Journal of Power and Energy Engineering, 2014, 2, 151-160
Published Online April 2014 in SciRes.
How to cite this paper: Ronilaya, F., Miyauchi, H. and Kurniawan, A. (2014) PID-Type Fuzzy Controller for Grid-Supporting
Inverter of Battery in Embedded Small Variable Speed Wind Turbine. Journal of Power and Energy Engineering, 2, 151-160.
PID-Type Fuzzy Controller for
Grid-Supporting Inverter of Battery in
Embedded Small Variable Speed
Wind Turbine
Ferdian Ronilaya1,2, Hajime Miyauchi1, Adi Kurniawan3
1Department of Frontier Technology for Energy and Devices, Kumamoto University, Kumamoto-shi, Japan
2Department of Electrical Power Engineering, The State Polytechnic of Malang, Malang, Indonesia
3Institut Teknologi Sepuluh Nopember, Surabaya, Indonesia
Email: ferdian@st.cs.ku
Received December 2013
Frequency and voltage of embedded variable speed wind turbine (VSWT) driving a permanent
magnet synchronous generator (PMSG) is strongly affected by wind speed fluctuations. In practice,
power imbalance between supply and demand is also common, especially when VSWT-PMSG is
connected to a weak micro grid (MG). If load demand fluctuations become high, isolated MG may
be unable to stabilize the frequency and voltage so that battery storage needs to be installed into
the MG to adjust energy supply and demand. To allow flexible control of active and reactive power
flow from/to battery storage, grid-supporting inverters are used. For a system that contains highly
nonlinear components, the use of conventional linear proportional-integral-derivative (PID) con-
trollers may cause system performance deterioration. Additionally, these controllers show slow,
oscillating responses, and complex equations are required to obtain optimum responses in other
controllers. To cope with these limitations, this paper proposes PID-type fuzzy controller (PIDfc)
design to control grid-supporting inverter of battery. To ensure safe battery operating limits, we
also propose a new controller scheme called intelligent battery protection (IBP). This IBP is inte-
grated into PIDfc. Several simulation tests are performed to verify the scheme’s effectiveness. The
results show that the proposed PIDfc controller exhibits improved performance and acceptable
responses, and can be used instead of conventional controllers.
Battery; PID-Type Fuzzy Controller; Inverter; Permanent Magnet Synchronous Generator (PMSG);
Variable Speed Wind Turbine (VSWT)
1. Introduction
VSWT-PMSG is a promising wind turbine system due to its ability to extract maximum power from fluctuating
F. Ronilaya et al.
wind resources. PMSG can operate at low wind speeds so that a gearbox is no longer needed. This feature may
reduce the cost and is suitable for small-scale applications. The VSWT-PMSG power varies with the wind speed,
so this can lead to poor performance of frequency and voltage of MG [1]. To obtain better performance, both in
terms of technical and economical, VSWT-PMSG needs some reliable control strategies. This is usually done
with the help of power electronics devices, for example an inverter. In most cases in the distributed generations
(DG), the inverter is connected in parallel with other inverters so that the inverter controller should be designed
Many researchers have considered the control strategies of parallel-connected inverters [2-5]. Most control
strategies are based on the droop method. Use of this method for inverter control was proposed by Chandorkar et
al. [3]. Their goal was to have accurate load sharing among the inverters. Parallel inverter operation was also
discussed by Kawabata et al. [4]. However, this study only focused on safe inverter operation. The dynamic
voltage and frequency responses of the output inverters were not discussed. Control strategies for nonlinear load
sharing by regulating the inverter output voltage bandwidth have been proposed [6]. Novel control structures
that include virtual resistors and reactors to adjust the output impedance were developed in [7,8], respectively.
However, these control strategies have two limitations. First, there is a trade-off between voltage regulation and
load sharing. Second, the proposed controllers have slower dynamic responses. To overcome these limitations, a
new control strategy was proposed by Guerrero et al. [9]. The scheme uses a wireless controller by including a
supplemental transient droop characteristic in the conventional static droop method. However, the main problem
of this scheme lies in the measurement noise and proportional load sharing accuracy. The first problem was
overcome using H- controllers [10-12]. But, the main drawback of this controller for real implementation is
that the H- technique requires high levels of mathematical understanding [13]. The second limitation of the
droop controller proposed in [9] was overcome by Qing-Chang [14]. However, for a system that contains highly
nonlinear components, the use of linear conventional PID controllers may weaken the system performance.
Rather than using high levels of mathematical understanding, this paper proposes a simple method based on
PIDfc to define the relationship between input information and output action. Furthermore, fuzzy-based control-
ler is suitable for highly non-linear components. Based on the literature review carried out by Pandey et al. [15],
however, little research has been conducted in this area. Similar studies were discussed in [16]. However,
grid-connected inverter control, which is the subject of this work, has not been discussed. Additionally, voltage
control is not taken into account.
As flooded deep-cycle lead-acid batteries are used, safe battery operation should be considered in designing
controller of the grid-supporting inverter. The battery should not be overcharged or overdischarged because it
may shorten the battery life and damage the battery. To accommodate these limitations, we propose a new con-
troller strategy scheme called intelligent battery protection (IBP). The SOC (state of charge) is used as input
signal in the IBP scheme. The output of IBP acts as intelligent protection signal which is then sent to PIDfc as
additional input.
2. System Model
Figure 1 shows the configuration of isolated embedded VSWT system. It is assumed that this prototype is used
for non-urban electrification and is located on coastal area with flat terrain. MG/diesel generator behaves as the
main power supply and has 31 kVA capacity. All of the energy sources are used to meet dynamic household and
agricultural electricity demand.
2.1. VSWT-PMSG Characteristics and Modeling
The wind turbine model used in this simulation is manufactured by Fortis Wind Energy. It has 7-meter span
blades. The company claims that this turbine can generate 10 kW of power at 13 m/s of wind speed [17]. The
v-P (wind speed-power) characteristic of this turbine is shown in Figure 2. By using polynomial regression,
such a characteristic can be modeled by (1).
1.124.361.841.5 9.5
Pvvvv=−+−+ −
where v is wind speed in m/s and Pw is wind turbine power. The tower on which the wind turbine is mounted has
30 m height and is located on 5 m from MG bus.
F. Ronilaya et al.
Figure 1. Isolated MG with embedded small VSWT-PMSG
and batteries.
Figure 2. Alize Fortis wind turbine characteristics.
2.2. Lead-Acid Battery Model
The non-linear dynamic model of lead acid battery used is adopted from [18,19] and is available in Matlab/
Simpowersystem library. Table 1 lists the parameters used in the simulation.
One of important parameters of lead acid battery is SOC. SOC represents available energy capacity of the
lead acid battery. In this paper, the SOC value is limited between 30% - 90% and is used as one of primary input
signals of IBP.
3. Inverter Control
The output impedance of the inverter is usually very inductive, Z∠θ = X90˚, therefore, the flow of real and
reactive power from grid-connected inverters can be expressed by (3) and (4) [9].
cosEV V
P: active power
Q: reactive power
X: output reactance of inverter
E: output voltage of inverter
V: MG bus voltage
: phase angle between E and V
Based on (2) and (3), the active and reactive power is predominantly be affected by the value of phase angle,
F. Ronilaya et al.
and inverter output voltage, E, respectively. In other words, by adjusting P and Q, the frequency and voltage
can be determined.
3.1. VSWT-PMSG Inverter Controller
Figure 3 shows a schematic diagram of the proposed inverter controller of the VSWT-PMSG. The use of PMSG
can provide some major simplifications because it does not require excitation current so as to save costs and im-
prove efficiency [20]. Voltage and frequency output of VSWT-PMSG fluctuate with wind speed so that the
output of VSWT-PMSG need to be converted into DC voltage by using rectifier and is stabilized by synchron-
ous buck converter. The PMSG and rectifier are completely uncontrollable. Thus, active and reactive power
transfer can only be realized by controlling the inverter.
For a given MG bus voltage, based on (2) and (3), the real and reactive power can be controlled by regulating
the angle δ and amplitude of inverter output voltage, E. The regulation can be realized by varying the sinusoidal
reference magnitude, U and phase angle,
w, with respect to the MG voltage. The frequency of sinusoidal mod-
ulating reference signal must equal to that of MG. Therefore, signal from PLL (
t) is used as reference. Limiter
Lv is used to limit the amplitude of reference voltage U not higher than that of the carrier voltage. VSWT-PMSG
is operated under unity power factor so that reactive power reference is set to 0.
The rated wind speed for the VSWT-PMSG is 13 m/s. If the wind speed is higher than 13 m/s, VSWT-PMSG
power is limited at the rated power value. A limiter Lf is used to realize this strategy. Below the rated wind
speed, the VSWT-PMSG follows the power output conditions as it is defined in (1).
3.2. Battery Inverter Controller
PIDfc is first introduced by Wu Zhi and Mizumoto in 1996 [21]. The interesting feature of this controller
scheme is that it is quite simple and has reduced overshoot if the gain factors are selected properly. The imple-
mentation of PIDfc scheme in the inverter controller is shown in Figure 4. Two PIDfc blocks are used to regu-
late frequency and voltage of the MG. The PIDfc scheme is designed such that the deviation of the frequency
Table 1. The battery parameters.
Parameters Value
Internal resistance 0.008 ohm
Fully charge voltage 261.3158 V
Exponential voltage zone 244.34 V
Exponential capacity zone 1 Ah
Maximum capacity 312.5 Ah
Nominal discharge current 60 A
Capacity at nominal voltage 93.0833 Ah
Figure 3. VSWT -PMSG inverter controller.
F. Ronilaya et al.
Figure 4. Battery inverter controller.
and voltage of the MG does not exceed 2% and 5% respectively [9]. Figure 5 shows input and output member-
ship functions (MFs) of FLCf and FLCv. The rule bases table for both FLCf and FLCv is listed in Table 2. IBP
is realized by using two hysteresis relays (Relay 1 and 2). Relay 1 is used to protect the battery from overcharge
condition. Relay 1 will be ON only if the SOC of the battery is equal to 90%. It remains ON until the SOC drops
to 80%. When the SOC is equal to 90% and the battery current is negative, then the battery is disconnected from
the MG. Meanwhile, Relay 2 is intended to protect the battery from over-discharge conditions. If the SOC is
equal to 30%, Relay 2 will be ON. With this condition, the controller will force the battery to absorb electrical
power from MG whose value is proportional to the available power from diesel generators. Available power
from diesel generator can be expressed by (4).
._avd ndd
PS QP= −−
Sd_n is nominal power capacity of diesel generator (31 kVA). Meanwhile, Qd and Pd are measured reactive and
active power of diesel generator respectively. To prevent excessive charging current, limiter block is used. It
should be noted that Relay 2 will remain ON until the SOC of the battery rises to 50%.
4. Simulation Results
Several simulation tests are performed to verify the performance of proposed PIDfc controller. Then, the simu-
lation results of PIDfc are compared with that of conventional PID controller. To analyze the significance of
battery in regulating frequency and voltage of MG, the simulation results are also compared with that of the
system in which the battery is not used. Load changes dynamically with constant power factor, i.e. 0.93. To ob-
tain best performance, parameters of PIDfc and conventional PID controllers for all simulation cases are ob-
tained by using trial-and-error method. The parameters of PIDfc controller are listed in Table 3. Matlab/Sim-
power system software is used to simulate the model.
Active power profiles of diesel generator, VSWT-PMSG, battery and load are shown in Figure 5. Meanwhile
F. Ronilaya et al.
Figure 5. MFs of FLCf and FLCv.
Table 2. The Rules Base for FLCf and FLCv.
F. Ronilaya et al.
Table 3. The PIDfc parameters.
Parameters Value
G1 1
G2 2
G3 5 x 10-4
G4 3 x 10-6
G5 1
G6 1
G7 5 x 10-4
G8 2 x 10-5
Kb 2 x 10-5
Figure 5. Active power profiles.
the reactive power and battery current responses are shown in Figures 6 and 7 respec tively. The oscillating res-
ponses are the simulation results when the conventional PID controller is used. On the other hand, smoother
responses are the results of PIDfc. These results confirm that PIDfc exhibits improved performance than that of
conventional PID controller. The sudden load changes are compensated by diesel generator because it can ab-
sorb or deliver power faster than the battery. The negative value of battery current indicates that the battery is
being charged and vice versa. We can see from Figure 7 that the maximum current of the battery is equal to 60
A which confirms to the nominal discharge current value listed in Table 1.
The active power delivered by VSWT-PMSG is proportional to the v-P characteristics as defined in Figure 2
and (1). In other hand, the reactive power delivered from VSWT-PMSG is equal to 0. It indicates the VSWT-
PMSG is operating at unity power factor as aforementioned design in Section 3.1.
Figures 8 and 9 show the frequency and voltage profiles for both using PIDfc and conventional PID control-
ler. To verify the effectiveness of PIDfc controller quantitatively, the following performance index is used:
ISEe dt=
ISE is integral squared error. Meanwhile, e and T are frequency/voltage error and simulation time respectively.
Table 4 summarizes the calculated ISE of MG frequency and voltage for different conditions.
This table verifies that PIDfc shows better performance than conventional PID controller, as shown in the
lower value of ISE for both voltage and frequency. Moreover, this table signifies that the battery plays an im-
portant role particularly when power unbalance between supply and demand occurs in isolated MG system. Bat-
tery may also provide a way to operate diesel generator in more efficient manner. Any extra power can be used
F. Ronilaya et al.
Figure 6. Reactive power profiles.
Figure 7. Battery current profiles.
Table 4. The calculated ISE.
Parameters ISE
Frequency (PIDfc based) 0.021617
Frequency (PID based) 0.037741
Frequency (wthout battery support) 21.10304
Voltage (PIDfc based) 5.454813
Voltage (PID based) 14.34414
Voltage (without battery support) 1018.58
to charge the battery during low load time and energy can be extracted during peak times. Although the initial
cost of the battery installation is expensive, fuel savings are much greater than the initial extra cost for remote
area application.
5. Conclusion
In this paper, a PIDfc controller for battery grid-supporting inverter in embedded VSWT-PMSG has been pre-
sented. In order to protect the battery from overcharge and overdischarge condition, IBP is introduced and inte-
grated with PIDfc controller. Simulation results show that PIDfc has lower ISE than that of conventional PID
controller. It indicates that PIDfc exhibits significantly improved performance and can be used instead of con-
ventional controllers.
F. Ronilaya et al.
Figure 8. Frequency profiles.
Figure 9. Voltage profiles.
The authors gratefully acknowledge the contributions of Directorate General of Higher Education of Indonesia
for the financial supports of this work.
[1] Muyeen, S. M., Takahashi, R., Murata, T. and Tamura, J. (2009) Integration of an Energy Capacitor System with a Va-
riable-Speed Wind Generator. IEEE Transactions on Energy Conversion, 24, 740-749.
[2] Barsali, S., Ceraolo, M., Pelacchi, P. and Poli, D. (2002) Control Techniques of Dispersed Generators to Improve the
Continuity of Electricity Supply. IEEE Power Engineering Society Winter Meeting, 789-794.
[3] Chandorkar, M.C., Divan, D.M. and Adapa, R. (1993) Control of Parallel Connected Inverters in Standalone AC
Supply Systems. IEEE Transactions on Industry Applications, 29, 136-143.
[4] Kawabata, T. and Higashino, S. (1988) Parallel Operation of Voltage Source Inverters. IEEE Transactions on Industry
Applications, 24, 281-287.
[5] Matthias, H. and Helmut, S. (2002) Control of a Three Phase Inverter Feeding an Unbalanced Load and Operating in
Paralel with Other Power Sources. EPE-PEMC, Dubrovnik & Cavtat, 1-10.
[6] Tuladhar, A., Jin, K., Unger, T. and Mauch, K. (1997) Parallel Operation of Single Phase Inverter Modules with No
Control Interconnections. Twelfth Annual APEC, 23-27 Feb 1997, 94-100.
[7] Chiang, S. J. , Yen, C.Y. and Chang, K. T. (2001) A Multimodule Parallelable Series-Connected PWM Voltage Regula-
tor. IEEE Transactions on Industrial Electronics, 48, 506-516.
F. Ronilaya et al.
[8] Engler, A. (2000) Control of Parallel Operating Battery Inverters. PV Hybrid Power System.
[9] Guerrero, J.M. , Garcia de Vicuna, L., Matas, J., Castilla, M. and Miret, J. (2004) A Wireless Controller to Enhance
Dynamic Performance of Parallel Inverters in Distributed Generation Systems. IEEE Transactions on Power Electron-
ics, 19, 1205-1213.
[10] Goya, T., Omine, E., Kinjyo, Y., Senjyu, T., Yona, A., Urasaki, N. and Funabashi, T. (2011) Frequency Control in
Isolated Island by Using Parallel Operated Battery Systems Applying H-Inf; Control Theory Based on Droop Charac-
teristics. IET Renewable Power Generation, 5, 160-166.
[11] Patra, S., Sen, S. and Ray , G. (2007) Design of Robust Load Frequency Controller: H Loop Shaping Approach. Elec-
tric Power Components and Systems, 35, 769-783.
[12] Singh, V.P., Mohanty, S.R., Kishor, N. and Ray, P.K. (2013) Robust H-Infinity Load Frequency Control in Hybrid
Distributed Generation System. International Journal of Electrical Power & Energy Systems, 46, 294-305,.
[13] Cubillos, X.C.M. and Souza, L.C. G.D. (2009) Using H-Infinity Control Method in Attitude Control System of Rigid-
Flexible Satellite. Mathematical Problems in Engineering, 2009, 1-9.
[14] Qing-Chang, Z. (2013) Robust Droop Controller for Accurate Proportional Load Sharing among Inverters Operated in
Parallel. IEEE Transactions on Industrial Electronics, 60, 1281-1290.
[15] Pandey, S.K., Mohanty, S. R. and Ki shor, N. (2013) A Literature Survey on Load-Frequency Control for Conventional
and Distribution Generation Power Systems. Renewable and Sustainable Energy Reviews, 25, 318-334.
[16] Yeşil, E., Güzelkaya, M. and Eksin, I. (2004) Self Tuning Fuzzy PID Type Load and Frequency Controller. Energy
Conversion and Management, 45, 377-390.
[17] Fortis Wind Energy (2013) Alize Wind Turbine.
[18] Matlab/Simpowersystem (2013) Battery-Implement Generic Battery Model.
[19] Tremblay , O. and Dessaint, L.A. (2009) Experimental Validation of a Battery Dynamic Model for EV Application.
World Electric Vehicle Journal, 3, 1-10.
[20] Chen, Z. and Spooner, E. (1998) Grid Interface Options for Varia ble -Speed , Permanent-Magnet Generators. IEE Pro-
ceedings: Electric Power Applications, 145, 273-283.
[21] Wu, Z.Q. and Mizumoto, M. (1996) PID Type Fuzzy Controller and Parameters Adaptive Method. Fuzzy Sets and
Systems, 78, 23-35.