- Email: [email protected]
Fuzzy Logic Based MPPT for Permanent Magnet Synchronous Generator in wind Energy Conversion System
4th International Conference on Advances in Control and Optimization of Dynamical Systems 4th International Conference on Advances in Control and 4th International Conference on in and 4th International Conference on Advances Advances in Control Control February 1-5, of 2016. NIT Tiruchirappalli, India Optimization Dynamical Systems Available onlineand at www.sciencedirect.com Optimization of Dynamical Systems Optimization Dynamical Systems February 1-5, of 2016. NIT Tiruchirappalli, India February 1-5, 2016. NIT Tiruchirappalli, India February 1-5, 2016. NIT Tiruchirappalli, India
ScienceDirect
49-1 (2016) 462–467 FUZZY LOGIC BASEDIFAC-PapersOnLine MPPT FOR PERMANENT MAGNET SYNCHRONOUS FUZZY LOGIC BASED MPPT FOR PERMANENT MAGNET SYNCHRONOUS FUZZY BASED FOR PERMANENT MAGNET SYNCHRONOUS GENERATOR IN WIND ENERGY CONVERSION FUZZY LOGIC LOGIC BASED MPPT MPPT FOR PERMANENT MAGNETSYSTEM SYNCHRONOUS GENERATOR IN WIND ENERGY CONVERSION SYSTEM GENERATOR IN WIND ENERGY CONVERSION SYSTEM GENERATOR IN Ramji WIND ENERGY CONVERSION SYSTEM Tiwari*, Ramesh Babu. N**
Ramji Tiwari*, Tiwari*, Ramesh Babu. Babu. N** Ramji Ramji Tiwari*, Ramesh Ramesh Babu. N** N** School of Electrical Engineering, VIT University, Vellore, India. School of of Electrical Engineering, Engineering, VIT University, University, Vellore, India. India. School *[email protected],**[email protected] School of Electrical Electrical Engineering, VIT VIT University, Vellore, Vellore, India. *[email protected],**[email protected] *[email protected],**[email protected] *[email protected],**[email protected] Abstract: In this paper, a comparative analysis of different control methods to extract the maximum Abstract: InPermanent this paper, paper,Magnet a comparative comparative analysis of different different control methods to extract extract the maximum maximum power fromIn Synchronous Generator (PMSG) based methods Wind Energy Conversion System Abstract: this analysis of control to the Abstract: InPermanent this paper,Magnet aa comparative analysis of different control methods to extract the maximum power from Synchronous Generator (PMSG) based Wind Energy Conversion Systema (WECS) under differentMagnet wind speed condition is presented. Thebased WECS consists ofConversion a wind turbine, power from Permanent Synchronous Generator (PMSG) Wind Energy System power from Permanent Magnet Synchronous Generator (PMSG) based Wind EnergyofConversion Systema (WECS) under different wind speed condition is presented. The WECS consists a wind turbine, PMSG and a DC/DC converter whichcondition is connected to a DC load. The Maximum Power Pointturbine, Trackinga (WECS) under different wind is The WECS consists of (WECS) under different wind speed speed is presented. presented. The The WECS consistsPower of aa wind wind turbine, a PMSG and and a DC/DC DC/DC converter whichcondition is connected connected to aa DC DC load. load. Maximum Point Tracking (MPPT) control technique compared here are Proportional Integral (PI) control, Power PerturbPoint and Tracking Observe PMSG a converter which is to The Maximum PMSG and a DC/DC converter which is connected to a DC load. The Maximum Power Point Tracking (MPPT) control and technique compared here are Proportional Proportional Integral (PI) considered control, Perturb and Observe (P&O) method Fuzzycompared Logic Controller (FLC). The Integral parameters for analysing the (MPPT) control technique here (PI) control, and (MPPT) control and technique compared here are are Proportional Integral (PI) considered control, Perturb Perturb and Observe Observe (P&O) method Fuzzy Logic Controller (FLC). The parameters for analysing the efficiency of the MPPT controller is the output DC voltage and power across the load. The steady state (P&O) method and Fuzzy Logic (FLC). The parameters considered for analysing the (P&O) method and Fuzzy Logicis Controller Controller (FLC). The and parameters considered forThe analysing the efficiency of the MPPT controller the output DC voltage power across the load. steady state voltage and the dynamic response of the system under different wind speed is considered to justify the efficiency of the MPPT is the output DC voltage and across load. state efficiency ofthe thedynamic MPPT controller controller is the output DC voltage and power power acrossisthe the load. The Thetosteady steady state voltage and response of the system under different wind speed considered justify the The system is designed and configured in MATLAB/SIMULINK overall efficiency of the controllers. voltage and response the under different wind speed to voltage and the the dynamic dynamic response of of The the system system under different wind speed is isinconsidered considered to justify justify the the system is designed and configured MATLAB/SIMULINK overall efficiency of the controllers. software and the results are validated. The overall of controllers. The system system is is designed designed and and configured configured in in MATLAB/SIMULINK MATLAB/SIMULINK overall efficiency efficiency of the theare controllers. software and the results validated. software and the results are validated. © 2016, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved. software and the results are validated. Keywords: Wind turbine, Permanent Magnet Synchronous Generator, Fuzzy Logic Controller, MPPT. Keywords: Wind turbine, Permanent Magnet Synchronous Generator, Fuzzy Logic Controller, MPPT. Keywords: Wind turbine, Permanent Magnet Synchronous Generator, Fuzzy Logic Controller, MPPT. Keywords: Wind turbine, Permanent Magnet Synchronous Generator, Fuzzy Logic Controller, MPPT. control is very simple to implement. PI control lacks in 1. INTRODUCTION control is very implement. PI control lacks in efficiency of the simple overall to system due to its arbitrary selection control is very simple to implement. PI control lacks 1. INTRODUCTION control is of very simple to implement. PIarbitrary control selection lacks in in 1. INTRODUCTION efficiency the overall system due to its of parameters. In order to overcome this problem FLC is 1. INTRODUCTION In recent years, the use of renewable energy resource is efficiency of system due its selection efficiency of the theInoverall overall system due to to this its arbitrary arbitrary selection of parameters. order to overcome problem FLC is employed to extract the to maximum power from wind. FLC In recent years, the use of renewable energy resource is increased due to increasing demand of power and depletion of parameters. In order overcome this problem FLC is In recent years, the use of renewable energy resource is of parameters. In order to overcome this from problem FLC is In recent years, the use of renewable energy resource is employed to extract the maximum power wind. FLC can track to theextract non linearity of thepower systemfrom andwind. givesFLC the increased due to increasing demand of power and depletion of fossil fuel such as coal for electricity generation. employed the maximum increased due to increasing demand of power and depletion employed to extract the maximum power from wind. FLC increased due to such increasing demand ofelectricity power andgeneration. depletion can track the nonforlinearity of thewind. system gives maximum output the available Theand input for the of fossil fuel coal for Moreover, the issue ofas climate change are causing concern can track the of system and gives the of fossil fuel such as coal for electricity generation. can track output the non nonforlinearity linearity of the thewind. system and gives the of fossil fuel such as coal for electricity generation. maximum the available The input for the FLC is the DC voltage and current across the load, whereas Moreover, the issue of climate change are causing concern and hence many regulations are proposed to reduce the maximum output for the available wind. The input for the Moreover, the issue of climate change are causing concern maximum output for the available wind. The input for the Moreover, the issue of climate change are causing concern FLC is the DC voltage and current across the load, whereas the output is the duty cycle for the DC/DC Converter. and hence many regulations are proposed to reduce the ) emission (Krishna et al., 2015). Carbon dioxide (CO FLC is the DC voltage and current across the load, whereas 2 and hence many regulations are proposed to reduce the FLC is the DC voltage and current across the load, whereas and hence many regulations are proposed to reduce the the output is the duty cycle for the DC/DC Converter. )) emission et al., 2015). Carbon dioxide (CO Among the renewable solar (Krishna energy and wind energy 2energy, the the cycle for (Krishna et al., 2015). Carbon dioxide (CO output isconverter the duty duty is cycle for the the DC/DC DC/DC Converter. Theoutput buck is employed here to Converter. interface the wind ) emission emission (Krishna et wind al., energy 2015). the Carbon dioxide (CO22energy, Among the renewable solar energy and is moretheutilised because ofsolar itsenergy abundant availability Among renewable energy, and wind energy The buck converter is employed here to interface the generator to the DC Load. The output of the PMSG iswind AC Among the renewable energy, solar energy and wind energy The is employed here interface the is more utilised because its abundant everywhere. The Wind Energyof is gaining interest availability because of generator The buck bucktoconverter converter is employed here to to interface theiswind wind is more utilised because of its abundant availability the DC Load. The output of the PMSG AC which is converted in to DC using diode controlled rectifier is more utilised because of its abundant availability generator to The output of the is everywhere. The Wind Energy gaining interest because of technology enhancement andis significant power cost generator to the the DC DCinLoad. Load. The output of controlled the PMSG PMSGrectifier is AC AC everywhere. The Wind Energy is gaining interest because of which is converted to DC using diode in order to eliminate the ripple present in the AC component everywhere. The Wind Energy is gaining interest because of which is in to DC diode controlled rectifier technology enhancement significant cost reduction. More effective and control strategiespower are under which is toconverted converted inthe to ripple DC using using diode controlled rectifier technology enhancement and significant power cost in order eliminate present in the AC component andorder smoothing capacitor is placed across the rectifier to technology enhancement and significant power cost in to the present in AC component reduction. More control are under research in order toeffective obtain reliable, coststrategies effective and quality in order to eliminate eliminate the ripple ripple presentacross in the the the AC rectifier component reduction. More effective control strategies are under and smoothing capacitor is placed to minimise the ripple due to non linearity. When the wind reduction. More effective control strategies are under and smoothing capacitor is placed across the rectifier to research in order obtain cost effective power from the to wind. Asreliable, stated in (Errami et and al., quality 2015) minimise and smoothing capacitor is non placed across When the rectifier to research in order to obtain reliable, cost effective and quality the ripple due to linearity. the wind power varies the FLC tracks the output voltage and current to research in order to obtain reliable, cost effective and quality minimise the ripple due to non linearity. When the wind power from the wind. As stated in (Errami et al., 2015) Permanent Magnet Synchronous Generator (PMSG) is most minimise the ripple due to non linearity. When the wind power from the wind. As stated in (Errami et al., 2015) power varies the FLC tracks the output voltage and current to generate an efficient duty cycle for the converter operation. power from the wind. As stated in (Errami et al., 2015) power the tracks the voltage and current Permanent Magnet Synchronous (PMSG) is most preferred wind generator due to Generator its reliability and size for generate power varies varies the FLC FLC duty trackscycle the output output voltage and operation. current to to Permanent Magnet Synchronous Generator (PMSG) is most an efficient for the converter Thus the maximum power is produced based on available Permanent Magnet Synchronous Generator (PMSG) is most generate an efficient duty cycle for the converter operation. preferred due to its reliability and size for stand alonewind windgenerator energy conversion system. generate an efficient duty cycle for the converter operation. preferred wind generator due to its reliability and size for the maximum power is produced based on available wind speed. preferred wind generator due to itssystem. reliability and size for Thus Thus the stand alone wind energy conversion Thus the maximum maximum power power is is produced produced based based on on available available stand alone wind energy conversion system. wind speed. stand alone wind energy conversion system. MPPT control algorithms can be employed in order to wind speed. wind speed. This paper analyses the maximum power point tracking MPPT control algorithms can be employed order to capture the maximum power from availablein wind by MPPT control algorithms can be employed in order to This paper analyses the maximum power point tracking control approach for stand alone WECS. The performance of MPPT control algorithms can be employed in order to This paper analyses the maximum power point capture the maximum from available maintaining the optimum power steady voltage across thewind load. by A This paper analyses the alone maximum power point tracking tracking capture the maximum power from available wind by control approach for stand WECS. The performance of the conventional PI controller, P&O controller and FLC capture the maximum power from available wind by control approach for stand alone WECS. The performance of maintaining the optimum steady across the A variety of MPPT techniques havevoltage been employed forload. Wind control approach for stand alone WECS. The performance of maintaining the optimum steady voltage across the load. A the conventional PI controller, P&O controller and FLC under variable wind speed is evaluated. The proposed control maintaining the optimum steady voltage across the load. A the conventional PI controller, P&O controller and FLC variety of MPPT techniques have been employed for Wind Energy Conversion System (WECS) in previous literature the conventional PI controller, P&O controller and FLC variety of MPPT have for Wind variable is evaluated. The proposed control strategy posseswind the speed improved capability of capturing the variety of MPPT techniques techniques have been beeninemployed employed for Wind under under variable wind speed is The proposed control Energy Conversion System previous literature such as Hill Climbing Search(WECS) (HCS) algorithm, Incremental under variable wind speed is evaluated. evaluated. Theof proposed control Energy Conversion System (WECS) in previous literature strategy posses the improved capability capturing maximum power from wind. Comparative efficiency of Energy Conversion System (WECS) in previous literature strategy posses the improved capability of capturing the the such as Hill Climbing Search (HCS) algorithm, Incremental and Conductance method (INC), Perturb and Observe (P&O) strategy posses the improved capability of capturing the such as Hill Climbing Search (HCS) algorithm, Incremental maximum power from wind. of the controllers is analysed fromComparative the outputefficiency power of such as Hill Climbing Search (HCS) algorithm, Incremental maximum power from wind. Comparative efficiency of the and Conductance method (INC), Perturb and Observe (P&O) method, Fuzzy method Logic (INC), Controller (FLC) and (P&O) many maximum power from wind. Comparative efficiency of the and Conductance Perturb and Observe controllers is analysed from the output power of the converter. and Conductance method (INC), Perturb and Observe (P&O) controllers is analysed from the output power of the method, Fuzzy Logic Controller (FLC) and many Evolutionary Algorithms. controllers is analysed from the output power of the method, Fuzzy Logic Controller (FLC) and many converter. method, Fuzzy Logic Controller (FLC) and many converter. Evolutionary Algorithms. converter. Evolutionary Algorithms. 2. MODELLING OF WIND ENERGY CONVERSION Evolutionary P&O methodAlgorithms. is well known MPPT technique due to its 2. ENERGY SYSTEM P&O method is well known MPPT technique due to its simplicity and effectiveness. But due to high non linearity in 2. MODELLING MODELLING OF OF WIND WIND ENERGY CONVERSION CONVERSION P&O method is well known MPPT technique due to its 2. MODELLING OF WIND ENERGY CONVERSION P&O method is well known MPPT technique due to its SYSTEM simplicity and effectiveness. due high non in wind speed, P&O techniqueBut fails toto track the linearity maximum SYSTEM simplicity and effectiveness. But due to high non linearity in SYSTEM simplicity andP&O effectiveness. But duetoto track high non linearity in The simulation models of wind turbine, PMSG and power wind speed, technique fails the maximum power point and hence introduces high fluctuations resulting The simulation models of wind turbine, PMSG and power wind speed, P&O technique fails to track the maximum electronics converters which comprises the whole WECS wind speed, P&O technique fails to track the maximum simulation models of turbine, PMSG and power power and hence introduces high fluctuations in low point power output (Dailii et al., 2015). Anotherresulting control The The simulation models which of wind wind turbine, the PMSG and WECS power power point and hence introduces high fluctuations resulting electronics converters comprises whole system are explained in which this section. power point andoutput hence(Dailii introduces high fluctuations resulting electronics converters comprises the whole WECS in low power et al., 2015). Another control technique which is used frequently is2015). PI control method. PI electronics converters which comprises the whole WECS in low power output (Dailii et al., Another control system are explained in this section. in low power output (Dailii et al., is2015). Another control technique which is used frequently PI control method. PI system system are are explained explained in in this this section. section. technique technique which which is is used used frequently frequently is is PI PI control control method. method. PI PI Copyright © 2016, 2016 IFAC 462Hosting by Elsevier Ltd. All rights reserved. 2405-8963 © IFAC (International Federation of Automatic Control) Copyright ©under 2016 responsibility IFAC 462Control. Peer review of International Federation of Automatic Copyright © 462 Copyright © 2016 2016 IFAC IFAC 462 10.1016/j.ifacol.2016.03.097
IFAC ACODS 2016 February 1-5, 2016. NIT Tiruchirappalli, IndiaRamji Tiwari et al. / IFAC-PapersOnLine 49-1 (2016) 462–467
Fig. 1 shows the overall view of WECS. A PMSG generator is mostly preferred for standalone WECS. The PMSG is direct driven thus there is no need of gear box to drive the generator thus reducing the complexity and size of entire system (Dehghan et al., 2009). They require low maintenance when compared to other generators. The diode controlled rectifier is used to convert AC voltage obtained from generator to DC voltage in order to eliminate the harmonics present in it due to linearity. A smoothing capacitor CDC is used to remove ripple present in DC voltage. A DC/DC converter is designed to obtain the desired output. DIODE RECTIFIER
Tm
Pm
From the above equation (4) it can be determined that the performance of the wind turbine is highly dependent on the wind speed (Galdi et al., 2008) Table 1 summarizes the parameters of the wind turbine which is used for the simulation analysis. Table 1. Parameter of Wind Turbine 8.5kW
Nominal Mechanical Output Power Base Wind Speed Radius of wind Turbine Air Density,
DC/DC CONVERTER
+
Cdc
(4)
m
PMSG
w
463
-
Pitch Angle,
12m/s 1.001m 1.225kg/m3 0o
DC GRID
2.2 Modelling of PMSG generator
GRID SIDE CONTROL
PMSG is widely used for stand alone small wind turbines because they have high efficiency and less maintenance (Baroudi et al., 2007). The PMSG is modelled in dq reference frame. Both d and q axis contains a voltage induced by the armature. The generator is implemented with DC Voltage and current. The current of d axis and q axis is determined by the equation 5 and 6 respectively (Alizadeh et al., 2015)
Fig. 1. Overall schematic of PMSG based WECS 2.1 Turbine Modelling The mechanical power, Pm captured by the turbine is given by the equation 1 3 Pm C p , R 2Vwind 2
L sq disd R 1 i sq u sd sa i sd s dt Lsd L sd L sd
(1)
disq
Where C p is a rotor power coefficient, is blade pitch angle, is a Tip Speed Ratio (TSR), is air density, R is
dt
1 L Rsa 1 i sq s sd i sd p u Lsq sq Lsq Lsq Lsq
(5)
(6)
The electromagnetic torque obtained from the rotor of PMSG is given by the equation 7 (Phankong et al., 2013)
radius of wind turbine blade and Vwind is the wind speed. The rotor power coefficient is defined by the fraction of available wind power that can be transformed to mechanical power. C p depends on the blade aerodynamics, which is the
Te 1.5
P p i sq i sd i sq Lsd Lsq 2
(7)
function of and . The power coefficient of turbine is determined by TSR. C p and TSR is determined by the shape
d and q axis respectively.
of the blade. However, C p in general blade design is
generator. L sd and Lsq are the inductance of the generator.
assumed for simplicity (Babu et al., 2013).
p is the permanent flux, R sa is the resistance of the stator
2 C p 0.44 0.0167 sin 0.0018 2 13 0.3
Where,
m R V wind
s is the angular frequency of the
and P is the number of poles.
(2)
Table 2. Parameters of PMSG
The TSR ( ) can be defined as the function of a wind speeds.
isd , isq , u sd and u sq are the currents and voltages of
Rated Power Stator Phase Resistance Armature Inductance Friction Factor Pole Pairs inertia
(3)
where, m is the rotor speed of a wind turbine. The input torque for the generator is obtained from the formula:
463
8.5kW 0.425 Ω 0.000835 H 0.001189 Nms 5 0.01197 kg.m2
IFAC ACODS 2016 464 February 1-5, 2016. NIT Tiruchirappalli, IndiaRamji Tiwari et al. / IFAC-PapersOnLine 49-1 (2016) 462–467
The Parameter of PMSG which is used for the study purpose in this work is presented in Table 2. According to author (Chen et al., 2015) the diode controlled rectifier are widely used for voltage conversion in small scale stand alone WECS due to its simplicity and low cost. However, Rectifier draws non sinusoidal current from PMSG. To overcome the problem a DC link capacitor is used, which can mitigate partial current harmonic elimination, thus achieving near sinusoidal current.
based MPPT control mainly focus on converting variable voltage and frequency to fixed voltage and frequency. The most commonly used power electronics converter configuration is analysed in this paper and the block diagram of the model used is shown in Fig. 3. DIODE RECTIFIER
2.4 Modelling of DC/DC Converter The DC/DC converter employed here is buck converter. Buck converter is used to step down the input DC voltage. By varying the duty cycle the output voltage can be controlled. This converter regulates the input voltage through the switch to reach the reference voltage which consists of maximum power. Fig. 2. shows the basic circuit of the buck converter. Buck converter operates in mainly two different modes namely continuous conduction mode and discontinuous conduction mode. (Bendib et al., 2014). The switch operates at high frequency to deliver chopped DC voltage output. The buck converter controls the power flow using the ON/OFF condition of switch which is controlled by the duty cycle switching. The average output voltage is given by the equation (8). Vout Vin * D
In this paper three MPPT techniques such as conventional PI controller, P&O method and FLC MPPT method are utilized and comparative study is done to choose the efficient and appropriate MPPT technique so that the maximum power is extracted from the available wind. 3.1 PI Control In PI based MPPT technique, an error signal is generated using the actual DC voltage and the reference DC voltage. The error signal is fed into the PI Controller from where an output signal is obtained. The output signal is then compared with the frequency repetitive triangular waveform to deliver a duty cycle which operates the DC/DC Converter switch thus obtaining the maximum power based on the variance in the wind speed. The equation used for PI controller is
Inductor
Vin
Diode
-
Capacitor
Load
Idc
Fig. 3. Converter configuration of PMSG based WECS
+ +
Vdc
MPPT
where Vout is the output Voltage Vin is the input voltage and D represents the duty cycle of the converter switch.
CONTROLLER
-
BUCK CONVERTER
Duty Cycle
(8)
Switch
Cdc
LOAD
+
PMSG
K Ds K p i * E s s
Vout
(9)
Where, K p is proportional parameter and K i is integral parameter. E s , represents the error between reference voltage and output voltage and Ds is the duty cycle generated by the PI controller (Martin et al., 2015). Fig. 4. represents basic structure of PI controller.
-
Fig. 2. Basic design of buck converter Table 3. Parameters of buck converter
Vref
Inductor (L) 3.65 mH Capacitor (C) 3 mF Load(R) 150 Ω In Table 3. the inductor and capacitor values which is used for simulation study is determined.
+
Error Signal
Duty Cycle PI Controller
-
Vout
3. CONTROL STRATEGIES OF WECS Fig. 4. PI controller The MPPT based control strategy is used here to obtain the maximum power. Wind energy even though available in abundant, the wind speed varies rapidly. The efficiency of the WECS depends upon the accuracy in which the maximum power is extracted by the MPPT controller. The PMSG
3.2 Perturb and Observe Control The P&O method is used to search for the maximum optimal
464
IFAC ACODS 2016 February 1-5, 2016. NIT Tiruchirappalli, IndiaRamji Tiwari et al. / IFAC-PapersOnLine 49-1 (2016) 462–467
point for the given wind speed. The P&O method does not require any prior wind turbine knowledge. It is independent, flexible and simple technique. Here the P&O method uses the perturbed output voltage across the load to determine the optimal operating point that will extract the maximum power. If the power of the current cycle is greater than the previous one then the voltage is modified in same technique as the previous one. Whereas, if the power is lesser than the previous technique the voltage must be varied in the opposite direction. The only disadvantage of P&O technique in wind energy conversion is that they cannot track the rapid variation of the wind speed thus affecting the efficiency of the overall system and the speed of convergence (Dalala et al., 2013). The flow chart of P&O method is described in Fig. 5.
465
Read: V(t), I(t)
ΔI =I(t) - I(t-Δt) ΔV =V(t) - V(t-Δt)
No
ΔI/ΔV=-I/V
3.3 Fuzzy Logic Controller To overcome the drawbacks of P&O method FLC algorithm is proposed. FLC has an advantage of fast convergence, imprecise input and handling non linearity. FLC generally consist of three stages Fuzzification, Rule base lookup table and Defuzzification as shown in Fig.6. The rules are designed on the basis of previous knowledge of the system (Simoes et al., 1997). An FLC is the artificial decision making controller that operates in closed loop. The inputs for fuzzy controllers are error signal and change in error signal. Once the signals are calculated and linguistic variables are obtained, the output of FLC is the duty cycle for buck converter is which is generated using the rules. FLC is termed to be the most efficient MPPT controller when compared with the PI and P&O controller (Tripathi et al., 2015). The efficiency of FLC is purely depend upon the previous knowledge of the system and right error computation and framing of rule based table.
Yes
ΔV = 0
Yes
Yes
ΔI>0
No
No
No
Yes
Yes
Increase Vref
Increase Vref
ΔI/ΔV>-I/V
Decrease Vref
No
ΔI>0
Decrease Vref
I(t-Δt) = I(t) V(t-Δt) = V(t)
Table. 4, represents the set of rules used for modelling FLC. Where, E represents the error signal and CE represents the Change in error. The rules are framed in five level namely Negative Big(NB), Negative Small(NS), Zero(ZE), Positive Small(PS) and Positive Big(PB).
Return
Fig. 5. Flowchart of P&O based control technique Table 4. FLC Set of rules E/CE
NB
NS
ZE
PS
PB
NB
ZE
PB
ZE
NB
NS
NS
PS
ZE
ZE
NB
NS
ZE
ZE
ZE
ZE
ZE
ZE
PS
PS
PB
ZE
ZE
NS
PB
PS
PB
ZE
NB
ZE
FUZZY SET
FUZZY RULES
ERROR
DUTY CYCLE
FUZZIFICATION
DEFUZZIFICATION
CHANGE IN ERROR
Inference mechanism is basically defined by membership functions of FLC which determines the relevance of rules from Table 4. Fig. 7, 8 and 9 represents the input and output membership function of FLC controller. Methods for implication and aggregation are defined as Minimum (min) and Maximum (max) respectively. The Defuzzification method uses centroid for processing.
INFERENCE MECHANISM
Fig. 6. Basic structure of FLC
465
IFAC ACODS 2016 466 February 1-5, 2016. NIT Tiruchirappalli, IndiaRamji Tiwari et al. / IFAC-PapersOnLine 49-1 (2016) 462–467
maximum and constant voltage during the steady state characteristic of wind turbine. To analyse the performance of each MPPT controller power output is compared as shown in Fig. 13.
Fig. 7. Input membership function of error signal
Fig. 10. MPPT controller model for WECS in Simulink Wind Speed Variation
Fig. 8. Input membership function of change in error signal
14 12
Wind Speed (m/s)
10 8 6 4
Fig. 9. Output membership function of duty cycle
2
FLC tracks the sudden change in wind speed more swiftly and precisely. The maximum power point is traced by the controller from the inference system which is mapped by the human knowledge earlier in form of rules. The controller tracks the change in output voltage, current and generates an error signal which is given as an input for fuzzification process, here the input data is converted into a suitable fuzzy linguistic sets using Mamdani method. Then the fuzzy set is processed in inference system where an appropriate fuzzy output is obtained using fuzzy rules. Then the fuzzy output is converted in to the systematic crisp value as a form of duty cycle in defuzzification. Thus the duty cycle is used to control the switching pattern of the converter switch.
0
0
2
4 6 Time (seconds)
8
10
Fig. 11. Wind speed variation 600 P&O
PI
Fuzzy
500
Voltage (Vdc)
400
4. RESULTS AND DISCUSSION
300
200
The detailed implementation of PMSG based WECS using buck converter incorporated with fuzzy logic controller in MATLAB/ Simulink is shown in Fig. 10. To check the tracking ability of MPPT techniques, the wind speed is varied from 3m/s to 12m/s as shown in Fig. 11. Three MPPT approaches in simulated in same varying wind speed conditions. From the analysis it can be stated that the FLC based MPPT technique is most efficient maximum power tracking power when compared with PI and P&O based technique.
100
0
0
2
4 6 Time (Sec)
8
10
Fig. 12. Output voltage for different MPPT techniques Table 5. shows the performance comparison of PI, P&O and FLC based MPPT controllers. The FLC has the high power out when compared with the other two controllers. The output voltage of FLC is equal to the reference DC voltage estimated to be the DC bus voltage of 500V. Whereas, the PI
Fig. 12 shows the output voltage comparison of MPPT controllers. The FLC based MPPT controller gives the 466
IFAC ACODS 2016 February 1-5, 2016. NIT Tiruchirappalli, IndiaRamji Tiwari et al. / IFAC-PapersOnLine 49-1 (2016) 462–467
and P&O controllers have an output voltage of 491V and 494V. Though there is only little difference in the power and voltage but when large wind turbines are connected then the difference will become huge.
Babu, N.R. and Arulmozhivarman, P. (2013). Wind Energy Conversion System- A Techical Review. Journal of Engineering Science and Technology, volume (8), 493 507. Chen, W.L. and Jiang, B.Y. (2015). Harmonic Suppression and Performance Improvement for a Small-scale Gridtied Wind Turbine using Proportional-Resonant Controllers. Electric Power Components and Systems, volume (43), 970 - 981. Baroudi, J.A., Dinavahi, V., and Knight, A.M. (2007). A review of power converter topologies for wind generators. Renewable Energy, volume (32), 2369 2385. Bendib, B., Krim, F., Belmili, H., Almi, M.F., and Boulouma, S. (2014). Advanced Fuzzy MPPT Controller for a Stand-alone PV system. Energy Procedia, volume (50), 383 - 392. Dalala, Z.M., Zahid, Z.U., Yu, W., Cho, Y., and Lai, J. (2013). Design and analysis of an MPPT technique for small-scale wind energy conversion systems. IEEE Transaction on Energy Conversion, volume (28),756 766. Dehghan, S/M., Mohamadian, M., and Varjani, A.Y. (2009). A new variable-speed wind energy conversion system using permanent-magnet synchronous generator and Zsource inverter. IEEE Transaction on Energy Conversion, volume (24),714 - 724. Errami, Y., Ouassaid, M., and Maaroufi, M. (2015). A performance comparison of a nonlinear and a linear control for grid connected PMSG wind energy conversion system,. International Journal of Electrical Power & Energy Systems, volume (68), 180 - 194. Galdi, V., Piccolo, A., and Siano, P. (2008). Designing an adaptive fuzzy controller for maximum wind energy extraction. IEEE Transaction on Energy Conversion, volume (23),559 - 569. Krishna, K.S., and Kumar, K.S. (2015). A review on hybrid energy systems. Renewable and Sustainable Energy Reviews, volume (52), 907 - 916. Martin, A.D. and Vazquez, J.R. (2015). MPPT algorithm comparison in PV systems: P&O, PI, neuro-fuzzy and backstepping controls. International Conference on Industial Technology, 2841 - 2847. Phankong, N., Manmai, S., Bhumkittipich, K., and Nakawiwat, P. (2013). Modeling of grid-connected with permanent magnet synchronous generator (PMSG) using voltage vector control. Energy Procedia, volume (34), 262 - 272. Tripathi, S.M., Tiwari, A.N., and Singh, D. (2015). Gridconnected permanent magnet synchronous generator based wind energy conversion systems: A technology review. Renewable and Sustainable Energy Reviews, volume (51), 1288 - 1305. Simoes, M.G., Bose, B.K., and Spiegel, R.J. (1997). Design and performance evaluation of fuzzy-logic-based variable-speed wind generation system. IEEE Transaction on Industrial Applications, volume (33), 956 - 965.
6000 Fuzzy
P&O
PI
5000
Power (W)
4000
3000
2000
1000
0
0
2
4
6
8
10
Time (sec)
Fig. 13. Power obtained by different MPPT techniques Table 5. Output of various MPPT techniques MPPT Technique/Parameters
PI Controller
P&O Controller
FLC
Power
4992 W
5093 W
5112 W
Voltage
491 V
494 V
500 V
467
5. CONCLUSIONS In this paper, three MPPT controllers like PI, P&O and FLC controller is modelled and the output is compared for wind energy under varying wind speed condition. The performance of each controller is analysed and it is verified that FLC based controller is more efficient and reliable than PI and P&O based controller. The PI controller fails to track the non linearity of the wind speed thus providing poor output power. The P&O based technique is suitable for the condition where the system is stable of with minimum variance. Wind Speed being high non linear the P&O algorithm oscillates around optimal point thus making it difficult to track the next point. The FLC method has a rapid tracking ability. The FLC control method requires previous knowledge of the system so that it can provide an efficient and steady state output. FLC is fast and efficient technique to track the maximum power point in WECS. The results obtained from FLC are superior and efficient than that of PI and P&O technique in terms of stability, faster tracking ability and fluctuations. Hence, it is concluded that the FLC based MPPT method is the best option for stand alone WECS. REFERENCES Alizadeh, M. and Kojori, S.S. (2015). Augmenting effectiveness of control loop of a PMSG based wind energy conversion system by a virtually adaptive PI controller. Energy, volume (91), 610 - 629. 467
ScienceDirect
49-1 (2016) 462–467 FUZZY LOGIC BASEDIFAC-PapersOnLine MPPT FOR PERMANENT MAGNET SYNCHRONOUS FUZZY LOGIC BASED MPPT FOR PERMANENT MAGNET SYNCHRONOUS FUZZY BASED FOR PERMANENT MAGNET SYNCHRONOUS GENERATOR IN WIND ENERGY CONVERSION FUZZY LOGIC LOGIC BASED MPPT MPPT FOR PERMANENT MAGNETSYSTEM SYNCHRONOUS GENERATOR IN WIND ENERGY CONVERSION SYSTEM GENERATOR IN WIND ENERGY CONVERSION SYSTEM GENERATOR IN Ramji WIND ENERGY CONVERSION SYSTEM Tiwari*, Ramesh Babu. N**
Ramji Tiwari*, Tiwari*, Ramesh Babu. Babu. N** Ramji Ramji Tiwari*, Ramesh Ramesh Babu. N** N** School of Electrical Engineering, VIT University, Vellore, India. School of of Electrical Engineering, Engineering, VIT University, University, Vellore, India. India. School *[email protected],**[email protected] School of Electrical Electrical Engineering, VIT VIT University, Vellore, Vellore, India. *[email protected],**[email protected] *[email protected],**[email protected] *[email protected],**[email protected] Abstract: In this paper, a comparative analysis of different control methods to extract the maximum Abstract: InPermanent this paper, paper,Magnet a comparative comparative analysis of different different control methods to extract extract the maximum maximum power fromIn Synchronous Generator (PMSG) based methods Wind Energy Conversion System Abstract: this analysis of control to the Abstract: InPermanent this paper,Magnet aa comparative analysis of different control methods to extract the maximum power from Synchronous Generator (PMSG) based Wind Energy Conversion Systema (WECS) under differentMagnet wind speed condition is presented. Thebased WECS consists ofConversion a wind turbine, power from Permanent Synchronous Generator (PMSG) Wind Energy System power from Permanent Magnet Synchronous Generator (PMSG) based Wind EnergyofConversion Systema (WECS) under different wind speed condition is presented. The WECS consists a wind turbine, PMSG and a DC/DC converter whichcondition is connected to a DC load. The Maximum Power Pointturbine, Trackinga (WECS) under different wind is The WECS consists of (WECS) under different wind speed speed is presented. presented. The The WECS consistsPower of aa wind wind turbine, a PMSG and and a DC/DC DC/DC converter whichcondition is connected connected to aa DC DC load. load. Maximum Point Tracking (MPPT) control technique compared here are Proportional Integral (PI) control, Power PerturbPoint and Tracking Observe PMSG a converter which is to The Maximum PMSG and a DC/DC converter which is connected to a DC load. The Maximum Power Point Tracking (MPPT) control and technique compared here are Proportional Proportional Integral (PI) considered control, Perturb and Observe (P&O) method Fuzzycompared Logic Controller (FLC). The Integral parameters for analysing the (MPPT) control technique here (PI) control, and (MPPT) control and technique compared here are are Proportional Integral (PI) considered control, Perturb Perturb and Observe Observe (P&O) method Fuzzy Logic Controller (FLC). The parameters for analysing the efficiency of the MPPT controller is the output DC voltage and power across the load. The steady state (P&O) method and Fuzzy Logic (FLC). The parameters considered for analysing the (P&O) method and Fuzzy Logicis Controller Controller (FLC). The and parameters considered forThe analysing the efficiency of the MPPT controller the output DC voltage power across the load. steady state voltage and the dynamic response of the system under different wind speed is considered to justify the efficiency of the MPPT is the output DC voltage and across load. state efficiency ofthe thedynamic MPPT controller controller is the output DC voltage and power power acrossisthe the load. The Thetosteady steady state voltage and response of the system under different wind speed considered justify the The system is designed and configured in MATLAB/SIMULINK overall efficiency of the controllers. voltage and response the under different wind speed to voltage and the the dynamic dynamic response of of The the system system under different wind speed is isinconsidered considered to justify justify the the system is designed and configured MATLAB/SIMULINK overall efficiency of the controllers. software and the results are validated. The overall of controllers. The system system is is designed designed and and configured configured in in MATLAB/SIMULINK MATLAB/SIMULINK overall efficiency efficiency of the theare controllers. software and the results validated. software and the results are validated. © 2016, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved. software and the results are validated. Keywords: Wind turbine, Permanent Magnet Synchronous Generator, Fuzzy Logic Controller, MPPT. Keywords: Wind turbine, Permanent Magnet Synchronous Generator, Fuzzy Logic Controller, MPPT. Keywords: Wind turbine, Permanent Magnet Synchronous Generator, Fuzzy Logic Controller, MPPT. Keywords: Wind turbine, Permanent Magnet Synchronous Generator, Fuzzy Logic Controller, MPPT. control is very simple to implement. PI control lacks in 1. INTRODUCTION control is very implement. PI control lacks in efficiency of the simple overall to system due to its arbitrary selection control is very simple to implement. PI control lacks 1. INTRODUCTION control is of very simple to implement. PIarbitrary control selection lacks in in 1. INTRODUCTION efficiency the overall system due to its of parameters. In order to overcome this problem FLC is 1. INTRODUCTION In recent years, the use of renewable energy resource is efficiency of system due its selection efficiency of the theInoverall overall system due to to this its arbitrary arbitrary selection of parameters. order to overcome problem FLC is employed to extract the to maximum power from wind. FLC In recent years, the use of renewable energy resource is increased due to increasing demand of power and depletion of parameters. In order overcome this problem FLC is In recent years, the use of renewable energy resource is of parameters. In order to overcome this from problem FLC is In recent years, the use of renewable energy resource is employed to extract the maximum power wind. FLC can track to theextract non linearity of thepower systemfrom andwind. givesFLC the increased due to increasing demand of power and depletion of fossil fuel such as coal for electricity generation. employed the maximum increased due to increasing demand of power and depletion employed to extract the maximum power from wind. FLC increased due to such increasing demand ofelectricity power andgeneration. depletion can track the nonforlinearity of thewind. system gives maximum output the available Theand input for the of fossil fuel coal for Moreover, the issue ofas climate change are causing concern can track the of system and gives the of fossil fuel such as coal for electricity generation. can track output the non nonforlinearity linearity of the thewind. system and gives the of fossil fuel such as coal for electricity generation. maximum the available The input for the FLC is the DC voltage and current across the load, whereas Moreover, the issue of climate change are causing concern and hence many regulations are proposed to reduce the maximum output for the available wind. The input for the Moreover, the issue of climate change are causing concern maximum output for the available wind. The input for the Moreover, the issue of climate change are causing concern FLC is the DC voltage and current across the load, whereas the output is the duty cycle for the DC/DC Converter. and hence many regulations are proposed to reduce the ) emission (Krishna et al., 2015). Carbon dioxide (CO FLC is the DC voltage and current across the load, whereas 2 and hence many regulations are proposed to reduce the FLC is the DC voltage and current across the load, whereas and hence many regulations are proposed to reduce the the output is the duty cycle for the DC/DC Converter. )) emission et al., 2015). Carbon dioxide (CO Among the renewable solar (Krishna energy and wind energy 2energy, the the cycle for (Krishna et al., 2015). Carbon dioxide (CO output isconverter the duty duty is cycle for the the DC/DC DC/DC Converter. Theoutput buck is employed here to Converter. interface the wind ) emission emission (Krishna et wind al., energy 2015). the Carbon dioxide (CO22energy, Among the renewable solar energy and is moretheutilised because ofsolar itsenergy abundant availability Among renewable energy, and wind energy The buck converter is employed here to interface the generator to the DC Load. The output of the PMSG iswind AC Among the renewable energy, solar energy and wind energy The is employed here interface the is more utilised because its abundant everywhere. The Wind Energyof is gaining interest availability because of generator The buck bucktoconverter converter is employed here to to interface theiswind wind is more utilised because of its abundant availability the DC Load. The output of the PMSG AC which is converted in to DC using diode controlled rectifier is more utilised because of its abundant availability generator to The output of the is everywhere. The Wind Energy gaining interest because of technology enhancement andis significant power cost generator to the the DC DCinLoad. Load. The output of controlled the PMSG PMSGrectifier is AC AC everywhere. The Wind Energy is gaining interest because of which is converted to DC using diode in order to eliminate the ripple present in the AC component everywhere. The Wind Energy is gaining interest because of which is in to DC diode controlled rectifier technology enhancement significant cost reduction. More effective and control strategiespower are under which is toconverted converted inthe to ripple DC using using diode controlled rectifier technology enhancement and significant power cost in order eliminate present in the AC component andorder smoothing capacitor is placed across the rectifier to technology enhancement and significant power cost in to the present in AC component reduction. More control are under research in order toeffective obtain reliable, coststrategies effective and quality in order to eliminate eliminate the ripple ripple presentacross in the the the AC rectifier component reduction. More effective control strategies are under and smoothing capacitor is placed to minimise the ripple due to non linearity. When the wind reduction. More effective control strategies are under and smoothing capacitor is placed across the rectifier to research in order obtain cost effective power from the to wind. Asreliable, stated in (Errami et and al., quality 2015) minimise and smoothing capacitor is non placed across When the rectifier to research in order to obtain reliable, cost effective and quality the ripple due to linearity. the wind power varies the FLC tracks the output voltage and current to research in order to obtain reliable, cost effective and quality minimise the ripple due to non linearity. When the wind power from the wind. As stated in (Errami et al., 2015) Permanent Magnet Synchronous Generator (PMSG) is most minimise the ripple due to non linearity. When the wind power from the wind. As stated in (Errami et al., 2015) power varies the FLC tracks the output voltage and current to generate an efficient duty cycle for the converter operation. power from the wind. As stated in (Errami et al., 2015) power the tracks the voltage and current Permanent Magnet Synchronous (PMSG) is most preferred wind generator due to Generator its reliability and size for generate power varies varies the FLC FLC duty trackscycle the output output voltage and operation. current to to Permanent Magnet Synchronous Generator (PMSG) is most an efficient for the converter Thus the maximum power is produced based on available Permanent Magnet Synchronous Generator (PMSG) is most generate an efficient duty cycle for the converter operation. preferred due to its reliability and size for stand alonewind windgenerator energy conversion system. generate an efficient duty cycle for the converter operation. preferred wind generator due to its reliability and size for the maximum power is produced based on available wind speed. preferred wind generator due to itssystem. reliability and size for Thus Thus the stand alone wind energy conversion Thus the maximum maximum power power is is produced produced based based on on available available stand alone wind energy conversion system. wind speed. stand alone wind energy conversion system. MPPT control algorithms can be employed in order to wind speed. wind speed. This paper analyses the maximum power point tracking MPPT control algorithms can be employed order to capture the maximum power from availablein wind by MPPT control algorithms can be employed in order to This paper analyses the maximum power point tracking control approach for stand alone WECS. The performance of MPPT control algorithms can be employed in order to This paper analyses the maximum power point capture the maximum from available maintaining the optimum power steady voltage across thewind load. by A This paper analyses the alone maximum power point tracking tracking capture the maximum power from available wind by control approach for stand WECS. The performance of the conventional PI controller, P&O controller and FLC capture the maximum power from available wind by control approach for stand alone WECS. The performance of maintaining the optimum steady across the A variety of MPPT techniques havevoltage been employed forload. Wind control approach for stand alone WECS. The performance of maintaining the optimum steady voltage across the load. A the conventional PI controller, P&O controller and FLC under variable wind speed is evaluated. The proposed control maintaining the optimum steady voltage across the load. A the conventional PI controller, P&O controller and FLC variety of MPPT techniques have been employed for Wind Energy Conversion System (WECS) in previous literature the conventional PI controller, P&O controller and FLC variety of MPPT have for Wind variable is evaluated. The proposed control strategy posseswind the speed improved capability of capturing the variety of MPPT techniques techniques have been beeninemployed employed for Wind under under variable wind speed is The proposed control Energy Conversion System previous literature such as Hill Climbing Search(WECS) (HCS) algorithm, Incremental under variable wind speed is evaluated. evaluated. Theof proposed control Energy Conversion System (WECS) in previous literature strategy posses the improved capability capturing maximum power from wind. Comparative efficiency of Energy Conversion System (WECS) in previous literature strategy posses the improved capability of capturing the the such as Hill Climbing Search (HCS) algorithm, Incremental and Conductance method (INC), Perturb and Observe (P&O) strategy posses the improved capability of capturing the such as Hill Climbing Search (HCS) algorithm, Incremental maximum power from wind. of the controllers is analysed fromComparative the outputefficiency power of such as Hill Climbing Search (HCS) algorithm, Incremental maximum power from wind. Comparative efficiency of the and Conductance method (INC), Perturb and Observe (P&O) method, Fuzzy method Logic (INC), Controller (FLC) and (P&O) many maximum power from wind. Comparative efficiency of the and Conductance Perturb and Observe controllers is analysed from the output power of the converter. and Conductance method (INC), Perturb and Observe (P&O) controllers is analysed from the output power of the method, Fuzzy Logic Controller (FLC) and many Evolutionary Algorithms. controllers is analysed from the output power of the method, Fuzzy Logic Controller (FLC) and many converter. method, Fuzzy Logic Controller (FLC) and many converter. Evolutionary Algorithms. converter. Evolutionary Algorithms. 2. MODELLING OF WIND ENERGY CONVERSION Evolutionary P&O methodAlgorithms. is well known MPPT technique due to its 2. ENERGY SYSTEM P&O method is well known MPPT technique due to its simplicity and effectiveness. But due to high non linearity in 2. MODELLING MODELLING OF OF WIND WIND ENERGY CONVERSION CONVERSION P&O method is well known MPPT technique due to its 2. MODELLING OF WIND ENERGY CONVERSION P&O method is well known MPPT technique due to its SYSTEM simplicity and effectiveness. due high non in wind speed, P&O techniqueBut fails toto track the linearity maximum SYSTEM simplicity and effectiveness. But due to high non linearity in SYSTEM simplicity andP&O effectiveness. But duetoto track high non linearity in The simulation models of wind turbine, PMSG and power wind speed, technique fails the maximum power point and hence introduces high fluctuations resulting The simulation models of wind turbine, PMSG and power wind speed, P&O technique fails to track the maximum electronics converters which comprises the whole WECS wind speed, P&O technique fails to track the maximum simulation models of turbine, PMSG and power power and hence introduces high fluctuations in low point power output (Dailii et al., 2015). Anotherresulting control The The simulation models which of wind wind turbine, the PMSG and WECS power power point and hence introduces high fluctuations resulting electronics converters comprises whole system are explained in which this section. power point andoutput hence(Dailii introduces high fluctuations resulting electronics converters comprises the whole WECS in low power et al., 2015). Another control technique which is used frequently is2015). PI control method. PI electronics converters which comprises the whole WECS in low power output (Dailii et al., Another control system are explained in this section. in low power output (Dailii et al., is2015). Another control technique which is used frequently PI control method. PI system system are are explained explained in in this this section. section. technique technique which which is is used used frequently frequently is is PI PI control control method. method. PI PI Copyright © 2016, 2016 IFAC 462Hosting by Elsevier Ltd. All rights reserved. 2405-8963 © IFAC (International Federation of Automatic Control) Copyright ©under 2016 responsibility IFAC 462Control. Peer review of International Federation of Automatic Copyright © 462 Copyright © 2016 2016 IFAC IFAC 462 10.1016/j.ifacol.2016.03.097
IFAC ACODS 2016 February 1-5, 2016. NIT Tiruchirappalli, IndiaRamji Tiwari et al. / IFAC-PapersOnLine 49-1 (2016) 462–467
Fig. 1 shows the overall view of WECS. A PMSG generator is mostly preferred for standalone WECS. The PMSG is direct driven thus there is no need of gear box to drive the generator thus reducing the complexity and size of entire system (Dehghan et al., 2009). They require low maintenance when compared to other generators. The diode controlled rectifier is used to convert AC voltage obtained from generator to DC voltage in order to eliminate the harmonics present in it due to linearity. A smoothing capacitor CDC is used to remove ripple present in DC voltage. A DC/DC converter is designed to obtain the desired output. DIODE RECTIFIER
Tm
Pm
From the above equation (4) it can be determined that the performance of the wind turbine is highly dependent on the wind speed (Galdi et al., 2008) Table 1 summarizes the parameters of the wind turbine which is used for the simulation analysis. Table 1. Parameter of Wind Turbine 8.5kW
Nominal Mechanical Output Power Base Wind Speed Radius of wind Turbine Air Density,
DC/DC CONVERTER
+
Cdc
(4)
m
PMSG
w
463
-
Pitch Angle,
12m/s 1.001m 1.225kg/m3 0o
DC GRID
2.2 Modelling of PMSG generator
GRID SIDE CONTROL
PMSG is widely used for stand alone small wind turbines because they have high efficiency and less maintenance (Baroudi et al., 2007). The PMSG is modelled in dq reference frame. Both d and q axis contains a voltage induced by the armature. The generator is implemented with DC Voltage and current. The current of d axis and q axis is determined by the equation 5 and 6 respectively (Alizadeh et al., 2015)
Fig. 1. Overall schematic of PMSG based WECS 2.1 Turbine Modelling The mechanical power, Pm captured by the turbine is given by the equation 1 3 Pm C p , R 2Vwind 2
L sq disd R 1 i sq u sd sa i sd s dt Lsd L sd L sd
(1)
disq
Where C p is a rotor power coefficient, is blade pitch angle, is a Tip Speed Ratio (TSR), is air density, R is
dt
1 L Rsa 1 i sq s sd i sd p u Lsq sq Lsq Lsq Lsq
(5)
(6)
The electromagnetic torque obtained from the rotor of PMSG is given by the equation 7 (Phankong et al., 2013)
radius of wind turbine blade and Vwind is the wind speed. The rotor power coefficient is defined by the fraction of available wind power that can be transformed to mechanical power. C p depends on the blade aerodynamics, which is the
Te 1.5
P p i sq i sd i sq Lsd Lsq 2
(7)
function of and . The power coefficient of turbine is determined by TSR. C p and TSR is determined by the shape
d and q axis respectively.
of the blade. However, C p in general blade design is
generator. L sd and Lsq are the inductance of the generator.
assumed for simplicity (Babu et al., 2013).
p is the permanent flux, R sa is the resistance of the stator
2 C p 0.44 0.0167 sin 0.0018 2 13 0.3
Where,
m R V wind
s is the angular frequency of the
and P is the number of poles.
(2)
Table 2. Parameters of PMSG
The TSR ( ) can be defined as the function of a wind speeds.
isd , isq , u sd and u sq are the currents and voltages of
Rated Power Stator Phase Resistance Armature Inductance Friction Factor Pole Pairs inertia
(3)
where, m is the rotor speed of a wind turbine. The input torque for the generator is obtained from the formula:
463
8.5kW 0.425 Ω 0.000835 H 0.001189 Nms 5 0.01197 kg.m2
IFAC ACODS 2016 464 February 1-5, 2016. NIT Tiruchirappalli, IndiaRamji Tiwari et al. / IFAC-PapersOnLine 49-1 (2016) 462–467
The Parameter of PMSG which is used for the study purpose in this work is presented in Table 2. According to author (Chen et al., 2015) the diode controlled rectifier are widely used for voltage conversion in small scale stand alone WECS due to its simplicity and low cost. However, Rectifier draws non sinusoidal current from PMSG. To overcome the problem a DC link capacitor is used, which can mitigate partial current harmonic elimination, thus achieving near sinusoidal current.
based MPPT control mainly focus on converting variable voltage and frequency to fixed voltage and frequency. The most commonly used power electronics converter configuration is analysed in this paper and the block diagram of the model used is shown in Fig. 3. DIODE RECTIFIER
2.4 Modelling of DC/DC Converter The DC/DC converter employed here is buck converter. Buck converter is used to step down the input DC voltage. By varying the duty cycle the output voltage can be controlled. This converter regulates the input voltage through the switch to reach the reference voltage which consists of maximum power. Fig. 2. shows the basic circuit of the buck converter. Buck converter operates in mainly two different modes namely continuous conduction mode and discontinuous conduction mode. (Bendib et al., 2014). The switch operates at high frequency to deliver chopped DC voltage output. The buck converter controls the power flow using the ON/OFF condition of switch which is controlled by the duty cycle switching. The average output voltage is given by the equation (8). Vout Vin * D
In this paper three MPPT techniques such as conventional PI controller, P&O method and FLC MPPT method are utilized and comparative study is done to choose the efficient and appropriate MPPT technique so that the maximum power is extracted from the available wind. 3.1 PI Control In PI based MPPT technique, an error signal is generated using the actual DC voltage and the reference DC voltage. The error signal is fed into the PI Controller from where an output signal is obtained. The output signal is then compared with the frequency repetitive triangular waveform to deliver a duty cycle which operates the DC/DC Converter switch thus obtaining the maximum power based on the variance in the wind speed. The equation used for PI controller is
Inductor
Vin
Diode
-
Capacitor
Load
Idc
Fig. 3. Converter configuration of PMSG based WECS
+ +
Vdc
MPPT
where Vout is the output Voltage Vin is the input voltage and D represents the duty cycle of the converter switch.
CONTROLLER
-
BUCK CONVERTER
Duty Cycle
(8)
Switch
Cdc
LOAD
+
PMSG
K Ds K p i * E s s
Vout
(9)
Where, K p is proportional parameter and K i is integral parameter. E s , represents the error between reference voltage and output voltage and Ds is the duty cycle generated by the PI controller (Martin et al., 2015). Fig. 4. represents basic structure of PI controller.
-
Fig. 2. Basic design of buck converter Table 3. Parameters of buck converter
Vref
Inductor (L) 3.65 mH Capacitor (C) 3 mF Load(R) 150 Ω In Table 3. the inductor and capacitor values which is used for simulation study is determined.
+
Error Signal
Duty Cycle PI Controller
-
Vout
3. CONTROL STRATEGIES OF WECS Fig. 4. PI controller The MPPT based control strategy is used here to obtain the maximum power. Wind energy even though available in abundant, the wind speed varies rapidly. The efficiency of the WECS depends upon the accuracy in which the maximum power is extracted by the MPPT controller. The PMSG
3.2 Perturb and Observe Control The P&O method is used to search for the maximum optimal
464
IFAC ACODS 2016 February 1-5, 2016. NIT Tiruchirappalli, IndiaRamji Tiwari et al. / IFAC-PapersOnLine 49-1 (2016) 462–467
point for the given wind speed. The P&O method does not require any prior wind turbine knowledge. It is independent, flexible and simple technique. Here the P&O method uses the perturbed output voltage across the load to determine the optimal operating point that will extract the maximum power. If the power of the current cycle is greater than the previous one then the voltage is modified in same technique as the previous one. Whereas, if the power is lesser than the previous technique the voltage must be varied in the opposite direction. The only disadvantage of P&O technique in wind energy conversion is that they cannot track the rapid variation of the wind speed thus affecting the efficiency of the overall system and the speed of convergence (Dalala et al., 2013). The flow chart of P&O method is described in Fig. 5.
465
Read: V(t), I(t)
ΔI =I(t) - I(t-Δt) ΔV =V(t) - V(t-Δt)
No
ΔI/ΔV=-I/V
3.3 Fuzzy Logic Controller To overcome the drawbacks of P&O method FLC algorithm is proposed. FLC has an advantage of fast convergence, imprecise input and handling non linearity. FLC generally consist of three stages Fuzzification, Rule base lookup table and Defuzzification as shown in Fig.6. The rules are designed on the basis of previous knowledge of the system (Simoes et al., 1997). An FLC is the artificial decision making controller that operates in closed loop. The inputs for fuzzy controllers are error signal and change in error signal. Once the signals are calculated and linguistic variables are obtained, the output of FLC is the duty cycle for buck converter is which is generated using the rules. FLC is termed to be the most efficient MPPT controller when compared with the PI and P&O controller (Tripathi et al., 2015). The efficiency of FLC is purely depend upon the previous knowledge of the system and right error computation and framing of rule based table.
Yes
ΔV = 0
Yes
Yes
ΔI>0
No
No
No
Yes
Yes
Increase Vref
Increase Vref
ΔI/ΔV>-I/V
Decrease Vref
No
ΔI>0
Decrease Vref
I(t-Δt) = I(t) V(t-Δt) = V(t)
Table. 4, represents the set of rules used for modelling FLC. Where, E represents the error signal and CE represents the Change in error. The rules are framed in five level namely Negative Big(NB), Negative Small(NS), Zero(ZE), Positive Small(PS) and Positive Big(PB).
Return
Fig. 5. Flowchart of P&O based control technique Table 4. FLC Set of rules E/CE
NB
NS
ZE
PS
PB
NB
ZE
PB
ZE
NB
NS
NS
PS
ZE
ZE
NB
NS
ZE
ZE
ZE
ZE
ZE
ZE
PS
PS
PB
ZE
ZE
NS
PB
PS
PB
ZE
NB
ZE
FUZZY SET
FUZZY RULES
ERROR
DUTY CYCLE
FUZZIFICATION
DEFUZZIFICATION
CHANGE IN ERROR
Inference mechanism is basically defined by membership functions of FLC which determines the relevance of rules from Table 4. Fig. 7, 8 and 9 represents the input and output membership function of FLC controller. Methods for implication and aggregation are defined as Minimum (min) and Maximum (max) respectively. The Defuzzification method uses centroid for processing.
INFERENCE MECHANISM
Fig. 6. Basic structure of FLC
465
IFAC ACODS 2016 466 February 1-5, 2016. NIT Tiruchirappalli, IndiaRamji Tiwari et al. / IFAC-PapersOnLine 49-1 (2016) 462–467
maximum and constant voltage during the steady state characteristic of wind turbine. To analyse the performance of each MPPT controller power output is compared as shown in Fig. 13.
Fig. 7. Input membership function of error signal
Fig. 10. MPPT controller model for WECS in Simulink Wind Speed Variation
Fig. 8. Input membership function of change in error signal
14 12
Wind Speed (m/s)
10 8 6 4
Fig. 9. Output membership function of duty cycle
2
FLC tracks the sudden change in wind speed more swiftly and precisely. The maximum power point is traced by the controller from the inference system which is mapped by the human knowledge earlier in form of rules. The controller tracks the change in output voltage, current and generates an error signal which is given as an input for fuzzification process, here the input data is converted into a suitable fuzzy linguistic sets using Mamdani method. Then the fuzzy set is processed in inference system where an appropriate fuzzy output is obtained using fuzzy rules. Then the fuzzy output is converted in to the systematic crisp value as a form of duty cycle in defuzzification. Thus the duty cycle is used to control the switching pattern of the converter switch.
0
0
2
4 6 Time (seconds)
8
10
Fig. 11. Wind speed variation 600 P&O
PI
Fuzzy
500
Voltage (Vdc)
400
4. RESULTS AND DISCUSSION
300
200
The detailed implementation of PMSG based WECS using buck converter incorporated with fuzzy logic controller in MATLAB/ Simulink is shown in Fig. 10. To check the tracking ability of MPPT techniques, the wind speed is varied from 3m/s to 12m/s as shown in Fig. 11. Three MPPT approaches in simulated in same varying wind speed conditions. From the analysis it can be stated that the FLC based MPPT technique is most efficient maximum power tracking power when compared with PI and P&O based technique.
100
0
0
2
4 6 Time (Sec)
8
10
Fig. 12. Output voltage for different MPPT techniques Table 5. shows the performance comparison of PI, P&O and FLC based MPPT controllers. The FLC has the high power out when compared with the other two controllers. The output voltage of FLC is equal to the reference DC voltage estimated to be the DC bus voltage of 500V. Whereas, the PI
Fig. 12 shows the output voltage comparison of MPPT controllers. The FLC based MPPT controller gives the 466
IFAC ACODS 2016 February 1-5, 2016. NIT Tiruchirappalli, IndiaRamji Tiwari et al. / IFAC-PapersOnLine 49-1 (2016) 462–467
and P&O controllers have an output voltage of 491V and 494V. Though there is only little difference in the power and voltage but when large wind turbines are connected then the difference will become huge.
Babu, N.R. and Arulmozhivarman, P. (2013). Wind Energy Conversion System- A Techical Review. Journal of Engineering Science and Technology, volume (8), 493 507. Chen, W.L. and Jiang, B.Y. (2015). Harmonic Suppression and Performance Improvement for a Small-scale Gridtied Wind Turbine using Proportional-Resonant Controllers. Electric Power Components and Systems, volume (43), 970 - 981. Baroudi, J.A., Dinavahi, V., and Knight, A.M. (2007). A review of power converter topologies for wind generators. Renewable Energy, volume (32), 2369 2385. Bendib, B., Krim, F., Belmili, H., Almi, M.F., and Boulouma, S. (2014). Advanced Fuzzy MPPT Controller for a Stand-alone PV system. Energy Procedia, volume (50), 383 - 392. Dalala, Z.M., Zahid, Z.U., Yu, W., Cho, Y., and Lai, J. (2013). Design and analysis of an MPPT technique for small-scale wind energy conversion systems. IEEE Transaction on Energy Conversion, volume (28),756 766. Dehghan, S/M., Mohamadian, M., and Varjani, A.Y. (2009). A new variable-speed wind energy conversion system using permanent-magnet synchronous generator and Zsource inverter. IEEE Transaction on Energy Conversion, volume (24),714 - 724. Errami, Y., Ouassaid, M., and Maaroufi, M. (2015). A performance comparison of a nonlinear and a linear control for grid connected PMSG wind energy conversion system,. International Journal of Electrical Power & Energy Systems, volume (68), 180 - 194. Galdi, V., Piccolo, A., and Siano, P. (2008). Designing an adaptive fuzzy controller for maximum wind energy extraction. IEEE Transaction on Energy Conversion, volume (23),559 - 569. Krishna, K.S., and Kumar, K.S. (2015). A review on hybrid energy systems. Renewable and Sustainable Energy Reviews, volume (52), 907 - 916. Martin, A.D. and Vazquez, J.R. (2015). MPPT algorithm comparison in PV systems: P&O, PI, neuro-fuzzy and backstepping controls. International Conference on Industial Technology, 2841 - 2847. Phankong, N., Manmai, S., Bhumkittipich, K., and Nakawiwat, P. (2013). Modeling of grid-connected with permanent magnet synchronous generator (PMSG) using voltage vector control. Energy Procedia, volume (34), 262 - 272. Tripathi, S.M., Tiwari, A.N., and Singh, D. (2015). Gridconnected permanent magnet synchronous generator based wind energy conversion systems: A technology review. Renewable and Sustainable Energy Reviews, volume (51), 1288 - 1305. Simoes, M.G., Bose, B.K., and Spiegel, R.J. (1997). Design and performance evaluation of fuzzy-logic-based variable-speed wind generation system. IEEE Transaction on Industrial Applications, volume (33), 956 - 965.
6000 Fuzzy
P&O
PI
5000
Power (W)
4000
3000
2000
1000
0
0
2
4
6
8
10
Time (sec)
Fig. 13. Power obtained by different MPPT techniques Table 5. Output of various MPPT techniques MPPT Technique/Parameters
PI Controller
P&O Controller
FLC
Power
4992 W
5093 W
5112 W
Voltage
491 V
494 V
500 V
467
5. CONCLUSIONS In this paper, three MPPT controllers like PI, P&O and FLC controller is modelled and the output is compared for wind energy under varying wind speed condition. The performance of each controller is analysed and it is verified that FLC based controller is more efficient and reliable than PI and P&O based controller. The PI controller fails to track the non linearity of the wind speed thus providing poor output power. The P&O based technique is suitable for the condition where the system is stable of with minimum variance. Wind Speed being high non linear the P&O algorithm oscillates around optimal point thus making it difficult to track the next point. The FLC method has a rapid tracking ability. The FLC control method requires previous knowledge of the system so that it can provide an efficient and steady state output. FLC is fast and efficient technique to track the maximum power point in WECS. The results obtained from FLC are superior and efficient than that of PI and P&O technique in terms of stability, faster tracking ability and fluctuations. Hence, it is concluded that the FLC based MPPT method is the best option for stand alone WECS. REFERENCES Alizadeh, M. and Kojori, S.S. (2015). Augmenting effectiveness of control loop of a PMSG based wind energy conversion system by a virtually adaptive PI controller. Energy, volume (91), 610 - 629. 467