WK,QWHUQDWLRQDO&RQIHUHQFHRQ5HQHZDEOH(QHUJLHVIRU'HYHORSLQJ&RXQWULHV 5('(&

Backstepping Controllers Design for a Grid Connected

Wind-Photovoltaic Hybrid Power System
Mohammed El malah, Abdellfattah Ba-razzouk, El Hassane Abdelmounim, Mhamed Madark
System Analysis and Information Technology Laboratory
Science and Technical Faculty, University Hassan 1st
Settat, Morocco.
[email protected] ; [email protected] ; [email protected] ; [email protected]

Abstract—This paper focuses on the maximum power point Tem Generator electromagnetic torque.
tracking (MPPT) and unity power factor operation of a grid Tr Rotor circuit transformer.
connected wind-photovoltaic hybrid energy conversion system
based on the doubly fed induction generator (DFIG). The ܸǡ ‫ ݒ‬ Voltage symbols.
proposed research work is based on the nonlinear mathematical Xrefǡ ܺ෠ Reference of measurement X, estimated of X.
models of the wind turbine, the DFIG, the photovoltaic generator, 1, 2, 3 Three phases sub-indices.
the DC-DC boost converter and the three phase inverters. Field- r,t,Vw rotor and turbine speeds, wind speed.
oriented control technique is used to obtain a simplified model that
ߠǡ ߱ Phase angle and angular pulsation symbols.
facilitates the design of controllers, and a backstepping control
strategy is used to develop the control laws for power converters ߛ Ideality factor of the equivalent PV cell diode.
of the hybrid system. Numerical simulation, realized in ,  ǡߩ Blade pitch angle, tip speed ratio, air density.
MATLAB/Simulink environment under different atmospheric ߪ Leakage coefficient ߪ ൌ ͳ െ ‫ܯ‬ଶ Τ‫ܮ‬௦ ‫ܮ‬௥
conditions acting on the hybrid system, confirm the validity and
effectiveness of the proposed controllers. I. INTRODUCTION
Keywords- Backstepping control, hybrid pv-wind system, doubly Hybrid wind-photovoltaic energy conversion systems are of
fed induction generator, maximum power point tracking, unity great interest due to the complementarity of photovoltaic and
power factor, vector control. wind energy sources and therefore present an effective solution
to the problem of the intermittency of these two renewable
NOTATIONS AND SYMBOLS
sources.
b, f Boost converter and filter indices. An MPPT controller is needed to make the most of the
C, c Symbol and index of the DC-bus capacitor. available energy, on the other hand, grid codes, such as FERC
Cp Wind turbine power conversion efficiency. order No.661-A, require operation at power factors greater than
d, q Synchronous reference frame indices. 0.95 for renewable energy interconnection. Several control
E, Eref Solar radiation, reference radiation(1kW/m2). strategies have been developed to accommodate these two
requirements [1] - [6], but most are based on linear controllers
Eg0, q Band-gap energy, electron charge.
[2] - [5]. What is needed now is to exploit the advantage of non-
F Total viscous coefficient (Turbine & DFIG).
linear controls. Indeed, in recent years, the use of non-linear
G, R, J Gearbox ratio, blade radius, total inertia. analytical controls has brought improvements to the control
‫ܫ‬or݅, fs Current symbols, switching frequency. performance of several systems [8] - [11], and this article
‫ܫ‬௦௔௧ ሺ‫ܫ‬௥௦ ሻ PV cell saturation (reverses saturation) currents. contributes in the same direction.
‫ܫ‬௣௛ ǡ ‫ܫ‬௦௖ PV-cell photocurrent and short-circuits current. Based on the stator-flux oriented control technique, the
‫ܫ‬௥௖ DC-current absorbed by the RSC proposed controllers are developed in d-q synchronous
‫ܭ‬ǡ ‫ܭ‬௜ Boltzmann constant, PV-cell short-circuit reference frame. The outputs of the controllers are transformed
current temperature coefficient. into the three-phase coordinate system using the inverse dq-abc
‫ܮ‬,߮ Inductances and flux symbols. transformation, and then a PWM command is used to make
M, m Mutual inductance, transformation ratio of Tr. them suitable for switching the power converters. The principle
cell, pv
objectives are as follows:
PV-cell and PV generator indices.
p, Pn DFIG pole pair number, DFIG rated power. • To achieve maximum power point tracking for each source.
P, Q Active and reactive powers. • To establish unity power factor operation of the hybrid
Paer Aerodynamic power captured by the turbine. system.
ܴ௦௛ǡ ܴ௦௘ PV-cell intrinsic parallel and series resistors. • To stabilize the DC-bus voltage.
RSC, GSC Rotor-side converter, grid-side converter. The paper is organized as follows. In section II, the models
s, r, g Stator, rotor and grid-side converter indices. of hybrid system elements are presented. Section III is devoted
T , Tref Temperature, reference temperature (298,15 K). to the design of hybrid system controllers. Simulation results
Ta/r Aerodynamic torque applied to DFIG rotor. are discussed in section IV followed by a conclusion.

k,(((
Authorized licensed use limited to: Cornell University Library. Downloaded on September 16,2020 at 19:48:46 UTC from IEEE Xplore. Restrictions apply.
 WK,QWHUQDWLRQDO&RQIHUHQFHRQ5HQHZDEOH(QHUJLHVIRU'HYHORSLQJ&RXQWULHV 5('(&

II. SYSTEM MODELING connected in parallel, one string containing Ns cells connected
The basic schematic diagram of the proposed control system in series, can be deduced from (2) and the following equations:
is shown in figure1. ܸ௔௥௥௔௬ ൌ ܰ௦ ܸ௖௘௟௟ Ǣ ‫ܫ‬௔௥௥௔௬ ൌ ܰ௉ ‫ܫ‬௖௘௟௟ (3)
The PV generator considered in this paper comprises ten
SM55 panels connected in series, the electrical specifications
Pm for one panel are given in [1]. The power-voltage characteristics
Ps Pt
for solar radiance change are depicted in the following figure.
559.775
Tem , Ωr 559.77
559.765
Pg 500 176.8 177 177.2

400

P pv [W]
300 438.73
Vc
438.72
200 173.5 174
2
100 T=25°C;E=1000W/m
2
T=25°C;E=800W/m
p v.j pg

0 20 40 60 80 100 120 140 160 180 200

Vp[V]
Fig 3. Power-voltage characteristics of the PV-Generator

C. Overall State Space Model

In order to design a simple control system, the three-phase
V c − ref
variables are transformed into a synchronous orthogonal frame.
The d-q transformation matrix utilized is as follows:
ଶగ ଶగ
Fig 1. Proposed hybrid system.
ଶ
ܿ‫ݏ݋‬ሺߠሻ ܿ‫ݏ݋‬ሺߠ െ ሻ ܿ‫ݏ݋‬ሺߠ ൅ ሻ
ଷ ଷ
A. Wind Turbine Model ܶఏ ൌ ට ቎ ଶగ ଶగ ቏ (4)
ଷ
െ‫݊݅ݏ‬ሺߠሻ െ‫݊݅ݏ‬ሺߠ െ ሻ െ‫݊݅ݏ‬ሺߠ ൅ ሻ
ଷ ଷ
The basic equations used to model the behavior of the wind
where ߠ is the angular position of the d-q frame with respect to
turbine are as follows: [2]:
ଵ ଶ ଷ
the Ƚ-axis of Ƚ-Ⱦ reference frame. The state space model of the
‫ܲۓ‬௔௘௥ ൌ ‫ܥ‬௣ ሺߣǡ ߚሻߩߨܴ ܸ௪ଶ DFIG wind system in a d-q synchronous frame, whose d-axis is
ష೎ఱ
ۖ
ۖ‫ ܥ‬ሺߣǡ ߚሻ ൌ ܿ ቀ௖మ െ ܿ ߚ െ ܿ ቁ ݁ ഊ೔ ൅ ܿ ߣ aligned with the stator flux vector, is as follows [1]:
௣ ଵ ଷ ସ ଺ ோೝ ௏೎
ఒ೔ ሶ
π೟ ோ
(1) ‫ܫۓ‬ௗ௥ ൌ ߱௥ ‫ܫ‬௤௥ െ ௅ೝ ఙ ‫ܫ‬ௗ௥ ൅ ௅ೝ ఙ‫ܭ‬ௗ௥
‫ ߣ۔‬ൌ
௏ೢ ۖ‫ ܫ‬ሶ ൌ െ߱ ‫ ܫ‬െ ఠೝ ெ ܸ െ ோೝ ‫ ܫ‬൅ ௏೎ ‫ܭ‬
ۖ
ۖଵ ଵ ଴Ǥ଴ଷହ
ۖ ௤௥ ௥ ௗ௥ ఠೞ ௅ೞ ௅ೝ ఙ ௦ ௅ೝ ఙ ௤௥ ௅ೝ ఙ ௤௥
ۖ
‫ ە‬ൌ
ఒ೔ ఒା଴Ǥ଴଼ఉ
െ
ఉ య ାଵ
‫ܫ‬ௗ௚
ோ೑
ሶ ൌ ߱௚ ‫ܫ‬௤௚ െ ‫ܫ‬ௗ௚ െ ‫ܭ‬ௗ௚ ௏೎
௅೑ ௅೑ (5)
Fig.2 shows characteristics Cp(), for different values of . ‫۔‬
ோ೑ ௏೒
0.48
ۖ‫ܫ‬௤௚ ሶ ൌ െ߱௚ ‫ܫ‬ௗ௚ െ ‫ܫ‬௤௚ ൅ െ ௏೎ ‫ܭ‬௤௚
β =0 ۖ ௅೑ ௅೑ ௅೑
ۖ
β =2 0.4795
0.4
ଵ
‫ܸە‬௖ሶ ൌ ஼ ൫‫ܭ‬ௗ௚ ‫ܫ‬ௗ௚ ൅ ‫ܭ‬௤௚ ‫ܫ‬௤௚ െ  ‫ܫ‬௥௖ ൯
8 8.1 8.2
β =5
0.3
β =10
where:ܸ௦ ൌ ܸ௤௦ ǡ ܸ௚ ൌ ܸௗ௚ ൌ ݉Ǥ ܸ௦ , ߱௚ ൌ ߱௦ ൌ ߱௥ ൅ ‫݌‬π௥ , ‫ܫ‬௥௖
Cp

0.2 β =15 is the DC-current absorbed by the RSC, ሺ‫ܭ‬ௗ௚ ǡ ‫ܭ‬௤௚ ሻǡ ሺ‫ܭ‬ௗ௚ ǡ ‫ܭ‬௤௚ ሻǡ
0.1
β =20
൫‫ܫ‬ௗ௚ ǡ ‫ܫ‬௤௚ ൯ andሺ‫ܫ‬ௗ௥ ǡ ‫ܫ‬௤௥ ሻ are respectively, the components in the
β =25

β =30
d-q frame of the GSC control signal, the RSC control signal, the
0 2 4 6 8 10
λ
12 14 16 18 20 GSC current and the rotor current,ܸௗ௚ ൌ ܸ݉ௗ௦ ؆ Ͳ.
Fig 2. Power coefficient curves. Photovoltaic power conversion is governed by ([6]):
ூ೛ೡ ିூಽ್
B. Photovoltaic generator Model ሶ ൌ
ܸ௣௩ 
஼೛ೡ
An efficient mathematical model describing the behaviors of ቐ ௏೛ೡ ିሺଵି௨ሻ௏೎
(6)
photovoltaic cells is given by [6]: ‫ܫ‬௅ሶ ್ ൌ
௅್
೎೐೗೗ ೎೐೗೗ ೞ೐ ௏ ାூ ோ ௏ ାூ ோ
‫ܫۓ‬௖௘௟௟ ൌ ‫ܫ‬௣௛ െ ‫ܫ‬௦௔௧ ቂ݁‫ ݌ݔ‬ቀ‫ݍ‬ ቁ െ ͳቃ െ ೎೐೗೗ ೎೐೗೗ ೞ೐ where ‫ܫ‬௅ is the current of the boost converter inductance (‫ܮ‬௕ )
ఊ௄் ோೞ೓
ۖ ா and u the switching signal. Neglecting ohmic losses, the DC-
‫ܫ‬௣௛ ൌ ൣ‫ ܫ‬൅ ‫ܭ‬௜ ൫ܶ െ ܶ௥௘௙ ൯൧
ாೝ೐೑ ௦௖ (2) bus voltage dynamics can be expressed as follows:
‫۔‬ ଷ ଵ ௉
ۖ‫ ܫ‬ൌ ‫ ܫ‬൬ ் ൰ ݁‫݌ݔ‬ ି௤ா ೒బ ଵ ଵ ܸ௖ሶ ൌ ቀ‫ܭ‬ௗ௚ ‫ܫ‬ௗ௚ ൅ ‫ܭ‬௤௚ ‫ܫ‬௤௚ ൅ ೛ೡ െ  ‫ܫ‬௥௖ ቁ (7)
൬ െ ൰ ஼ ௏೎
‫ ە‬௦௔௧ ௥௦ ்
ೝ೐೑ ఊ௄ ் ்ೝ೐೑
Then the overall state space model of the hybrid system is
The model of the PV generator formed by Np strings deduced from (5), (6) and (7):

Authorized licensed use limited to: Cornell University Library. Downloaded on September 16,2020 at 19:48:46 UTC from IEEE Xplore. Restrictions apply.
 WK,QWHUQDWLRQDO&RQIHUHQFHRQ5HQHZDEOH(QHUJLHVIRU'HYHORSLQJ&RXQWULHV 5('(&

ோೝ ௏೎ ଵ
ሶ ‫ܥ‬௣̴௠௔௫ ߩߨܴହ .We therefore deduce:
‫ܫۓ‬ௗ௥ ൌ ߱௥ ‫ܫ‬௤௥ െ ௅ೝ ఙ ‫ܫ‬ௗ௥ ൅ ௅ೝ ఙ‫ܭ‬ௗ௥ where ‫ܭ‬௢௣௧ ൌ
ଶீ య ఒ೚೛೟
య

ۖ‫ ܫ‬ሶ ൌ െ߱ ‫ ܫ‬െ ఠೝ ெ ܸ െ ோೝ ‫ ܫ‬൅ ௏೎ ‫ܭ‬ ܶ௘௠̴௥௘௙ ൌ െ‫݌‬

ெ
߮௦ ‫ܫ‬௤௥̴௥௘௙ ൌ ‫ܨ‬π௥ െ ܶ௔Ȁ௥̴௥௘௙
ۖ ௤௥ ௥ ௗ௥ ఠೞ ௅ೞ ௅ೝ ఙ ௦ ௅ೝ ఙ ௤௥ ௅ೝ ఙ ௤௥ ௅ೞ
ۖ ூ೛ೡ ିூಽ್ ௅ೞ
ܸۖ௣௩ ሶ ൌ ‫ܫ‬௤௥̴௥௘௙ ൌ ൫‫ܭ‬௢௣௧ π௥ ଶ െ ‫ܨ‬π௥ ൯ (15)
஼೛ೡ ௣ெఝೞ
ۖ Ÿ ቐ ௅ೞ
ሶ‫ܫ‬௅ ൌ ௏೛ೡିሺଵି௨ሻ௏೎ ሶ
‫ܫ‬௤௥̴௥௘௙ ൌ πሶ௥ ൫ʹ‫ܭ‬௢௣௧ π௥ െ ‫ܨ‬൯
್ (8) ௣ெఝೞ
௅್
‫۔‬ In order to obtain UPF operation on the stator-side:
ۖ‫ܫ‬ௗ௚ ሶ ൌ ߱௦ ‫ܫ‬௤௚ െ ோ೑ ‫ܫ‬ௗ௚ െ ௏೎ ‫ܭ‬ௗ௚ ሺଵ଴ሻ
௅೑ ௅೑ ெ ௏ೞ మ ௏
ۖ ሳልሰ ܳ௦̴௥௘௙ ൌ െܸ௦ ‫ܫ‬ ൅ Ÿ ‫ܫ‬ௗ௥̴௥௘௙ ൌ ெఠೞ (16)
ோ
ۖ‫ ܫ‬ሶ ൌ െ߱ ‫ ܫ‬െ ‫ ܫ‬൅ ௏೒ െ ௏೎ ‫ܭ‬೑ ௅ೞ ௗ௥̴௥௘௙ ఠೞ ௅ೞ ೞ
ۖ ௤௚ ௦ ௗ௚ ௅೑ ௤௚ ௅೑ ௅೑ ௤௚ In order to obtain UPF operation of the GSC:
ۖ ଵ ௉೛ೡ ܳ௚̴ೝ೐೑ ൌ ܸ௚ ‫ܫ‬ௗ௚ೝ೐೑ Ÿ ‫ܫ‬ௗ௚̴௥௘௙ ൌ Ͳ (17)
ሶ
‫ܸە‬௖ ൌ ஼ ቀ‫ܭ‬ௗ௚ ‫ܫ‬ௗ௚ ൅ ‫ܭ‬௤௚ ‫ܫ‬௤௚ ൅ ௏೎ െ  ‫ܫ‬௥௖ ቁ The PV power derivative with respect to its voltage is chosen
III. CONTROLLER DESIGN as a controlled output. When the PV generator is operating in
its maximum state (fig. 3) this output becomes zero, therefore:
A. Description డ௉೛ೡ
൬ ൰ ൌͲ (18)
The objectives of the GSC controller are to maintain a డ௏೛ೡ
௥௘௙
constant DC bus voltage and ensure unit power factor operation
C. Rotor Side Inverter Control Laws
regardless of the direction of the rotor power flow. Whereas the
RSC controller provides independent control of the MPPT and Let us define the error ݁ଵ between ‫ܫ‬ௗ௥ and its desired value:
stator reactive power, which are implicitly controlled by acting ݁ଵ ൌ ‫ܫ‬௤௥ െ ‫ܫ‬௤௥̴௥௘௙ (19)
on the rotor current components, and the boost converter is used Its derivative with respect to time, using (8), is:
to track the maximum power point of the PV generator. The ெ௏ೞ ோೝ ௄೜ೝ ௏೎
݁ሶଵ ൌ െ߱௥ ‫ܫ‬ௗ௥ െ ݃ െ ‫ܫ‬ ൅ ሶ
െ ‫ܫ‬௤௥ (20)
grid voltage as well as the grid frequency are assumed to be ௅ೞ ௅ೝ ఙ ௅ೝ ఙ ௤௥ ௅ೝ ఙ ೝ೐೑ 
stable and the stator flux has then been estimated as follows: The first Lyapunov function candidate (LFC) is defined as:
߮ො௜ ൌ ‫׬‬ሺ‫ݒ‬௦௜ െ ܴ௦ ݅௜ ሻ݀‫  א ݅ݐ݀ ݅ݒ ׬ ≈ ݐ‬ሼͳǡ ʹǡ ͵ሽ (9) ଵ ௌ௢
ܸଵ ൌ ݁ଵ ଶ ֜ ܸଵሶ ൌ ݁ଵ ݁ଵሶ (21)
The stator flux phase angle Ʌୱ is extracted by using a phase ଶ
locked loop (PLL). The PLL algorithm tracksɅୱ by converging If we choose a dynamic control ‫ܭ‬௤௥ as follows:
the quadrature component of stator flux to zero [17]. ௅ೝ ఙ ோ ௚ெ௏
‫ܭ‬௤௥ ൌ ቀെܿଵ ݁ଵ ൅ ߱௥ ‫ܫ‬ௗ௥ ൅ ೝ ‫ܫ‬௤௥ ൅ ೞ ሶ
൅ ‫ܫ‬௤௥ ቁ (22)
௏೎ ௅ೝ ఙ ௅ೞ ௅ೝ ఙ ೝ೐೑

with ܿଵ ൐ Ͳ. ܸଵሶ become ܸଵሶ ൌ െܿଵ ݁ଵ ൑ Ͳଶ

The error between ‫ܫ‬ௗ௥ and its desired value is defined:

݁ଶ ൌ ‫ܫ‬ௗ௥ െ ‫ܫ‬ௗ௥̴௥௘௙ (23)
Its derivative with respect to time, using (8), is:
ோ ௄ ௏
ሶ
‫ܫ‬ௗ௥̴௥௘௙ ൌ ͲŸ݁ሶଶ ൌ ߱௥ ‫ܫ‬௤௥ െ ೝ ‫ܫ‬ௗ௥ ൅ ೏ೝ ೎ (24)
௅ೝ ఙ ௅ೝ ఙ
Let us consider the following new LFC:
ଵ ௌ௢
Fig 4. Principle of the PLL and its Simulink model. ܸଶ ൌ ܸଵ ൅ ݁ଶଶ ֜ ܸଶሶ ൌ ܸଵሶ ൅ ݁ଶ ݁ଶሶ (25)
ଶ
In synchronous d-q frame the electromagnetic torque, the The stabilizing RSC direct control Kdr is chosen as:
active and reactive powers are simplified as follows [1]: ௅ೝ ఙ ோೝ
‫ܭ‬ௗ௥ ൌ ሺെܿଶ ݁ଶ ൅  ‫ܫ‬ െ ߱௥ ‫ܫ‬௤௥ ሻ (26)
ெ ெ ௏೎ ௅ೝఙ ௗ௥
‫ܶۓ‬௘௠ ൌ െ‫ ݌‬௅ೞ ሺ߮ௗ௦ ‫ܫ‬௤௥ െ ߮௤௦ ‫ܫ‬ௗ௥ ሻ ൌ െ‫ ݌‬௅ೞ ߮௦ ‫ܫ‬௤௥ where ܿଶ is a strictly positive design parameter. We obtain:
ۖ ெ
ܲ௦ ൌ ܸௗ௦ ‫ܫ‬ௗ௦ ൅ ܸ௤௦ ‫ܫ‬௤௦ ൌ ܸ௤௦ ‫ܫ‬௤௦ ൌ െܸ௦ ‫ܫ‬௤௥
௅ೞ (10) ܸଶሶ ൌ െܿଵ ݁ଵଶ െ ܿଶ ݁ଶଶ ൑ Ͳ (27)
‫۔‬ మ
ۖܳ ൌ ܸ ‫ ܫ‬െ ܸ ‫ ܫ‬ൌ ‫ ܫ ݒ‬ൌ െܸ ‫ ܫ‬൅ ெ ௏ೞ D. Grid Side Inverter Control Laws
‫ ە‬௦ ௤௦ ௗ௦ ௗ௦ ௤௦ ௤௦ ௗ௦ ௦ ௗ௥ ௅ೞ ఠೞ ή௅ೞ
An integral action is added to improve the control
B. Dynamic Outputs and Their References. performance of the GSC. The first tracking error is:
Rotor dynamics is given by this mechanical equation: ݁ଷ ൌ ‫ܫ‬ௗ௚ െ ‫ܫ‬ௗ௚̴௥௘௙ (28)
‫ܬ‬πሶ௥ ൌ ܶ௔Ȁ௥ ൅ܶ௘௠ െ ‫ܨ‬π௥ (11) Its time-derivative, using (8), is:
When the turbine is working on the maximum power point, ோ೑ ௏೎
ሶ
݁ሶଷ ൌ ߱௚ ‫ܫ‬௤௚ െ ‫ܫ‬ െ ‫ܭ‬ௗ௚ െ ‫ܫ‬ௗ௚̴௥௘௙ (29)
=opt and Cp=Cp_max, from (1) we get: ௅೑ ௗ௚ ௅೑
ଷ
ଵ ଵ π೟ ோ Now, the LFC is chosen as follows:
ܲ௠௔௫ ൌ ‫ܥ‬௣̴௠௔௫ ߩߨܴଶ ܸ௪ଷ ൌ ‫ܥ‬௣̴௠௔௫ ߩߨܴଶ ൬ ൰ (12)
ଶ ଶ ఒ೚೛೟ ଵ ଵ
௉೘ೌೣ ܸଷ ൌ ݁ଷ ଶ ൅ ߪଷ ଶ Ÿܸଷሶ ൌ ݁ଷ ሺ݁ሶଷ ൅ ܿଷ௜ ߪଷ ሻ (30)
Ÿ ܶ௔Ȁ௥̴௥௘௙ ൌ ൌ ‫ܭ‬௢௣௧ π௥ ଶ (13) ௧
ଶ ଶ
πೝ
with ߪଷ = ܿଷ௜ ‫׬‬଴ ݁ଷ ሺ߬ሻ ݀߬, ܿଷ௜ : integral action design parameter.

Authorized licensed use limited to: Cornell University Library. Downloaded on September 16,2020 at 19:48:46 UTC from IEEE Xplore. Restrictions apply.
 WK,QWHUQDWLRQDO&RQIHUHQFHRQ5HQHZDEOH(QHUJLHVIRU'HYHORSLQJ&RXQWULHV 5('(&

௘ ௘
Then, ‫ܭ‬ௗ௚ is chosen as follows: ܸହሶ ൌ െܿହ ݁ହଶ െ ݂ ఱ ల (47)
஼೛ೡ
௏೎ ோ೑
‫ܭ‬ௗ௚ ൌ ܿଷ ݁ଷ ൅ ܿଷ௜ ߪଷ ൅ ߱௚ ‫ܫ‬௤௚ െ ‫ܫ‬ (31) ௘ ௘ ௏ ିሺଵି௨ሻ௏೎
௅೑ ௅೑ ௗ௚ Ÿ ܸ଺ሶ ൌ െܿହ ݁ହଶ െ ݂ ఱ ల ൅ ݁଺ ሺ ೛ೡ െ ߙሶ ଵ ሻ (48)
஼೛ೡ ௅್
where ܿଷ is a strictly positive design parameter.ܸଷሶ becomes:
Then the command u is chosen as follows:
ܸଷሶ ൌ െܿଷ ݁ଷ ଶ ൑ Ͳ (32) ௅್ ௘ఱ ௏೛ೡ
‫ݑ‬ൌ ൬െܿ଺ ݁଺ ൅ ߙሶ ଵ ൅ ݂ ൰െ ൅ͳ (49)
The DC-bus tracking error and its time-derivative are: ௏೎ ஼೛ೡ ௏೎
݁ସ ൌ ܸ௖ െ ܸ௖̴௥௘௙ (33) where ଺ is a strictly positive design parameter.ܸ଺ሶ becomes:
ሶ
ሺ଼ሻଵ ௉ ܸ଺ሶ ൌ െܿହ ݁ହଶ െ ܿ଺ ݁଺ଶ ൑ Ͳ (50)
ܸ௖̴௥௘௙ ൌ Ͳ ሳሰ ݁ሶସ ൌ ቀ‫ܭ‬ௗ௚ ‫ܫ‬ௗ௚ ൅ ‫ܭ‬௤௚ ‫ܫ‬௤௚ ൅ ೛ೡ െ ‫ܫ‬௥௖ ቁ (34)
஼ ௏೎
F. Overall Stability Analysis
Similarly, the LFC is chosen as:
ଵ ଵ ௌ௢ The overall LFC can be defined as:
ܸସ ൌ ݁ସ ଶ ൅ ߪସ ଶ  ֜ ܸସሶ ൌ ݁ସ ሺ݁ሶସ ൅ ܿସ௜ ߪସ ሻ (35) ௘೔ మ ఙ೔ మ
ଶ
௧
ଶ ்ܸ ൌ ܸଶ ൅ ܸଷ ൅ ܸସ ൅ ܸ଺ ൌ σ଺௜ୀଵ ൅ σସ௜ୀଷ (51)
ଶ ଶ
withߪସ ൌ ܿସ௜ ‫׬‬଴ ݁ସ ሺ߬ሻ ݀߬, ܿସ௜ : integral action design parameter. ሺଶ଻ሻǡሺଷଶሻǡሺଷ଻ሻƬሺହ଴ሻ
Now, if ‫ܭ‬ௗ௚ and ‫ܭ‬௤௚ are chosen as follows: ሳልልልልልልልልልልልልልልልሰ ܸሶ் ൌ െ σ଺௜ୀଵ ܿ௜ ݁௜ ଶ (52)
ூ೏೒ ூ೜೒ ଵ ௉೛ೡ ሶ
்ܸ is negative. Consequently, the tracking errors are
‫ܭ‬ௗ௚ ൅ ‫ܭ‬௤௚ ൌ െܿସ ݁ସ െ ܿସ௜ ߪସ ൅ ቀ‫ܫ‬௥௖ െ ቁ (36) asymptotically stable and converge towards zero.
஼ ஼ ஼ ௏೎
where ܿସ is a strictly positive design parameter.ܸଷሶ becomes:
IV. SIMULATION RESULTS
ܸସሶ ൌ െܿସ ݁ସ ଶ ൑ Ͳ (37)
In order to evaluate the robustness of the proposed
Kdg and Kqg are then calculated, using (31) and (36), as follows: algorithms under changing atmospheric conditions, the
‫ܭ‬ௗ௚ ‫ܣ ܣ‬ଶ ିଵ ‫ܤ‬ଵ simulation is performed with the following solar radiation and
൤ ൨ൌ൤ ଵ ൨ ൤ ൨ (38)
‫ܭ‬௤௚ ‫ܣ‬ଷ ‫ܣ‬ସ ‫ܤ‬ଶ wind speed profiles.
ோ೑ ௏೎ 12
where: ‫ܤ‬ଵ ൌ ܿଷ ݁ଷ ൅ ܿଷ௜ ߪଷ ൅ ߱௚ ‫ܫ‬௤௚ െ ‫; ܫ‬ ‫ܣ‬ଵ ൌ Ǣ ‫ܣ‬ଶ ൌ Ͳ 11
௅೑ ௗ௚
Irradiation [W /m ²]

1000
௅೑ 10.95

W in d sp ee d [m / s ]
11
ଵ ௉೛ೡ ூ೏೒ ூ೜೒ 10.9
‫ܤ‬ଶ ൌ െܿସ ݁ସ െ ܿସ௜ ߪସ ൅ ቀ‫ܫ‬௥௖ െ ቁ;‫ܣ‬ଷ ൌ Ǣ‫ܣ‬ସ ൌ 900 12.9 13 13.1
஼ ௏೎ ஼ ஼ 10

800
E. Boost Control Law 0 10 20 30 40 9
t [s]
Step 1: The tracking error is defined: Fig 5. solar irradiation profile 8
0 5 10 15 20 25 30 35 40
డ௉೛ೡ డ௉೛ೡ డூ೛ೡ t [s]
݁ହ ൌ െ൬ ൰ ൌ ‫ܫ‬௣௩ ൅ ܸ௣௩ (39) Fig 6. Wind speed profile.
డ௏೛ೡ డ௏೛ೡ డ௏೛ೡ
௥௘௙
The important parameters are listed in the following table.
Its derivative with respect to time, using (8):
TABLE I. CHARACTERISTICS
డమ ூ೛ೡ డூ೛ೡ ூ೛ೡ ିூಽ್
݁ሶହ ൌ ൬ܸ௣௩ మ ൅ʹ ൰ሺ ሻ (40) Hybrid system
డ௏೛ೡ డ௏೛ೡ ஼೛ೡ
ܲ௡ =1kW ߩ=1.22kg/m3 ܴ௥ =0.88Ω
The first LFC is defined: ܷ௦ =190V R=0.9 m C=3000μF
ଵ
ܸହ ൌ ݁ହ ଶ (41) p=3 F=0.00064 Nms/rad M=90.1 mH
ଶ ‫ܥ‬௣̴௠௔௫ =4.8 G=1.3 ‫ܮ‬௦ =93.1 mH
డூ ூ ିூಽ್
Ÿ ܸହሶ ൌ ݁ହ ݁ሶହ ൌ ݁ହ ൬݄ܸ௣௩ ൅ ʹ ೛ೡ ൰ ሺ
೛ೡ
ሻ (42) opt=8.13 ܴ௦ =1.1Ω ‫ܮ‬௙ =12.6 mH
డ௏೛ೡ ஼೛ೡ
ܸ௖̴௥௘௙ =300V ܴ௙ =0.5655Ω ‫ܮ‬௥ ൌ93.1 mH
డమ ூ೛ೡ ‫ܮ‬௕ =20݉‫ܪ‬
where: ݄ ൌ మ . To make ܸହሶ negative, ‫ܫ‬௅್ is taken as a ‫ܥ‬௣௩ =2000μF
డ௏೛ೡ
Controllers
virtual control provided that its desired value is: ݂௦ =10kHz ܿଵ =7×105 ܿଶ =6×105 ܿଷ =57×103 ܿଷ௜ =3×102
஼೛ೡ ܿସ =65×103 ܿସ௜ =5×104 ܿହ =104 ܿ଺ =104
ߙଵ ൌ ‫ܫ‬௣௩ ൅ ܿହ ݁ହ (43)
௙
డூ೛ೡ
where:݂ ൌ ݄ܸ௣௩ ൅ ʹ andܿହ is a positive design parameter. 1300
డ௏೛ೡ 1226
1224
Rotor speed [tr/min]

Step2: The error between ‫ܫ‬௅ and its desired value is:
1222
1200 13 13.2

݁଺ ൌ ‫ܫ‬௅್ െ ߙଵ (44)
Its derivative using (8) is: 1100
௏ ିሺଵି௨ሻ௏೎
݁ሶ଺ ൌ ‫ܫ‬௅ሶ െ ߙሶ ଵ ൌ ೛ೡ
್
െ ߙሶ ଵ (45)
௅್ 1000
௙௘ሶ ఱ ି௙ሶ ௘ఱ డூ೛ೡ
ሶ ; ݂ሶ ൌ ͵݄ܸ௣௩
ሶ ൅ ݄ሶܸ௣௩
Sub-synchronous
where: ߙሶ ଵ ൌ ܿହ ‫ܥ‬௣௩ ൅ ܸ௣௩
௙మ డ௏೛ೡ 0 5 10 15 20 25 30 35 40
t [s]
Let us consider an augmented LFC: Fig 7. Rotor Speed.
ଵ
ܸ଺ ൌ ܸହ ൅ ݁଺ଶ Ÿ ܸ଺ሶ ൌ ܸହሶ ൅ ݁଺ ݁ሶ଺ (46) The DFIG operates in super-synchronous and also in sub-
ଶ
According to (42), (43) and (44): synchronous mode as shown in Fig 7.

Authorized licensed use limited to: Cornell University Library. Downloaded on September 16,2020 at 19:48:46 UTC from IEEE Xplore. Restrictions apply.
 WK,QWHUQDWLRQDO&RQIHUHQFHRQ5HQHZDEOH(QHUJLHVIRU'HYHORSLQJ&RXQWULHV 5('(&

1
500
Stator flux [Wb]

Stator power
0.5 x 10
-4 0
5
0
-5 -500
0 5 10 15 20 25 30 35 40
0 -3
x 10
2 -1000 Ps[W]
0 Direct stator flux
-2 Qs[VAR]
0 5 10 15 20 25 Quadratue
30 stator
35 flux 40
-0.5 0 5 10 15 20 25 30 35 40
0 5 10 15 20 25 30 35 40 t[s]
t[s]
Fig 8. Stator flux behavior. Fig 13. stator power behavior.
Iqr 400
Rotor quadratue current [A]

8 Iqr-ref
6 300

4 310

Vc [V]
200 305 300.04
2 300 300.02
300
0 5 10 299.98
0 100
29.8 30 30.2
-2
0 5 10 15 20 25 30 35 40 0
0 5 10 15 20 25 30 35 40
t [s] t[s]
Fig 9. Rotor quadrature current behavior. Fig 14. DC-link voltage.
Tem 2000
Electromagnetic torque(Nm)

0
Tem-ref -200
0 50 -400
1000 0 -600

Pg(W),Qg(Var)
-50 -800
-5 10 12 14 16 18 0 10 20 30 40

0
-10

-1000
-15
Pg(W)
0 5 10 15 20 25 30 35 40 Qg(Var)
-2000
t(s) 0 5 10 15 20 25 30 35 40
Fig 10. Electromagnetic torque behavior. t [s]
Fig 15. Electrical power injected by DFIG rotor and the PV generator.
Figure 8 clearly shows that the PLL has achieved its
objective. Fig. 9 and 10 show the accurate convergence of the 5

rotor quadrature current, and thus the electromagnetic torque, 20 0

Rotor current [A]

to its reference values. Fig.11 confirms that the aerodynamic

-5
power has reached its maximum value (Cp=Cp_max)
10 23 23.5 24 24.5 25 25.5
0.5

0.45 0
0.48
0.4798
Cp

0.4 0 5 10 15 20 25 30 35 40
0 5 10 20 15 25 30 35 40
t[s]
0.35 Fig 16. Rotor quadrature current behavior.
FIG. 14 illustrates the accurate convergence of the DC-bus
0 5 10 15 20
t[s]
25 30 35 40 voltage. Fig.15 confirms that the GSC operates at the UPF
Fig 11. Cp behavior in presence of wind changes. regardless of the rotor current behavior affected by the DFIG
FIG. 12 shows the convergence of the direct component of operating mode (Fig.16).
the rotor current towards its reference, and FIG. 13 shows that 600
the stator reactive power is, as expected, close to zero.
PV-Power & V oltage

400
Rotor direct current [A]

10 559.7724 438.7346
559.7724
559.7723 438.7345
8.95 9 9.05 14.98 15 15.02
5 200
178
176
Pp [W]
174
0 Idr Vp[V]
10 15 20
Idr-ref 0
0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40
t [s] t [s]
Fig 12. Rotor direct current behavior. Fig 17. PV generator power/voltage behaviors.

Authorized licensed use limited to: Cornell University Library. Downloaded on September 16,2020 at 19:48:46 UTC from IEEE Xplore. Restrictions apply.
 WK,QWHUQDWLRQDO&RQIHUHQFHRQ5HQHZDEOH(QHUJLHVIRU'HYHORSLQJ&RXQWULHV 5('(&

As shown in Fig. 17, the PV power has also reached its HPC, PC and MPPT control," 2014 IEEE 23rd International Symposium
on Industrial Electronics (ISIE), Istanbul, 2014, pp. 2024-2029.
maximum value. The total injected power and current are
[3] D. Nguyen, G. Fujita,"Analysis of sensorless MPPT method for hybrid
presented in Figs. 18 and 19, respectively. The total reactive PV–Wind system using DFIG Wind Turbines"Sustain. Energy Grids
power injected is close to zero, confirming that the hybrid Netw., 5 (2016), pp. 50-57.
system is operating at UPF. [4] S. Datta, J. P. Mishra and A. K. Roy, "Active and reactive power control
of a grid connected speed sensor less DFIG based wind energy conversion
1000 system," 2015 International Conference on Energy, Power and
Pt [W]
Qt [VAR] Environment: Towards Sustainable Growth, Shillong, 2015, pp. 1-6.
Total electrical power

[5] Haisheng Sun et al., "DFIG wind power generation based on back-to-back
0
PWM converter," 2009 International Conference on Mechatronics and
Automation, Changchun, 2009, pp. 2276-2280.
-1000
[6] M. Kumar, N. Gupta and R. Garg, "Unity power factor control of grid
connected SPV system using instantaneous symmetrical component
theory," 2016 IEEE 1st International Conference on Power Electronics,
Intelligent Control and Energy Systems, Delhi, 2016, pp. 1-6.
-2000
[7] M. El Malah, A. Ba-razzouk, M. Guisser, E. Abdelmounim, M.Madark,
0 5 10 1520 25 30 35 40 H. Bahri, "Robust Nonlinear Controller Design for Three Phase Grid
t[s] Connected PV Systems with Improved Power Quality", International
Fig 18. Behavior of the total electrical power injected. Review Of Automatic Control, vol. 10, no. 4, pp. 359 – 370, 2017.
[8] A. Ba-Razzouk, A. Cheriti, G. Olivier and P. Sicard, "Field oriented
100 control of induction motors using neural networks decouplers,"
Proceedings of IECON '95 - 21st Annual Conference on IEEE Industrial
100 0 Electronics, Orlando, FL, USA, 1995, pp. 1428-1433 vol.2.
10 *T otal current [A ]
G rid v oltage [V ]

-100 [9] A. M. Omara, M. Sleptsov and A. A. Z. Diab, "Cascaded fuzzy logic

based direct torque control of interior permanent magnet synchronous
9.9 9.95 10 10.05 10.1 motor for variable speed electric drive systems," 2018 25th International
0
Workshop on Electric Drives: Optimization in Control of Electric Drives
100
(IWED), Moscow, 2018, pp. 1-6.
0
[10] M.Madark, A. Ba-razzouk, E. Abdelmounim, M. El malah, "Adaptive
-100 v1 Vector Control of Induction Motor Using CTMVC," International Review
it1 -100
on Modelling and Simulations (IREMOS), vol. 10, no. 4, pp. 303 – 312,
2017.
0 10 20 30 40 24.95 25 25.05 25.1 25.15 25.2 25.25
t [s] [11] Y. Zhao, S. Li, L. Zhao, X. Guo, Q. Wang and Y. Wen, "Sliding-mode
Fig 19. Behavior of the total injected current. control of permanent magnetic spherical motor based on co-simulation
platform," 2016 IEEE 11th Conference on Industrial Electronics and
Finally, the proposed control strategy has demonstrated that Applications (ICIEA), Hefei, 2016, pp. 119-123.
it is capable of achieving all the objectives set out in the first [12] Y. Majdoub, A. Abbou, M.Akherraz, R. El Akhrif, "Intelligent
section with very good performance. Backstepping Control of Variable Speed DFIG-Wind Turbine Under
Unbalanced Grid Voltage Conditions Using Genetic Algorithm
Optimization," International Review of Electrical Engineering
V. CONCLUSION (IREE),Vol. 10(Issue6):716-726, 2015.
This paper proposes nonlinear controllers for a grid- [13] M. El Azzaoui, H. Mahmoudi and K. Boudaraia, "Analysis and control of
connected wind-photovoltaic hybrid system. The main grid connected DFIG and solar PV based hybrid energy system," 2016
International Renewable and Sustainable Energy Conference (IRSEC),
objectives were to track accurately the maximum power point, Marrakech, 2016, pp. 1104-1109.
ensure an adequate grid connection with unity power factor. [14] S. Kumar Tiwari, B. Singh and P. K. Goel, "Design and Control of
Based on the stator-field oriented control technique a Microgrid Fed by Renewable Energy Generating Sources," in IEEE
backstepping control laws have been developed for each Transactions on Industry Applications, vol. 54, no. 3, pp. 2041-2050,
electronic converter. Simulation results demonstrated that the May-June 2018.
proposed controllers provides a stable hybrid system. The [15] A. Muhtadi and A. M. Saleque, "Modeling and simulation of a microgrid
consisting solar PV & DFIG based wind energy conversion system for St.
MPPT control, DC-bus stabilization and UPF operation were Martin's island," 2017 IEEE 3rd International Conference on Engineering
ensured even in case of large variations in atmospheric Technologies and Social Sciences (ICETSS), Bangkok, 2017, pp. 1-6.
conditions with good tracking properties. [16] G. P., A. Verma and A. Mitra, "Impact of Hybrid Renewable Sources on
the Transient Stability of Multi-machine Power System," 2018 IEEE
REFERENCES International Conference on Power Electronics, Drives and Energy
Systems (PEDES), Chennai, India, 2018, pp. 1-5.
[1] M. El Malah, A. Ba-razzouk, E. Abdelmounim, M.Madark,"Robust
Nonlinear Sensorless MPPT Control with Unity Power Factor for Grid [17] Se-Kyo Chung, "A phase tracking system for three phase utility interface
Connected DFIG Wind Turbines," International Review on Modelling inverters, " IEEE Transactions on Power Electronics, Vol. 15(Issue3):
and Simulations (IREMOS), vol. 11, no. 5, pp. 313 – 324, 2018. 431 – 438, MAY 2000.
[2] C. Rosa, D. Vinnikov, E. Romero-Cadaval, V. Pires and J. Martins,
"Hybrid low-power Wind Generation and PV gridconnected system with

Authorized licensed use limited to: Cornell University Library. Downloaded on September 16,2020 at 19:48:46 UTC from IEEE Xplore. Restrictions apply.