Modelling, Measurement and Control A

Vol. 91, No. 3, September, 2018, pp. 123-130

Journal homepage:http://iieta.org/Journals/MMC/MMC_A

Experimental investigations on grid integrated wind energy storage system using neuro fuzzy
controller
Krishnan Suresh1, Attuluri R.Vijay Babu2*, Perumal M. Venkatesh3

RC-RES, Department of Electrical and Electronics Engineering, Vignan’s Foundation for Science, Technology & Research,
Guntur 522213, Andhra Pradesh, India

Corresponding Author Email: [email protected]

https://doi.org/10.18280/mmc_a.910304 ABSTRACT

Received: 28 July 2018 This paper presents harnessing of maximum wind energy from natural resource whenever
Accepted: 30 September 2018 it’s available. The power electronic converters role is important In between sources and
load. The load may be linear and non-linear in nature, so converters performance decides
Keywords: the efficiency of the system. Proper controller can switch the converter in the desired time
speedgoat, bidirectional DC-DC converter, and improve the system performance and stability. Many controllers are suggests to
boost inverter, ANFIS control the converter to get better performance in at output side. The proposed system
also has boost converter, bidirectional DC-DC converter and inverter for grid and wind
energy integration. The boost inverter/buck rectifier in this system is controlled by
ANFIS controller is for better output, boost and bidirectional DC-DC converters are
controlled by PID controller in closed loop. Overall operations are based on modes main
controller speedgoat, which is control the system operation in different modes. Any
variation happening in the input, storage and load parameters speedgoat changing the
mode and operate the system is in effective way. Based on the system conditions
speedgoat generates control signal for the control breakers, these control breakers
changing modes of operation. ANFIS, PID and speedgoat are the three controllers
combined together which harness maximum wind energy and this system is applicable for
both linear and non-linear loads in domestic applications.

1. INTRODUCTION mode controller, these controller are robust [3-5]. Even

though these are in robust, it has some drawbacks related to
Modern controllers are very helpful for energy conversion control parameter selection constraints and complex theory.
system, now days most of the energy storage systems are So, this controller performance is not efficient and effective
fully depends on intelligent controllers such as digital logic in the inverter topology [6-8]. These drawbacks are overcome
and fuzzy logic [1-2]. by an Adaptive Neuro-Fuzzy Inference System (ANFIS), it
has advantages of fuzzy logic and neural networks which are
fulfill the alteration rule based demand and it can determine
the inference logic rules. ANFIS can control the grid tied
bidirectional converter in three modes such as grid tied boost
inverter mode, grid tied buck rectifier mode and stand-alone
mode in an effective way [9].
In conventional energy storage systems the converters like
uncontrolled AC-DC rectifier, PID controlled boost converter,
PID controlled bidirectional DC-DC converters are also using
for power conversion stages with grid tied inverter [10-12].
These converters are connected between Wind Energy
Conversion System (WECS), Energy Storage Device (ESD)
and load/grid for effective wind energy harnessing; the
converters in all the stages are controlled by separate
controllers [13]. But converters in these systems are
operating in closed loop but the overall system is operated by
open loop configuration. There are some serious dis-
Figure 1. Wind energy conversion system with linear and advantages related to this open loop configuration such as
non-linear load block diagram non-reliable, inefficient operation and less utilization of
natural energy [14-15]. To overcome these draw back the
This fuzzy logic behavior is good for tremendous system should be operate in closed loop from some parameter
intelligent control technique when compare to digital logic. considerations. This proposed system considers the
The fuzzy logic consist of fuzzy controller and fuzzy sliding parameters such as wind speed (v), state of charge (%) and

123
load position (ON/OFF). Based on these parameters the gets supply from primary source otherwise the generating
speedgoat generates control signal and given to control power is directly export to grid. If the load is in ON condition,
breakers which operates the system in different modes for an the system acts like stand-alone otherwise operates in in grid
effective operation with linear and non-linear load is as connected inverter mode.
shown in figure 1.

2. FIVE MODES OF OPERATION

2.1 Primary source storage mode

Primary source of the system is wind energy. Based on the

first two parameters consideration the wind speed (v) should
be more than 5 m/s and battery charging level has to be less
than 20%. Then the third parameter load position may be
ON/OFF. The primary source is directly supply to only
battery for the purpose of charging. In this mode load is said
to be ON condition, it gets supply from secondary source
otherwise the grid is isolated from the system. Figure 2
represents primary source storage mode.

Figure 3. Primary source wind storage-output mode

Figure 2. Primary source storage mode

2.2 Primary source storage-output mode

Figure 4. Primary source output mode
In the second mode of operation primary source wind
energy is given to output load/grid and battery for charging. 2.4 Battery power output mode
Under the parameters consideration the wind speed (v)
should be more than 5 m/s and battery charging level has to
be in between (40-80) %. Then the third parameter load
position may be ON/OFF. The primary source is supply to
both battery for the purpose of charging and output. In this
mode load is said to be ON condition, it gets supply from
primary source otherwise the generating power is export to
grid. (Fig 3)

2.3 Primary source output mode

Third mode (fig 4) of operation is primary source output

mode. In this mode primary source wind energy is directly
given to output load/grid. The parameters such as wind speed
(v) should be more than 5 m/s and battery charging level has Figure 5. Battery power output mode
to be reaching above 90 %. Then the third parameter load
position may be ON/OFF. The primary source is supply to Back-up source of the system is battery power. In this
only output. In this mode load is said to be ON condition, it mode primary source wind energy is less than 5 m/s. Due to

124
low wind speed, the power is not sufficient for output and the Table 1. By the modes conditions the converters like
battery charging level is already in above 90 %. Then the UDC, BDC and inverter are controlled by main controller in
third parameter load position should be in ON position. The different modes. The load and UDC has two positions ON-1
primary source is not available so battery supplies the power and OFF-0. BDC also two positions 1-boost and 2-buck. The
to load. In this mode load is said to be ON condition, it gets inverter has four modes namely 0-OFF, 1-standalone, 3-grid-
supply from back-up source otherwise the total system is in inverter and 4-grid-rectifier.
OFF until the wind speed reach above 5 m/s. If the load is in
ON condition, the system acts like stand-alone system. (fig 5).

2.5 Grid power storage-load mode

In the fifth mode of operation secondary source grid power

is given to output load and battery for charging. The
condition for this mode is wind speed (v) should be less than
5 m/s and battery charging level is less than 20 %. Then the
third parameter load position may be either ON/OFF. The
primary source is supply to both battery for the purpose of
charging and load. In this mode load is said to be ON
condition, it gets supply from secondary source otherwise
only charging the battery from grid. This mode will continue
until the wind speed reaches the level of 5 m/s. (fig 6).
Figure 6. Grid power storage-load mode
All the five modes of operations are clearly explained in

Table 1. Modes of operation

Modes condition
Mode Wind speed Control UDC BDC Inverter mode
Load Battery(%)
m/s
1 <40 >=5 1,2,5 & 11 1 0 0
1
0 <40 >=5 1,2 & 5 1 0 0
1 >=40 >=5 1,2,5, 6 & 10 1 1 1
2
0 >=40 >=5 1,2,5,6 & 7, 8, 9 1 1 2
1 >=40 &<80 >=5 5,6 & 10 1 1 1
3
0 >=40 &<80 >=5 5,6,7,8 & 9 1 1 2
4 1 >=40 <5 3,4,6 & 10 0 1 1
1 <40 <5 3,4,6,7,8,9,& 11 0 2 3
5
0 <80 <5 3,4,6,7,8 & 9 0 2 3

3. ANALYSIS OF WIND ENERGY SYSTEM PT v

2
v
Cp = = 0.5 * (1 − 22 ) * (1 − 2 )
P v1 v1
Each stage of conversion is analysed and composed by (3)
mathematical equations. Every stages of conversion consist
of converters and controllers in closed loop and overall 3.2 PMSG
system is controlled by separate main controller called
speedgoat. Mathematical modeling of Permanent Magnet
Synchronous
3.1 Wind energy conversion Generator (PMSG) can be written in the following
equations 4 and 5.
The design of wind turbine is taken from [1], horizontal
axis wind turbine has to analyse from equations 1, 2 and 3. did Ra Lq 1
Wind passes through in area area (A) and wind speed (v) of a = id = piq + u d
dt Ld Ld Ld (4)
particular velocity. The power of wind is expressed in
equation 1.
did Ra L 1 1
= iq = d pid + pm + u q
p = 0.5 *  * A * v 3
(1) dt Lq Lq Ld Lq
(5)

Mechanical power is expressed in the following eqn 2. The EMF equation of PMSG is written in equation 6.

pT = 0.5 *  * A * (v1 + v2 ) * (v1 + v2 )

2 2
(2)
E ph = 4.44 fT ph K w
(6)

The rotor coefficient is expressed by the following eqn. 3

125
3.3 Rectifier unit  di L   rc + rL 1 
 dt  − L −
L    i L  + 0 v
The three phase AC power from permanent magnet  dv  =  1 1   vc  0 s
synchronous generator is supply to 3 phase bridge rectifier  c  − 
 dt   C RC  (14)
for AC-DC conversion.
The phase voltage amplitude Vm and output average
voltage Vo is expressed by the equations 7 & 8. The small signal analysis of AC perturbation in a DC
operating point, the circuit variables is added to the ac
perturbations is expressed in equation 15.
Vm = Vrms 2 (7)
  
d ss = Dss + d ss x = X + x vs = Vs + vs
3 2 (15)
Vr = Vm
 (8) AC small signal is denoted by cap symbol. The
perturbation approximation is small so negligible. Steady
3.4 Uni-directional converter (UDC) state value is expressed in equation 16.
The output of unidirectional boost converter can be   
analysed in the input voltage and duty cycle. All stages of d ss x vs
 1  1  1
boost mode are controlled by adjusting the duty cycle Dss X Vs
(16)
automatically with closed loop controller. The output voltage
and duty cycle is expressed by the equations 9 and 10. It provides the final linearized ac small signal model.

Vr •
Vb =    
(1 − D) (9) x = A x+ BVs + ( A1 − A2 )X + (B1 − B2 )Vs d ss (17)

Ton T The transfer function of boost converter in the boost mode

= on = D
Ton + Toff Ts operation is expressed in s-domain. Thus the s-domain form
(10) of boost converter is expressed in an equation 18.
The current-programmed state equation (19) is minimized
3.5 Bidirectional converter (BDC) to a function of X(s) &Vcontrol(s). It allows x(s) to an
expression as in Vcontrol(s) in a matrix multiplication,
The primary switch Q1 is ON position during boost mode
it is expressed in equation 11  r +r 1 
   − c L −      1    Vs  
s i ( s )
 L  =  L L   iL ( s)  +   vs (s) +  2 LN  d ss (s)
•   2 LN
 s v ( s)  1 −  vc (s)  0 
1  0 
X = A1 X + B1VS   c   
 C RC  (18)
Vbatt = C1T X (11)
 
 −1 
X ( s ) = sI − A1 C 1 V control ( s ) (19)
where, x = state variable vector, IL is inductor current and Vc
is capacitor voltage. It consists of inductor and capacitor to
where, I = identity matrix. The small-signal expression for
represents current and voltage .It can be re-written in
output voltage is given by
equation 12.
 
 diL   rc + rL 1  v batt ( s ) = C x( s )
 dt  − L −  i   1 
L  L + (20)
 dv  =  1 v
1   vc   2 LN  s
 c  −   0  
 dt   C RC  s
 1+
i  s
= rc 1  L 
v batt ( s )
Vbatt 
= Gvf
v c  s s2
(12) v control ( s ) 1+ +
 L / C1  Lc / C1 (21)
The state vector X dot is expressed in state variables such
as A2 and B2. The time interval of the primary switches is ON s
are expressed in the equation 13.  1+
v batt ( s ) s

= Agvf
• s s2
X = A2 X + B2VS  v s (s) 1+ +
 L / C1  Lc / C1 (22)
Vbatt = C2T X (13)
where, the corner frequencies are identical to those in the
which are re-written in the equation 14. control-to-output transfer function.

126
3.6 ANFIS ANFIS control mode. During first and third mode the
inverter acts like a boost converter and during second mode it
acts like buck rectifier.

4. SIMULATION AND RESULTS

Output waveforms of the entire mode are as shown in the

figure 8. During the load is in ON (manual) position the
control breakers (U1&U2), U5&U6 are turn ON by speed goat
and wind power is supply to charging the battery and supply
to load or else load is in OFF (manual) the control breakers
Figure 7. ANFIS Interface (U7 U8 U9) turn ON and supply the wind power to grid and
remaining control breakers remain in OFF condition. The
ANFIS controller is connected to boost inverter is as other two conditions are wind speed (v) is >= 5m/s and state
shown in figure 7. This Set up operation is bidirectional and of charge is in between (40-80) %.
operates in three modes such as grid-inverter ANFIS control
mode, grid- uncontrolled rectifier mode and stand alone

127
 Figure 8. Simulation output

In this mode the battery is charging and also used to load, with this wind speed the output of wind profiles values of
then this mode is called wind sourced battery-load mode. The rotor torque is -0.5 N-m and rotor speed is 160 rad/sec. The
carrier and reference signal is compared and the output signal output of PMSG is 24V. This 24V AC supply is provided to
of PWM controlled by the converter switches. The carrier converters circuit, and the final 24V DC supply is given to
and reference signal are compared, and PWM signal is charge the battery and also load. The input is getting from
generated given to the converter switches. The unidirectional PMSG the variable 24V AC supply is initially converted to
converter output value is now obtained as 24 V, and the DC by three phase uncontrolled bridge rectifier.
bidirectional converter operates in boost mode and the The output of bridge rectifier is 20V DC because the
getting input voltage from UDC is step-up to 48 V, and the rectifier efficiency is only 81.2%. The variable 20V DC
current value is 1A. Then the filtered pure DC is given to the voltage is converted into fixed 24V DC by using
inverter then the inverter converts DC-AC and also step-up to unidirectional boost converter, and the converter current is
98V. Finally, the electrical AC output is applied to two types almost 2A. This 24V constant DC is supplied to the battery
of loads. ANFIS controller is used to control the inverter for charging and also supply to BDC. The bi-directional
switching and finally LC filter is used to filter the output converter operates in boost mode, and the input voltage of
signal Final inverter output is given to both linear and non- BDC 24V is step-up to 48V. Then the filtered pure DC is
linear loads. The Total Harmonics Distortions (THD) of given to the inverter. Finally, the electrical AC output is
linear load is 2.57% and THD value of non-linear load is applied to Load. In the other side battery also charging from
4.35%. UDC constant output 24V DC.
The input of the wind energy conversion system is 8 m/s.

128
5. HARDWARE AND RESULTS it supplied to charge the battery and also provided to BDC.
The BDC output 47.6V DC is step-up and converts to AC
Hardware implementation is shown in the figure 9. The using an inverter. The inverter output 93.8V is given to the
output waveform of modes wind sourced battery-load mode load.
is as shown in figure 10 the output is taken when the wind The output is taken at when the wind speed is at 7m/s.
speed is at 5.5m/s. During the load is in ON (manual) When compared to wind speed at 5.5 m/s there are some
position the control breakers (U1&U2), U5&U6 are turn ON parameters changed. The BDC again boost up the input 24V
by speed goat and wind power is supply to charging the constant DC into 48V DC, and finally, an inverter converts
battery and supply to load or else load is in OFF (manual) the 48V DC into AC then the value nearly doubled and given to
control breakers (U7 U8 U9) turn ON and supply the wind load. The source is provided to battery and load so, this mode
power to grid and remaining control breakers remain in OFF of operation is called wind sourced battery-load mode. Real
condition. The other two conditions are wind speed (v) is >= time output values from PMSG are 15V AC, its rectified into
5m/s and state of charge is in between (40-80) %. 11.9V DC and supplied to the UDC. The UDC maintain
constant 24V DC and is provided to charge the battery and
also supplied to BDC. The BDC output 48 DC is step-up and
converts to AC using an inverter.
The inverter output 98V is as given to the load.The
experimental platform of overall system consists of a boost
converter and bi-directional converter with PMSG. The
performance of the unidirectional boost and bi-directional
converters are controlled by DSPIC30F4011 controller with
PID control structure. Inverter is controlled by ANFIS
controller.

Table 2. Output results

Figure 9. Hardware Implementation
Simulation results Hardware results
Conversion Non- Non-
Linear Linear
stages Linear Linear
load load
load load
PMSG Output 22 22 15 16
Rectifier
20 20 11.9 12
output
UDC Output
24 23.9 23.9 23.8
(V)
UDC Output
4 4 4.2 4.11
(I)
Battery Output
4 4 4.1 4.1
(V)
Battery Output
24 24 23.9 23.6
(V)
BDC Output
48 48 47.3 46.8
(V)
BDC Output
25 2.45 2.3 2.41
(I)
Inverter
2 2.1 1.7 1.9
Output (I)
Inverter
98 97.3 93.8 92.7
Output (V)
THD (%) 2.57 4.35 7.65 9.35

The hardware result values verified with simulation results.

Its outputs are almost equal to the simulation result. The
system has a robust performance under all the modes. The
THD values are obtained in simulation is below 5% and
hardware results are below 10%.
Figure 10. Hardware output result

The BDC again boost up the input 24V constant DC 6. CONCLUSION

voltage into 48V DC voltage, and finally, an inverter
converts 48V DC voltage into 98V AC voltage the value is In this paper, bidirectional converter operation plays an
nearly doubled and given to load. The source is given to both important role for mode changing operation and its five
battery and load so, this mode of operation is called wind modes of operation were analysed. Wind energy is the main
sourced battery-load mode. Real time output values from input sources from renewable energy, based on this source
PMSG are 15V AC, its rectified into 11.9V DC and supplied level the operations are divided into five modes and also
to the UDC. The UDC maintains constant 24V DC, and again these modes based on state of charge in battery. The

129
 electrical energy conversion system consists of a Ind. Appl 32(2): 293–300.
unidirectional boost converter and bidirectional converter http://doi.org/10.1109/28.491477
operates four modes in boost operation and one mode in a [6] Ganesh G, Vijay Kumar G, VijayBabu,G.SrinivasaRao
buck operation. Simulink model of whole system including AR, Tagore YR. (2015). Performance analysis and
unidirectional DC-DC boost converter, bidirectional DC-DC MPPT control of a standalone hybrid power generation
converters were developed for wind energy electrical system system. Journal of Electrical Engineering. 15(1): 334-
and simulation results were obtained in 343.
MATLAB/SIMULINK. Real time implementation of whole [7] Norrga S. (2006). Experimental study of a soft-switched
system including speedgoat real time target machine, isolated bidirectional AC-DC converter without
bidirectional DC-DC converters controlled by auxiliary circuit. IEEE Trans. Power Electron 21(6):
DSPIC30F4011 were developed for wind energy electrical 1580–1587. http://doi.org/10.1109/TPEL.2006.882969
system and real-time results were obtained. A variety of [8] Norrga S, Meier S, Ostlund S. (2008). A three-phase
operating conditions from different inputs were analysed. soft-switched isolated ac/dc converter without auxiliary
The system has a robust performance under mode changing circuit. IEEE Trans. Ind. Appl 44(3): 836–844.
while input wind speed changes. The hardware results of the http://doi.org/10.1109/TIA.2008.921430
proposed model were verified with simulation results. The [9] Yang XD, Duan WY, Wang JQ, Hu GW.(2017).
mode changing operation is effectively done in both Analysis of the fault property of the sensors in H-Bridge
simulation and real-time platforms. converter. Modelling, Measurement and Control A
90(2): 183-195.
http://dx.doi.org/10.18280/mmc_a.900205
REFERENCES [10] Vijay Babu AR, Rajyalakshmi V, Suresh K. (2017).
Renewable energy integrated high gain DC-DC
[1] Kolar JW, Drofenik U, Zach FC. (1999). Vienna converter with multilevel inverter for water pumping.
rectifier II—A novel single-stage high-frequency Journal of Advanced Research in Dynamical and
isolated three-phase PWM rectifier system. IEEE Trans. Control Systems 9(1): 173-190.
Ind. Electron 46(4): 674–691. [11] Chen SJ. (2015). Bidirectional three-phase high-
http://doi.org/10.1109/41.778214 frequency ac link dc-ac converter used for energy
[2] De D, Ramanarayanan V. (2010). A dc-to-three-phase- storage. IET Power Electron 8(12): 2529–2536.
ac high-frequency link converter with compensation for http://doi.org/10.1049/iet-pel.2014.0840
nonlinear distortion. IEEE Trans. Ind. Electron 57(11): [12] Kusiak A, Zhang Z, Li MY. (2010). Optimization of
3669–3677. http://doi.org/10.1109/TIE.2010.2040566 wind turbine performance with data-driven models.
[3] Weerasinghe DSB, Madawala UK, Thrimawithana DJ, IEEE Trans. Sustain. Energy 1(2): 66–76.
Vilathgamuwa DM. (2013). A three-phase to single- http://doi.org/10.1109/TSTE.2010.2046919
phase matrix converter based bi-directional IPT system [13] Suresh K, Arulmozhiyal R. (2016). Design and
for charging electric vehicles. In Proc. 2013 IEEE implementation of bi-directional dc-dc converter for
ECCE Asia Downunder 1240–1245. wind energy system. Circuits and Systems 7: 3705-3722.
http://doi.org/10.1109/ECCE-Asia.2013.6579267 http://doi.org/ 10.4236/cs.2016.711311
[4] Singh AK, Das P, Panda SK. (2015). A novel matrix [14] Jain M, Daniele M, Jain PK. (2000). A bidirectional dc–
based isolated three phase ac-dc converter with reduced dc converter topology for low power application. IEEE
switching losses. In Proc. 2015 IEEE Appl. Power Trans. Power Electron 15(4): 595–606.
Electron. Conf. Expo 1875–1880. http://doi.org/10.1109/63.849029
http://doi.org/10.1109/APEC.2015.7104602 [15] Xu CY, Xie CJ, Jiang F, Zhao JY. (2016). Design and
[5] Matsui M, Nagai M, Mochizuki M, Nabae A. (1996). implementation of the power battery management
High-frequency link dc/ac converter with suppressed system of photovoltaic power generation based on Bi-
voltage clamp circuits-naturally commutated phase directional DC-DC equalization control. Modelling,
angle control with self-turn-off devices. IEEE Trans. Measurement and Control A 89(1): 156-172.

130