Fuzzy logic control of DFIG-based wind turbine

Sakina Elkhadiria
Pr.Lamiaa Elmenzhib Pr.Abdelouahid Lyhyaouic
Laboratory of Innovative Technologies
Laboratory of Innovative Technologies Laboratory of Innovative Technologies
University of ABDELMALEK ESSAADI
University of ABDELMALEK ESSAADI University of ABDELMALEK ESSAADI
(ENSA) Tangier, Morocco
[email protected] (ENSA) Tangier, Morocco (ENSA) Tangier, Morocco

Abstract --This paper proposes the control of the rotor side

PWM converter of a variable speed doubly fed induction II. WIND TURBINE MODELING
generator (DFIG) based wind turbine using rotor flux
oriented vector control based fuzzy logic in order to II.1. Turbine model
control the rotor currents. The fuzzy logic controllers and
PI controller are used to control the rotor currents to Wind turbine uses the power extracted from the wind to
overcome any disturbance. The simulations are developed produce electric power, then drive it to electrical generator.
in MATLAB Simulink. The power contained in the wind is given as:
ܲ௔௜௥ ൌ ͳȀʹ ‫ ܴߩ כ‬ଶ  ଷ (1)
Keywords-- Doubly fed induction generator (DFIG), Power control, The mechanical power and the torque on the shaft are given
fuzzy logic, PI controller, variable speed wind turbine. by the expressions:

I. INTRODUCTION ܲ௠௧ ൌ ͳȀʹ ‫ ܴߨߩ כ‬ଶ  ଷ ‫ܥ‬௣ ሺߣǡ ߚሻ (2)

Wind energy is the fastest growing capacity among all ܶ௧ ൌ ͳȀʹ ‫ ܴߨߩ כ‬ଶ  ଷ ‫ܥ‬௣ ሺߣǡ ߚሻ ‫ͳ כ‬Ȁπ‫ݐ‬ (3)
renewable energy sources. More than 54 GW of clean
Where, ܴǡ ߩǡ ǡ ‫ܥ‬௣ǡ ߣ andߚ are respectively: radius of the
renewable wind power was installed across the global market
turbine (m), air density (kg/ଷ ), wind speed (m/s), the power
in 2016, cumulative capacity grew by 12.6 % to reach a total
coefficient, the tip speed ratio and blade pitch angle (deg). To
of 486.8 GW, or cumulative capacity reached a total of 486.8
drive the DFIG, fixed pitch is chosen as ߚ ൌ ͲǤ
GW in 2016 [1]. The contribution of wind power in the world
The power coefficient ‫ܥ‬௣ of the wind turbine is given as
total generation capacity is expected to add 60 GW of new follow:
wind installations in 2017, rising of about 75 GW by 2021, to ଵଵ଺
షమభ

bring cumulative installed capacity of over 800 GW by the ‫ܥ‬௣ ሺߣǡ ߚሻ ൌ ͲǤͷͳ ቀ െ ͲǤͶߚ െ ͷቁ ݁ ഊ೔
൅ ͲǤͲͲ͸ͺߣ (4)
ఒ೔
end of 2021 [1]. ோπ೟ೠೝ
With: ߣൌ (5)
௏
Over the last few years, Doubly Fed Induction Generator Where π௧ is the wind turbine speed.
has attracted great attention and dominate the world market ͳ ͳ ͲǤͲ͵ͷ
ൌ െ
due to its advantages over fixed-speed-based WECS by less ߣ௜ ߣ ൅ ͲǤͲͺߚ ߚଷ ൅ ͳ
mechanical stress, less acoustical noise, maximum power Fig.1 shows the Cp curve for ߚ =0.
attraction and reactive power compensation using back to
back power converters of rating near to 25%-30% to the
generator capacity [1]. There are many structures and controls
which could be used to control the converters, and the most
common technique is controlling the rotor currents. This paper
presents the design of a control scheme using Fuzzy logic
controller. The fuzzy control has the ability to be robust and
simple to design, it doesn’t require the knowledge of the exact
model [2]. The first part explains the modelling of wind
turbine and the DFIG, and the second part presents the
performance of this type of controller. It is clear from results
that the fuzzy logic control has better stability of the DFIG. Fig.1 Turbine power coefficient.

978-1-5386-4396-9/18/$31.00 ©2018 IEEE

II.2. DFIG model 
Š‡ƒ –‹˜‡ƒ†”‡ƒ –‹˜‡•–ƒ–‘”’‘™‡”‡“—ƒ–‹‘•‹†“Ǧƒš‹•
The doubly fed induction generator is a system in which ƒ”‡ǣ
the electronic power controls the rotor currents to achieve the ெ
‫ܥ‬௘ ൌ െ‫ ݌‬௅ ߮ௗ௦ ‫ܫ‬௤௥
maximum energy with variable wind speed. ೞ
The stator is directly connected to the grid, and the rotor is ܸ݀‫ ݏ‬ൌ Ͳ
ቊ
connected via slip rings to a back-to-back converter. ܸ‫ ݏݍ‬ൌ ܸ‫ ݏ‬ൌ ߱‫ ݏ߮ ݏ‬

‡‘„–ƒ‹–Š‡‡™‡š’”‡••‹‘‘ˆ–Š‡•–ƒ–‘” —””‡–•ǣ

‫ܫۓ‬ ‫ܯ‬ ߮
ۖ ݀‫ݏ‬ ൌെ ‫ ݎ݀ܫ‬൅ ‫ݏ‬
‫ݏܮ‬ ‫ݏܮ‬

‫۔‬ ‫ܯ‬
ۖ ‫ ݏݍܫ‬ൌ െ ‫ݎݍܫ‬
‫ە‬ ‫ݏܮ‬

‫ܯ‬ଶ ݀‫ܫ‬ௗ௥ ‫ܯ‬ଶ
‫ۓ‬ ܸௗ௥ ൌ  ܴ௥ ‫ܫ‬ௗ௥ ൅ ቆ‫ܮ‬௥ െ
ቇ െ ߱௥ ቆ‫ܮ‬௥ െ ቇ‫ܫ‬
ۖ ‫ܮ‬௦ ݀‫ݐ‬ ‫ܮ‬௦ ௤௥

‫۔‬ ‫ܯ‬ଶ ݀‫ܫ‬௤௥ ‫ܯ‬ଶ 
ܸۖ௤௥ ൌ  ܴ௥ ‫ܫ‬௤௥ ൅ ቆ‫ܮ‬௥ െ ቇ െ ߱௥ ቆ‫ܮ‬௥ െ ቇ ‫ ܫ‬൅ ߱௥ ߮௦
‫ە‬ ‫ܮ‬௦ ݀‫ݐ‬ ‫ܮ‬௦ ௗ௥ ‫ܮ‬௦
Fig.2 Schematic diagram of DFIG-based wind generation
systems 
Š‡•–ƒ–‘”ƒ –‹˜‡ƒ†”‡ƒ –‹˜‡’‘™‡”ƒ”‡‰‹˜‡ƒ•ˆ‘ŽŽ‘™ǣ

The model of DFIG is established in the dq reference. The ܲ௦ ൌ ܸ௦ ‫ܫ‬௤௦ 
൜ (15)
stator and rotor voltages are given by the following ܳ௦ ൌ ܸ௦ ‫ܫ‬ௗ௦ 
expressions: 
ௗ
ܸௗ௦ ൌ  ܴ௦ ‫ܫ‬ௗ௦ ൅ ௗ௧ ߮ௗ௦ െ ߱௦ ߮௤௦ ܲ‫ ݏ‬ൌ  െܸ‫ ݎݍܫ ܯܮ ݏ‬
‫ݏ‬
ቐ ௗ (6), (7) ൞ ʹ (16)
ܸ௤௦ ൌ  ܴ௦ ‫ܫ‬௤௦ ൅ ௗ௧ ߮௤௦ ൅ ߱௦ ߮ௗ௦ ܳ‫ ݏ‬ൌ െܸ‫ ݎ݀ܫ ܯܮ ݏ‬൅ ‫߱ݏܸܮ‬
‫ݏ‬ ‫ݏ ݏ‬
ௗ
ܸௗ௥ ൌ  ܴ௥ ‫ܫ‬ௗ௥ ൅ ௗ௧ ߮ௗ௥ െ ߱௥ ߮௤௥ 
ቐ ௗ (8), (9)
ܸ௤௥ ൌ  ܴ௥ ‫ܫ‬௤௥ ൅ ௗ௧ ߮௤௥ ൅ ߱௥ ߮ௗ௥ 
III.VECTOR CONTROL OF DFIG

Where ܴǡ ‫ܮ‬ǡ ‫ܮ‬௠ and ߮and݅ represent respectively resistance ‡ Š‘‘•‡†“”‡ˆ‡”‡ ‡–‘”‡’”‡•‡––Š‡ ǡ†Ǧƒš‹•‹•
of windings, inductance, mutual inductance, flux and current. ሬሬሬሬԦǡ™‡Šƒ˜‡ǣ
ƒŽ‹‰‡†™‹–Š–Š‡•–ƒ–‘”ˆŽ—š߮ ௦
The subscripts s, r, d and q respectively indicate stator, rotor, 
d-axis and q-axis.
The flux equations are given as:

߮ௗ௦ ൌ  ‫ܮ‬௦ ‫ܫ‬ௗ௦ ൅ ‫ܫܯ‬ௗ௥ 

൜ ߮ ൌ  ‫ ܫ ܮ‬൅ ‫ܫܯ‬ (10), (11)
௤௦ ௦ ௤௦ ௤௥
߮ௗ௥ ൌ  ‫ܮ‬௥ ‫ܫ‬ௗ௥ ൅ ‫ܫܯ‬ௗ௦
൜ ߮ ൌ  ‫ ܫ ܮ‬൅ ‫ܫܯ‬ (12), (13)
௤௥ ௥ ௤௥ ௤௦

Š‡”‡‫ܫ‬ௗ௦ ǡ‫ܫ‬௤௦ ƒ†‫ܫ‬ௗ௥ ǡ ‫ܫ‬௤௥ ƒ”‡–Š‡†‹”‡ –ƒ†“—ƒ†”ƒ–‡

•–ƒ–‘”ƒ†”‘–‘” —””‡–•Ǥ
Š‡ ‡Ž‡ –”‘ƒ‰‡–‹ –‘”“—‡ ƒ„‡‡š’”‡••‡†ƒ•ǣ
ଷ 
‫ܥ‬௘௠ ൌ ଶ ሺ߮ௗ௦ ‫ܫ‬௤௦ െ ߮௤௦ ‫ܫ‬ௗ௦ ሻ (14) Fig 3: Orientation of the stator flux
 
߮݀‫ ݏ‬ൌ ߮‫ݏ‬ ߮ௗ௦ ൌ ߮௦ 
ቊ߮  ൜߮ ൌ  ‫ ܫ ܮ‬൅ ‫ ܫܯ‬ൌ Ͳ
‫ݏݍ‬ൌͲ 
(17)
௤௦ ௦ ௤௦ ௤௥
 • Fuzzy rules: It consists of the development of suitable rules
ௗ set.
ܸௗ௦ ൌ  ܴ௦ ‫ܫ‬ௗ௦ ൅ ௗ௧ ߮ௗ௦
Š‡ቊ (18)
ܸ௤௦ ൌ  ܴ௦ ‫ܫ‬௤௦ ൅ ߱௦ ߮ௗ௦
• Defuzzification: It consists of the conversion of a fuzzy
Š‡˜‘Ž–ƒ‰‡†”‘’•ƒ”‡‡‰Ž‡ –‡†ǣܴ௦ ‫ܫ‬ௗ௦ ൌ Ͳƒ†ܴ௦ ‫ܫ‬௤௦ ൌ Ͳ quantity to a precise quantity.
ௗ
ܸ  ؄ ௗ௧ ߮௦
so ቊ ௗ௦ (19) For the proposed control, the rotor current's error e(t) and the
ܸ௤௦ ؄  ߱௦ ߮௦ change of the error de(t) are used as inputs. The rotor voltage
 is used as output. The controller observes the pattern of the
error signal and its derivative of the direct and quadratic rotor
‡™‘”‹ƒ•–‡ƒ†›•–ƒ–‡߮௦ ؄ ܿ‫‡–‹”™‡™‘• ݁ݐ‬ǣ currents components control loops and correspondingly
 updates the output U so that the rotor current matches its
ܸௗ௦ ൌ Ͳ
൜ܸ ൌ  ߱ ߮ ሺʹͲሻ reference. These two signals are normalized through their
௤௦ ௦ ௦ respective scaling factors Ke and Kde. The output control
 signal dU is derived by multiplying the du/dt by the output
Š‡ ‹• ‘‡ –‡†–‘–Š‡‰”‹†•‘™‡Šƒ˜‡ǣ scale factor Kdu, and then integrated to generate the command
 signal.
ܸ௤௦ ൌ ܷ ൌ ܿ‫݁ݐ‬
൜ (21) The simplified block diagram of the fuzzy logic controller is
ܸௗ௦ ൌ Ͳ
presented on this figure 4.

௎ ௎
߮ௗ௦ ൌ ߮௦  ൌ ఠ ൌ ଶగ௙
†൝ ೞ (22)
߮௤௦ ൌ Ͳ

‘™„›—•‹‰–Š‡‡“—ƒ–‹‘•ሺͳʹǡͳ͵ሻǡ–Š‡‡“—ƒ–‹‘•ሺͺǡͻሻ
„‡ ‘‡ǣ

ௗூ೏ೝ Fig. 4: Fuzzy logic structure
ܸ௥ௗ ൌ  ܴ௥ ‫ܫ‬ௗ௥ ൅ ɐ‫ܮ‬௥ െ ߱௥ ɐ‫ܮ‬௥ ‫ܫ‬௤௥
ௗ௧
ௗூ೜ೝ ெ
ܸ௥௤ ൌ  ܴ௥ ‫ܫ‬௤௥ ൅ ɐ‫ܮ‬௥ ௗ௧
൅ ߱௥ ሺ௅ ߮௦ ൅ ɐ‫ܮ‬௥ ‫ܫ‬ௗ௥ ) Where the error e and its rate of change de are the input
ೞ variables; Ke , Kde and Ku are inputs and outputs scaling
‫ʹܯ‬
™‹–Šı =ቀͳ െ ௅ ቁ gains. For the proposed FLC, The inputs to the direct and
ೞ ௅ೝ quadrate axis rotor current fuzzy controllers are the d- and
q-axis rotor current errors.
ܸௗ௥ ܿ‫ܸ†ƒ݌݉݋‬௤௥ ‘’ƒ”‡–Š‡—Ž‹‡ƒ”˜‘Ž–ƒ‰‡•–Šƒ–
™‡Šƒ˜‡–‘ ‘’‡•ƒ–‡„›–Š‹• ‘–”‘ŽǤ a. Membership functions:
 A Gaussian curve membership function is used for inputs
ܸௗ௥ ܿ‫ ݌݉݋‬ൌ ‫ݓ‬௥ ߪ‫ܮ‬௥ ‫ܫ‬௤௥  and a triangular membership function is used for the output.
൝ ெ  (23)
ܸ௤௥ ܿ‫ ݌݉݋‬ൌ ‫ݓ‬௥ ሺߪ‫ܮ‬௥ ‫ܫ‬ௗ௥ ൅ ௅ ɔ௦ ሻ
ೞ

IV. FUZZY LOGIC CONTROLLER

The dynamic mathematical model of DFIG is a nonlinear,

complex and multivariable time-varying system. Fuzzy logic
control has the capability to control nonlinear, uncertain and
adaptive systems, which gives a strong robust performance for
parameters variation. So in order to design a fuzzy controller,
the following steps must be performed:
Fig 5: The membership functions for the inputs
• Fuzzification: The process of decomposing a system input
and/or output into one or more fuzzy sets using fuzzy
linguistic variables and membership functions.
 Fig 6: The membership functions for the output
Fig 9: The electromagnetic torque

b. The fuzzy rules

Rule bases of the fuzzy controller:

e(t)
P EZ N
de(t) P PB N NB
EZ PB EZ NB
N PB P NB

Table 1: Rule bases of the fuzzy controller

Fig 10: The wind speed profile
V. SIMULATIONS AND RESULTS

The simulation based on DFIG has been performed for a

variable speed wind profile. The parameters used for those
simulations are: ܴ௦ =0.455; ܴ௥ =0.62; ‫ܮ‬௦ =0.084; ‫ܮ‬௥ =0.081;
M=0.078; J=0.3125; F=6.73e-3; p=2;

To achieve the results of our study. The model is implemented

in Matlab/Simulink.

Fig 11: The mechanical speed

Fig 7: The wind speed profile

Fig 12: The electromagnetic torque

The results show that the fuzzy logic controller presents a

faster response in comparison with the PI controller.
A significant overshoot is seen with the PI controller, gives
good performances. The mechanical speed and the
electromagnetic torque track their references for wind speeds
Fig 8: The mechanical speed perfectly.
 VI. CONCLUSION

This paper presented the rotor flux oriented vector control

applied to a DFIG based wind turbine. A comparative study
was made between fuzzy logic controller and PI controller.
The fuzzy logic controller has the ability to track the
references compared to the conventional PI. FLC has the
advantage to be robust and simple to design, since it doesn’t
require the knowledge of the exact model.

VII. REFERENCES

[1] “Global wind energy outlook 2008”, Global Wind Energy Council,
Oct.J. Breckling, Ed., The Analysis of Directional Time Series:
Applications to Wind Speed and Direction, ser. Lecture Notes in
Statistics. Berlin, Germany: Springer, 1989, vol. 61.
[2] P. Thirumuraugan and R. Preethi, “Closed Loop Control of Multilevel
Inverter Using SVPWM for Grid Connected Photovoltaic System,”
vol. 2, no. 4, pp. 1561–1572, 2013
[3] Sayouti, Y.; Abbou, A.; Akherraz, M.; Mahmoudi, H., "Real-time DSP
implementation of DTC neural networkbased for induction motor
drive," Power Electronics, Machines and Drives (PEMD 2010), 5th
IET International Conference on, vol., no., pp.1, 5, 19-21 April 2010.
[4] G. V. Madhav and Y. P. Obulesu, “A Fuzzy Logic Control Strategy for
Doubly Fed Induction Generator for Improved Performance under
Faulty Operating Conditions,” vol. 4, no. 4, pp. 419–429, 2014
[5] Kh.Benyahiya,L.Boumediene, A.Mezouar,M.Moussaoui. ‘’A robust
control of doubly fed induction generator using in wind energy
conversion system’’.2eme conférence internationale des energies
renouvelables CIER 2014.
[6] J.N.razafinjaka, A.Tsiory Patrick,‘’Model reference adaptive control
based on fuzzy logic for a doubly fed induction generator used in a
chain of wind power’’
[7] Hamzaoui, F. Bouchafaa, and A. Talha, “DPC and DTC control
strategies for wind turbine of DFIG with flywheel energy storage
system dedicated to improving the quality of energy,” vol. 5176.
[8] M.Sharawy, N.Abdel-Rahim, Adel A.Shaltout, ‘’Modeling and control
of stand-alone doubly fed induction generator used in wind energy
conversion system’’ Recent trends in energy systems conference
(RTES) 2015