Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
Performance Improvement with a Robust Self Tuned
Fuzzy Logic Controller for Generator Control in
Wind Energy System
Swagat Pati , K.B.Mohanty, Benudhar Sahu Department of Electrical Engineering, National Institute of Technology, Rourkela, India
[email protected] [email protected] [email protected]
Abstract— In this paper a line excited cage generator is considered
which is connected with the grid through a bidirectional PWM converter- inverter system. The generator is controlled by indirect field oriented control (IFOC) scheme. Fuzzy logic controllers (FLC) are used for the control purpose. The first FLC is used in the outer speed loop to track the generator speed with the reference speed for maximum power extraction and the second and third FLCs are used in the inner current loops for control of active and reactive power. The FLCs use normalized values of error and change of error as their inputs. The outputs of the FLCs are again multiplied with gains to give the control signals. A trapezoidal membership function is taken for the error input and triangular membership functions are taken for change of error as well as output. Again a robust self tuned fuzzy logic controller (STFLC) scheme is used in place of the FLCs. In this scheme a tuning FLC (TFLC) is used to tune the output gain of the main FLC. The inputs to both the FLCs are normalized values of error and change of error. The output of the TFLC is the output gain of the main FLC. The main FLC is similar to the FLC as discussed in the previous scheme. In the TFLC triangular membership functions are used for all input as well as output variables. The performances of both the schemes are simulated and a comparison is given. The simulation work is done in MATLAB coding environment. Keywords- Induction Generator; Back to Back PWM converters; Indirect Vector control; FLC; Self Tuned FLC
I. INTRODUCTION
Wind energy is one of the most important and promising source of renewable energy all over the world, mainly because it is considered to be nonpolluting and economically viable. At the same time there has been a rapid development of related wind energy technology. However in the last two decades, wind power has been seriously considered to supplement the power generation by fossil fuel and nuclear methods. Normally induction machines are used for generation purpose in wind energy systems. But due to the coupling effect between active and reactive power the response becomes sluggish and the control becomes difficult and complex in case of induction generators. When induction machines are operated using vector control techniques, fast dynamic response and accurate torque control are obtained. All of these characteristics are advantageous in variable speed wind energy conversion systems (WECS). Squirrel cage generators with shunt passive or active VAR (volt ampere
reactive) generators was proposed in [3], which generate constant frequency power through a diode rectifier and line commutated thyristor inverter. A comparative study of fixed speed and DFIG during power system disturbances such as network voltage sags and three phase faults, as well as the possibility of network voltage instability is investigated in [4]. The performance of a DFIG driven by a wind turbine connected to large power systems is studied in [5]. Operation of several self excited induction generators connected to a common bus is analyzed in [6]. The control systems for the operation of indirect rotor flux oriented vector controlled induction machines for variable speed wind energy applications are discussed in [7]-[9]. Sensorless vector control scheme suitable to operate cage induction generator is discussed in [7]. In [8] cage induction machine is considered and a fuzzy control system is used to drive the WECS to the point of maximum energy capture for a given wind velocity. In [11] a stator flux oriented vector control scheme is discussed to decouple the active and reactive power generated from the variable speed wind generation system employing DFIG. A new control scheme using non-linear controller and FLC is employed for a variable speed grid connected wind energy system in [12] where as a vector control scheme for both supply side and machine converters is discussed and the independent control of active and reactive power is done in [13]. A FLC based speed control scheme is described in [14] employing feed-forward vector control scheme for performance improvement of an induction machine. Again a self tuned fuzzy controller is implemented for the performance improvement of a field oriented control of an induction motor in [15]. Several other fuzzy logic MRAS scheme have been developed in [16] - [20]. The induction machine is connected to the utility using back to-back converters. In this paper a variable speed wind turbine driven squirrel cage induction generator system with two double sided PWM converters is described. Fuzzy controllers are used to optimize efficiency and enhance performance. Again a self tuned fuzzy logic control scheme is implemented for performance enhancement purpose and a comparison is given. The control algorithms are evaluated by MATLAB simulation study.
2010 International Conference on Industrial Electronics, Control and Robotics
978-1-4244-8546-8/10/$26.00 ©2010 IEEE 185
II. INDUCTION GENERATOR MODEL
The basic configuration of a line excited i
generator is sketched in Fig.1. Normally the
interfaced with the grid through back-to-back PWM c
configuration.
.
Fig: 1 Basic block diagram of the wind energy syste
The sign convention applied in this model i
to that of the motor, i.e. the stator currents are posit
flowing towards the machine and real power and
power are positive when fed from the grid.
Stator voltages:
Vqs = Rs iqs +dt
dqsΨ
+ ωeΨds
Vds = Rsids +dt
ddsΨ
- ωeΨqs
Rotor voltages:
Vqr = Rr iqr +dt
dqrΨ
+ (ωe-ωr) Ψdr
Vdr = Rr idr + dt
ddrΨ
- (ωe-ωr) Ψqr
The flux linkages in these equations were calculated
Ψqs = Ls iqs + Lm iqr , Ψqr = Lr iqr + Lm iq
Ψds = Ls ids + Lm ids , Ψdr = Lr idr + Lm id
The final mathematical model for the squirrel cage
machine is given as
⎥⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢⎢
⎣
⎡
qr
dr
qs
ds
V
V
V
V
=( ) ( )
( ) ( )⎢⎢⎢⎢
⎣
⎡
+−−
−−+−−
+
−−+
rrrremmre
rerrmrem
mmessse
memsess
sLRLsLL
LsLRLsL
sLLsLRL
LsLLsLR
ωωωω
ωωωω
ωω
ωω
where ‘s’ is the laplace operator and ‘ωr’ is the rotor
speed.
induction
stator is
converter
em
is similar
ive when
d reactive
(1)
(2)
(3)
(4)
d from:
qs
ds
induction
⎥⎥⎥⎥
⎦
⎤
rL
⎥⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢⎢
⎣
⎡
qr
dr
qs
ds
i
i
i
i
(5)
electrical
For a singly fed machine as the cage motor
The machine dynamics is given as
m
m
turbineeB
dt
dJTT ω
ω
+=+
)(22
3qrdsdrqsme iiiiL
pT −=
The active and reactive power of the generat
qsqsdsds iViVp +=
qsdsdsqs iViVq −=
III. CONTROL STRUCTUR
A. Induction generator control scheme
The control scheme is shown in Fig.2. Indi
control scheme is implemented for high pe
of the cage generator. The key feature of fiel
is to keep the magnetizing current at a consta
T
Fig: 2 Control Structure
Thus the torque producing component o
adjusted according to the active power dem
dynamic response . With this assumption
formulation can be written as
ds
qs
r
rsl
i
i
L
R=ω
qsdr
r
me i
L
LPT ψ
22
3=
where ωsl is the slip speed and ψdr is the
linkage.
Vqr =Vdr = 0
(6)
(7)
or is given as
(8)
(9)
RE
irect field oriented
erformance control
ld oriented control
ant rated value.
of current can be
mand with a better
the mathematical
(10)
(11)
e d-axis rotor flux
186
The basic configuration of the system cons
line excited squirrel cage induction generator interfa
the grid through back-to-back PWM converters.
The converters are voltage controlled bidi
voltage source inverters having a dc link between th
rotor speed is fed back by a speed sensor which is th
with the slip speed to get the synchronous speed
further integrated to get the angular displacement
which the unit vectors are generated.
The controller FLC-1 makes the rotor speed
the reference speed such that maximum power can be
from the wind by the wind turbine. The controlle
controls the direct axis rotor current such that t
magnetizing flux remains constant to get better
response.
The controller FLC-3 controls the quadra
stator current such that the given active power deman
met.
B. Fuzzy Logic Controller Design
The structure of the fuzzy logic controllers is
Fig. 3. FLC1 uses normalized values of speed e
change of speed error as its inputs. Ke and Kce are
scaling factors which are chosen to normalize the e
change of error respectively. The output of FLC1
multiplied with output scaling factor Ko to give contr
iqs*.
FLC2 uses normalized values of direct ax
current error and change of direct axis current err
inputs and it gives control output uds*. FLC3 takes no
values of quadrature axis current error and ch
quadrature axis current error as its inputs and gives
and uqs* when added with (-ωeψqs) and (ωeψds) res
give control voltages vds* and vqs
*.
Fig: 3 Block diagram of fuzzy controller
Table-1
Input error membership functions are s
Fig.4(a). The input error fuzzy sets use both triang
trapezoidal membership fuctions, which are foun
optimum. Triangular membership functions as s
Fig.4(b) are used for change of error input. Th
sists of a
aced with
irectional
hem. The
hen added
which is
θe from
d to track
captured
er FLC-2
the rotor
dynamic
ature axis
nd can be
s given in
error and
the input
error and
is again
rol output
xis stator
ror as its
ormalized
hange in
uqs*. uds
*
spectively
hown in
gular and
nd to be
hown in
he output
membership functions are shown in Fig.4(c)
the FLC algorithm is given in Table-1
Mamdani type fuzzy inferencing is used
constants, membership functions, fuzzy se
output variables, and the rules used in the
by trial and error to obtain optimum drive pe
study, the centroid method of defuzzification
(a)
(b)
(c)
Fig: 4 Membership functions for
(a) input membership functions for error (b) infunctions for change in error (c) output memb
C. STFLC design
The robust self tuned fuzzy logic controller
is used in place of the FLCs. In this sche
(TFLC) is used to tune the output gain of the
Fig: 5 STFLC gain tuning sch
. The rule-base for
. For this study
d. The values of
ets for input and
study are selected
erformance. In this
n is used.
FLC
nput membership bership functions
r (STFLC) scheme
eme a tuning FLC
e main FLC.
heme
2010 International Conference on Industrial Electronics, Control and Robotics
187
The inputs to both the main FLC and the Tnormalized values of speed error and change of speThe output of the TFLC is the gain updating factor α wa value between 0 and 1.When α is multiplied with the output gain of the STFLC. The main FLC is simiFLC as discussed in the previous subsection.. The gamechanism of the STFLC is shown in Fig 5. In thtriangular membership functions are used for all inpuas output fuzzy sets. With a view to improving thcontrol performance, we use the rule base in Tacomputation of α.
Further modification of the rule base for ‘αrequired, depending on the type of response the controdesigner wishes to achieve. It is very important to notrule base for computation of ‘α’ will always be depethe choice of the rule base for the controller. Any sichange in the controller rule base may call for changrule base for ‘α’ accordingly.
The input and output membership functions for the Tgiven in Fig:6. For the designing of the TFLC .Mamfuzzy inferencing is used. The values of membership ffuzzy sets for input and output variables, and the rulethe study are selected by trial and error to obtain optimperformance. In this study the centroid medefuzzification is used.
Table-2
(a)
(b)
Fig: 6 Membership functions for TFLC(a) input memfunction, (b) output membership function
TFLC are eed error. which has Ko gives
ilar to the ain tuning he TFLC ut as well he overall ble-2 for
’ may be ol system te that the endent on ignificant ges in the
TFLC are dani type functions, es used in
mum drive ethod of
mbership
IV. SIMULATION RESULT
The drive system was first simulated w
controllers with different operating conditi
change in the reference speed and step c
torque and some sample results are presente
section.
A step change in command speed f
1690 rpm is given at t = 2.5sec. which con
and again returns to the previous value.
controller the machine takes around 0.11
steady state but with self tuned fuzzy logic
approximately 0.065 sec. to achieve steady s
seen clearly in Fig.7
.(a)
(b)
Fig: 7 Speed response with (a) FLC (b) STFLC
speed command
In Fig. 8 it can seen that due to a ch
command there is no variation in active popower, but the reactive power response wsluggish than with STFLC. Again a step torque from 10 Nm to 15 Nm is given aresponses are shown . The turbine torque is kNm till t = 2 sec. At t = 2 sec. the turbine tor15 Nm and at t = 3 sec. again the turbine to10 Nm.
(a)
TS
with Fuzzy logic
ions such as step
change in turbine
ed in the following
from 1890 rpm to
ntinues for 0.5sec.
With fuzzy logic
15sec. to achieve
controller it takes
state which can be
for step change in
hange in the speed ower and reactive with FLC is more change in turbine
and the simulation kept constant at 10 rque is increased to orque is restored at
188
(b)
Fig: 8 Active power and reactive power response with (b) STFLC
(a)
(b)
Fig: 9 Speed response for a change in turbine torque wFLC (b) STFLC
(a)
(b)
Fig: 10 active and reactive power responses for a step turbine torque with (a) FLC (b) STFLC
(a) FLC
with (a)
change in
The speed response of the machine due
turbine torque is given in Fig: 9. In fig: 9 i
the controller is made on at t = 1 sec. Before
starts up in open loop. From Fig: 10 we can see that the active
increases with the increase in turbine torquevariation in reactive power with STFLC than
(a)
(b)
Fig: 11 Power factor response with (FL
For which there is an improvement in theSTFLC scheme which is justified form Fischeme the power factor varies from 0.62 change in turbine torque, Where as in thepower factor varies from 0.58 to 0.72.
(a)
(b)
Fig: 12 Line current response with (a)FLC (b
e to the change in
it can be seen that
e that the machine
e power generation e, but there is less in FLC scheme.
LC) (b) STFLC
e power factor with ig: 11. In STFLC to 0.745 with the
e FLC scheme the
) STFLC
2010 International Conference on Industrial Electronics, Control and Robotics
189
V. CONCLUSION
The self excited induction generator system was simulated
with both FLC and STFLC controllers in MATLAB . Both the
controllers improve the responses of the generator system.
While the generator system with FLC achieves steady state in
0.02 sec, with STFLC the system achieves steady state in
0.013 sec, which makes the STFLC faster controller than the
FLC. During the sudden change in turbine torque the STFLC
gives a better performance in reactive power control than the
FLC due to which the power factor in case of STFLC control
becomes better than that with FLC control. This concludes
that the STFLC has an overall better performance than the
FLC.
APPENDIX
Symbols
Vas, Vbs, Vcs = Three phase supply voltages
Vdss, Vqs
s = d- q axis voltages in stationary reference frame
Vds, Vqs = d-q axis voltages in synchronously rotating reference
frame
Rs, Rr = stator and rotor resistances
Ls, Lr = stator and rotor inductances
Lm = magnetizing inductance
Ψds, ψqs, ψdr, ψqr = stator and rotor flux linkages
ωe = synchronous speed(electrical)
ωr = rotor electrical speed
ωm* = reference rotor speed (mechanical)
Te = electromagnetic torque
Tturbine = prime mover torque
J = moment of inertia
B = frictional damping coefficient
ωsl = slip speed
ωm = rotor mechanical speed
P = pair of poles of the machine
p = active power
q = reactive power
Machine parameters
3 hp, 3Ф, 220V, 60Hz, P=4, Rs = 0.435Ω, Rr = 0.816Ω, Xls = 0.754Ω,
Xls=Xlr, Xm = 26.13Ω, J = 0.089 kg-m2, B = 0
REFERENCES
[1] G. Saccomando, J. Svensson, and A. Sannino, “Improving voltage
disturbance rejection for variable speed wind turbines,” IEEE Trans.
Egy.Conv., vol. 17, no. 3, pp. 422-428, Sep 2002.
[2] J. Morren, and S. W. H. de Hann, “Ridethrough of wind turbines with
doubly-fed induction generator during a voltage dip,” IEEE Trans.
Egy.Conv., vol. 20, no. 2, pp. 435-441, Jun 2005.
[3] C. V. Nayar, and J. H. Bundell, “Output power controller for a wind
driven induction generator,” IEEE Trans. Aerospace Electronic Systems,vol.
23, pp. 388-401, May 1987.
[4] L.Holdsworth, X.G.Wu, J.B.Ekanayake, N.Jenkins, “Comparison of fixed
speed and doubly-fed induction wind turbines during power system
disturbances,” IEE Proc. Gener. Transm. Distrib. , vol.150, No. 3, pp 343-
352, May2003.
[5] Richard Gagnon, Gilbert Sybille, Serge Bernard, Daniel Paré, Silvano
Casoria, Christian Larose, “Modeling and Real-Time Simulation of a
Doubly-Fed Induction Generator Driven by a Wind Turbine,” International
Conference on Power Systems Transients (IPST’05) , Montreal, Canada,
Paper No. IPST05-162, June 2005.
[6] C. Chakraborty, S. N. Bhadra, and A. K. Chattopadhyay, “Analysis of
parallel operated self excited induction generators,” IEEE Trans. Egy.Conv.,
vol. 14, no. 2, pp. 209-216, Jun 1999.
[7] R. Cardenas, and R. Pena, “Sensorless vector control of induction
machines for variable speed wind energy applications,” IEEE Trans.
Egy.Conv., vol. 19, no. 1, pp. 196-205, Mar. 2004.
[8] M. G. Simoes, B. K. Bose, R. J. Spiegel, “Design and performance
evaluation of a fuzzy logic based variable speed wind generation system,”
IEEE Trans. Ind. Appl., vol. 33, no. 4, pp. 956-965, Aug. 1997.
[9] S. N. Bhadra, D. Kastha, and S .Banerjee, Wind Electrical Systems,
Oxford University Press, New Delhi, 2005.
[10] M.G.Simoes, B.K.Bose, and R.J.Spiegel, “Fuzzy logic based intelligent
control of a variable speed cage machine wind generation system,”IEEE
Trans. Power Electron., vol. 12, pp. 87-95, Jan 1997.
[11] A. Tapia, G.Tapia, J.Xabier Ostolaza, J.R.Saenz , “Modellimg and
control of a wind turbine driven doubly fed induction generator” IEEE Trans.
Energy conversion, vol.18, No. 2, pp. 194-204, June 2003.
[12] R.Pena, R.Cardenas, R.Blasco, G.Asher, J.Clare, “A cage induction
generator using back to back PWM converters for variable speed grid
connected wind energy system,” IEEE 27th annual Conf., IECON 01,
Industrial Electronics Society, pp. 1376-1381, 2001.
[13] R.Pena, J.C.Clare, G.M.Asher, “Doubly fed induction generator using
back to back PWM converters and its application to variable speed wind
energy generation,” IEE Proc. Electr. Power Appl. , vol. 143, No. 3, pp. 231-
241, May1996.
[14] M.NasirUddin, Tawfik S. Radwan, M.Azizur Rahman, “Performance
of Fuzzy-Logic-Based indirect vector control for induction motor drive,”
IEEE Trans. Industry applications, vol. 38, No. 5 , pp. 1219-1225, Oct 2002.
[15] M. Masiala , J.Salmon , “Fuzzy self tuning speed control of an indirect
field-oriented control induction motor,” IEEE Trans. industry appl. ,vol .
44 , No. 6 ,Dec 2008.
[16] A. El Dessouky and M. Tarbouchi, “Fuzzy model reference self-tuning
controller,” in Proc. 7th Int.Workshop Advanced Motion Control, Jul.35,2002,
pp. 153–158. [17] M. Cheng, Q. Sun, and E. Zhou, “New self-tuning fuzzy PI control of
anovel doubly salient permanent-magnet motor drive,” IEEE Trans. Ind. Electron., vol. 53, no. 3, pp. 814–821, Jun. 2006.
[18] Y. Miloud, A. Miloudi, M. Mostefai, and A. Draou, “Self-tuning fuzzy
logic speed controller for inductionmotor drives,” in Proc. IEEE Int. Conf.
Ind. Technol., Dec. 8–10, 2004, pp. 454–459. [19] L. Mokrani and R. Abdessemed, “A fuzzy self-tuning PI controller for
speed control of induction motor drive,” in Proc. IEEE Conf. Control Appl., Jun. 2003, vol. 2, pp. 785–790.
[20] J. Sun, P. Su, Y. Li, and L. Li, “Application of self-adjusting fuzzy
controller in a vector-controlled induction motor drive,” in Proc. 3rd IEEEInt.
Conf. Power Electron. Motion Control, Aug. 15–18, 2000, vol. 3.
190