind farm is slaved (using
PAC) in order to provide a wide range of ancillary services by scaling up the
response of small generator.
3
Novel approach to provide ancillary services
to the power grid from Wind Farms using
PAC: The Generator-Response Following
(GRF)
Advantages
– No direct power frequency measurements, which would be prone to high noise and lack of
accuracy
– Capable of providing a full range of ancillary services, inertia, governor-droop control, reactive
power, reserve, curtailment etc.
– Grid Code Compliant
4
Test network for GRF concept
5
Results of GRF considering no
communication delay
6
Results of GRF Considering communication
delay of 150ms
• Requirements:
– The HVDC converter station must
use the energy stored in the
capacitor to provide immediate
GRF.
– Once the energy from the offshore
wind farm arrives to the onshore
HVDC station, it must replenish the
lost energy in the capacitor without
affecting the GRF profile.
– HVDC voltage must not decrease
beyond a safety level.7
Feedforward Controller for GRF
8
Feedforward Controller for GRF
The accuracy of the feedforward action is based in the accuracy of the plant parameters used to tune the controller.
9
DC voltage plant disturbance rejection
21
2dcW Cv
Variable to control
Capacitor Energy2
1 322
dc
in d d
dvdWC P v i
dt dt
PlantTransfer Function
2 ( ) 3( )
( )
dc d
dc
d
v s vP s
i s Cs
The transfer function has a pole in the origin, making it very susceptible to disturbances
Plant
( )dcP s
( )di s
Output2
dcvΣ
+
Control
signal
( )inP s
ControllerΣ
2
_dc refv
reference
Disturbance
++
-
- Robust control of the dc voltage
- Internal Model Controller (IMC) with 2 degrees of freedom
Σ1( ) ' ( )L s M s Plant
( )P sΣ
-
++
( )r s
( )d s
Output
( )y sΣ
'( )M s
IMC controller G(s)
Plant model
+
+
+
( )F s
aG
Σ
Internal feedback loop
+
-
( )M s
Reference
Naturally damped system
Poorly damped system
disturbance disturbance
Active damping3
dc
d
d
dc
dcdc c
C
v
Ki G
Kp
1
3c
d
dc
d
C
vG
IMC constant for robust control
of dc voltage
10
Enhanced feedforward controller simulation
results for a 150ms delay
Feedforward controller enables using GFR with delay in communications
The stability of the grid is not compromised
11
Ideal results vs Feedforward controller results
Feedforward controller results follow very closely the ideal results
The extraction of energy from the DC capacitor produces variation on the DC voltage, but it is adjusted by the enhanced feedforward controller to match the set-point value
12
Second part of research: DFIG prototype
development
av
bv
cv
ravrbv
rcv
savsbv
scv
Neutral
Stator Rotor
in
in
in
out
out
out
in
in
in
in
out
out
out
a
b
c
nin n
a
b
c
ab
c
40V DC
(Sorennsen DCS-
40)
Labvolt 500W Wound Rotor
Induction Machine
12
3
45
6
7 8
9
Hawk 3P
Analog Instrumentation board
Hawk 3P
Analog Instrumentation board
DC Motor/Dynamometer
Mechanical
connexion CPP
(DSP F28335
Control Board)
Stator electric signal
measurement
(Vas,Ias,Vbs,Ibs,Vcs,Ics)
Ro
tor
sig
nal
s
Ia,I
b,I
c
ADC A0-A5
ADC B0-B2
ADC B3
Rotor speed
eP
WM
1,2
,3
DA
CS
IMO
DA
CSO
MI
A, B (rotor position) Voltage dividerstage Rotor speed
A
B
2kW
3 Phase VSC
3 Phase
AC Source
120 V 5.5kW
(California
Instruments
CSW5550)
PW
M
pu
lses
DFIG experimental
prototype
circuit
13
DFIG Experimental prototype
AC voltage source
DC voltage source
InductionMachine
DC Prime mover
VSC DSP
Measurements
VSCMeasurement DSPa) b)
c)
DC Prime mover
InductionMachine
14
Digital Controller development
A program that • reads 10 analog inputs using analog-to-
digital converters, • reads 2 digital inputs of a rotor position
encoder, • executes 4 closed-loop D2F-IMC
controllers, • executes a PLL, • executes 2 abc-to-dq0 transformation• generates switching pulses for the 6 IGBTs
of the VSC was designed in floating point architecture with a sampling frequency of 12kHz (i.e. 83.3 microsecond step program execution) using automatic code generation
15
Experimental results: Rotor Speed Controller
Reactive power consumption at machine terminals is 0
Rotor speed control is attained
16
Experimental results: Speed Control
Rotor speed is kept constant for changes in mechanical torque inputs
17
Experimental results: Reactive power control
Reactive power can be controlled without affecting the machine speed.
Experimental Results: Advanced control
Harmonic distortion in Voltage Signal produce highly distorted Rotor and Stator Currents
19
The Stator and Rotor Current Harmonic distortion is eliminated using our proposed advanced controller technique
Experimental Results: Advanced control
20
Experimental Results: Advanced control
Same result presented with shorter time division.
