smes based dfig generator reactive power …

51

Int. J. Elec&Electr.Eng&Telecoms. 2014 Kannan Kaliyappan et al., 2014

SMES BASED DFIG GENERATOR REACTIVEPOWER IMPROVEMENT USING SVPWM

Kannan Kaliyappan1*, S Sutha2 and R Kathirvel1

*Corresponding Author: Kannan Kaliyappan,[email protected]

The Doubly Fed Induction Generator (DFIG) equipped with self-commutated Insulated GateBipolar Transistor (IGBT) is one of the most popular topologies used in wind power systems. Ithas the ability to control active and reactive power independently. The reactive power capabilityis subject to several limitations which change with the operating point. This paper proposes aSuperconducting Magnetic Energy Storage (SMES) based excitation system for DFIG used inwind power generation and used to store energy. The superconducting magnet connects to theDC side of the two converters, which can handle the active power transfer with the rotor of DFIGand the power grid independently. This paper investigates the thermal behavior of the converterusing semiconductor losses. A Space Vector Pulse-Width Modulation (SVPWM) is necessaryat around synchronous speed to increase the maximum permissible rotor current as well asreactive power capability.

Keywords: Doubly Fed Induction Generator (DFIG), Superconducting Magnet Energy Storage(SMES), Space Vector Pulse-Width Modulation (SVPWM), Motor Side Converter(MSC), Load Side Converter (LSC)

INTRODUCTIONWind power has established itself as one ofthe most important renewable energy sourcesover the past decades. The contribution ofwind power to electrical power supply isincreasing rapidly. The worldwide installedcapacity currently reached of about 200 GWand will double toward 2014. According to thegrid code requirement, Wind Turbines (WTs)have to contribute not only to active power

ISSN 2319 – 2518 www.ijeetc.comVol. 3, No. 4, October 2014

© 2014 IJEETC. All Rights Reserved

Int. J. Elec&Electr.Eng&Telecoms. 2014

1 AP/EEE, RVSCET, Dindigul, India.2 ASP/EEE.UCE-Dindigul, India.

generation, but also to the provision of reactivepower into the grid during voltage dips and alsoin the steady-state operation. Modern powerelectronic converters are able to control activeand reactive power independently of oneanother.

The power capability of WTs generator issubject to several limitations resulting from thevoltage, current, and speed. Based on thepower system and generator technology, the

Research Paper

52


power capability of the conventional turbinegenerator is determined by the field heatinglimit, end heating limit, stator heating limit, andsteady-state stability limit (IEEE Guide forOperation and Maintenance of TurbineGenerators, 1990; and Nilsson and Mercuriot,1994). For the Doubly Fed Induction Generator(DFIG)-based WT, the power capabilitylimiting factors involves mechanical power,stator current, and rotor current (Engelhardtet al., 2011). The active-reactive powerdiagram has been presented considering thelimiting factors. The authors have alsoinvestigated that a rotor current duration of aMachine Side Converter (MSC) is necessaryat low frequencies that is around synchronousoperating points to maintain maximum junctiontemperature of IGBTs at about the same levelas at the rated speed. At synchronous speed,the junction temperature of IGBTs could exceedits maximum limit due to higher generatedpower losses which could affect overheatingin the converter and lead to IGBTs failure(Bruns et al., 2009; Jung and Hofmann, 2011;and Wei et al., 2011).

Reduction in the rotor current consequentlyreduces the output power of the MSC. Whenactive power is taken as the priority in thesystem, reduction in reactive power capabilityis needed. To improve the reactive powercapability, the basic idea is to have lowerpower losses of the converter. One approachis by using a converter topology with low powerloss performance. A three-level Neutral PointClamped (NPC) converter seems to havelower power losses compared with a two-levelconverter for the same installed switch power,and has been investigated in Krug et al. (2003)and Alemi and Lee (2011), and it improved

the reactive power capability of DFIG aroundsynchronous speed (Sujod and Erlich, 2013).However, this solution is not preferable by theconverter manufacturer due to higher cost andspace constraint.

A more sophisticated approach toovercome this problem, while maintaining thesize of IGBT module, is using a low switchingfrequency modulation technique.Discontinuous Pulse-Width Modulation(DPWM) type is a well-known modulationtechnique having a low switching frequencyand superior switching loss performance. Ithas been investigated and presented in manyliterature studies (Hava et al., 1998; Chung andSul, 1999; Solomon and Famouri, 2006;Bierhoff et al., 2007; and Wu et al., 2008 and2011). The switching losses of DPWM arelower than Continuous PWM (CPWM) andhave been validated experimentally. At adifferent rotor speed operation mode of DFIG,a different DPWM type showed a differentpower loss performance, which dependent onrotor power factor angle (Sujod and Erlich,2012).

The steady-state reactive powergeneration capability of a typical WT systemusing the DFIG around synchronous operatingpoint using various PWM types. Parametersof the machine and the control system arebased on manufacturer data, and thesimulation results are intended to reflect theexisting reality as accurately as possible. Theactive, reactive power diagram of the DFIGis developed step by step. First, IGBT anddiode power loss model of MSC aredeveloped using MATLAB/Simulinkenvironment. Based on the thermal model ofIGBT and diode, and the simulated power

53


losses data, the junction temperature valuesare calculated. Then, they are used tocharacterize the permissible output current inrelation to rotor frequency. Different PWMtypes are applied for the MSC and its resultsof power losses, junction temperature, outputcurrent, and reactive power capability arecompared. These results help in the intelligentchoice of the appropriate PWM type for DFIGapplication with improvement in the reactivepower capability around synchronousoperating point while maintaining theconverter size.

This paper investigates the steady-statereactive power generation capability of atypical WT system using the DFIG aroundsynchronous operating point using SVPWMtype. Parameters of the machine and thecontrol system are based on manufacturerdata, and the simulation results are intendedto reflect the existing reality as accurately aspossible. The active- reactive power diagramof the DFIG is developed step by step. First,IGBT and diode power loss model of MSC aredeveloped using MATLAB/Simulinkenvironment and converter section MSC andLSC implemented in hardware. The SVPWMcoding are fed in PIC microcontrollerdsPIC30F2010.

DFIG BASED WIND TURBINEThe tip speed ratio is defined as:

R

...(1)

where, , R and v represent turbine rotationalspeed,

Turbine blade radius and the wind velocityrespectively.

In general, the power captured from the windturbine can be written as:

3( , )2m p

AP C ...(2)

where Cp(, ) is the power coefficient, isthe air density, v is the wind speed, R is theblade radius, is the blade pitch angle and is the tip speed ratio. The power curve speedsof a typical wind turbine are shown in Figure1. The value of maximum wind turbine outputpower per unit can be obtained by putting zeropitch angle and Betz limit, when the velocity ofwind turbine is 12 m/s.

It is not suitable for real time application. Inthis proposed paper, maximum power can becaptured at different wind turbine speeds.Considering the generator efficiency G, thetotal power produced by the WG P is:

mP GP ...(3)

DFIGs are widely used in wind generation.It includes a slip ring induction generator, powerelectronic converter, and common DC-linkcapacitor. Power electronic converter, whichencompasses a back-to-back Ac-DC-AcVSC, has two main parts: the Line SideConverter (LSC) and MSC. The LSC isconnected to the stator windings via LC filters,while the MSC is connected to the rotorwindings. A schematic of a DFIG-based WTsystem is shown in Figure 2.

The SMES based excitation system iscomposed of the rotor - side converter, grid-side converter DC chopper, andsuperconducting magnet. In this excitationsystem, the energy storage unit must have thecharacteristic of high energy storage efficiencyand quick power response, and the energy

54


Figure 1: Turbine Output Characteristics (Zero Pitch Angle)

Figure 2: Schematic of a SMES Based DFIG Wind Turbine System

55


storage capacity is not the essential factor. So,the superconducting magnet is selected as theenergy storage unit of the excitation system, itis set at the DC side of the two converters.The DC chopper is utilized to control thecharging and discharging of thesuperconducting magnet. With the coordinatecontrol of the converter, the power transferbetween the super conducting magnet and thegenerator rotor or the power grid can becontrolled effectively. In normally, the maximumpower extraction from wind (Pw) opt isobtained using the indirect speed controltechnique.

The SMES based excitation system fordoubly-fed induction generator can fulfill thefollowing functions:

• In the condition that the energy storagecapacity of superconducting magnet islarge enough, the power transfer during theoperation of Variable Speed ConstantFrequency (VSCF) can be handled by thesuperconducting magnet. This operationmode can smooth the power fluctuation ofwind farms, and improve the power qualityissue induced by the wind power integration.

• During the system operation, the rotor-sideconverter and the grid-side converter ismutually independent. The system intuitivelyparticipates the operation and control of thepower grid, which can improve the securityand stability of the power system.

• In the condition of grid faults, the over currentin the rotor can be suppressed using theenergy storage of SMES. Moreover, theSMES and the grid-side converter areimpervious to the grid fault, which cansupply the power compensation to the grid

and recover the power coupling pointvoltage.

The MSC and LSC are controlled in a dqreference frame. The control structure andcontroller design are explained in detail inIEEE Guide for Operation and Maintenanceof Turbine Generators (1990). The LSCmaintains the dc voltage and provides reactivecurrent support for the optimization of reactivepower sharing between MSC and LSC duringthe steady state. The MSC controls active andreactive power of the WT and follows a trackingcharacteristic to adjust the generator speedfor optimal power generation depending onwind speed. Variable speed operation in theDFIG in accordance with the tracking curveresults in higher power output from the wind.The direction of the power flow through theconverters depends on the operating point ofthe generator. At low active power that is duringthe synchronous and subsynchronous speed,high reactive power capability is needed toallow the generator inject enough reactivecurrent into the grid in the condition when afault occurs at the grid side. This is importantto fulfill the grid code requirement where theWT has to stay connected to the grid duringsuch conditions. The simulation presented inthis paper is carried out with MATLAB/Simulinkwith SimPowerSystems Toolbox in the timedomain based on instantaneous values.

PWM TECHNIQUESThe commonly used MSC is based on a three-phase two level topology. It consists of threebridges with six IGBTs and its freewheelingdiodes. The converter is supplied by a voltagesource DC-link capacitor. This topology is thesimplest VSC configuration as it requires a

56


simple control. The IGBTs is controlled by aSVPWM circuit to achieve variable voltage andvariable frequency output from a fixed DCvoltage. The PWM circuit generates theswitching instants to each of the IGBT’s gates.It could generate different types of PWM whichprovide a different performance level of theconverter, including harmonics, switchinglosses, and output voltage linearity (or DC busutilization). Most common switching techniquesuse PWM which produces less harmonicscompared to the other switching techniques,such as the six-step switching. Several PWMtechniques have been developed and thesecan be generally classified into CPWM andDPWM types Examples of CPWM typesinclude the Sinusoidal PWM (SPWM), Third-Harmonic Injection PWM (THIPWM), andSpace Vector PWM (SVPWM). Theimprovement in reactive power capability couldbe achieved with low switching frequency ofPWM which causes low power losses in theIGBTs. Among the PWM types, SVPWM is thebest choice. A SVPWM generates a referencevoltage vector (VREF) whose angular speedcalculate the desired value of synchronousspeed of the motor and its magnitudedetermines the required voltage (Vs) that willmaintain a constant air gap flux. This referencevoltage vector (VREF) is rotating with respectto a plurality of voltage vectors that each pointout a respective one of the switching states ofan inverter. The concept of space vector isderived from the rotating field of AC machinewhich is used for modulating the inverter outputvoltage. In this modulating technique the threephase quantities can be transformed to theirequivalent 2-phase quantity either insynchronously rotating frame (or) stationary

frame. From this 2-phase component thereference vector (VREF) magnitude can befound and used for modulating the inverteroutput.

POWER LOSSESA power loss model is developed first,because its output data are required for thethermal model. There are different ways tocalculate power losses of IGBT module.However, some complex models require manyIGBT and diode parameters to perform thecalculations. The proposed power lossesestimator using MATLAB/Simulink presentedis more flexible and shows similar results whencompared with other simulation programs suchas PSIM and SEMISEL (Semikron) (Jung andHofmann, 2011). The power losses of theIGBT/diode are divided into the static powerlosses and the non static power losses. Thestatic power losses are the conduction losses(on-state losses) and the blocking losses. Thenon static losses are divided into the switchinglosses (turn-on/off losses) and driving losses.The blocking losses and driving losses areneglected since they account for a smallportion of the overall power losses.

IGBT is a not an ideal switch. Therefore,there will be a voltage drop across the IGBTand its freewheeling diode while it conductscurrent. Switching losses occur when there isa turn-on and a turn-off signal at IGBT's gate.The switching losses of the IGBT and diodeare dependent on load current, DC-link voltage,and switching frequency. When SVPWMpulses are applied to the converter side, itreduces the conduction and switching losses.This creates a path to improve the reactivepower capability.

57


SIMULATION RESULTSAs the MSC currents consist of active andreactive components, reduction of onecomponent may be enough to limit the currentmagnitude by defining the priority in thecontrol design. In this simulation, activecurrent priority is defined in the MSC control.Therefore, the reactive current is reduced first,and the active current is kept unchangedwhen the magnitude exceeds the limit. Toobtain the active-reactive current capabilityof DFIG based WT, the stator and LSCreactive current capabilities have to bedefined. For the stator reactive currentcapability, the stator current limitation due toheating limit, the rated MSC current limitationand the maximum rated stator active powerhave to be considered. In this simulation,SVPWM is applied for LSC in the entirespeed range. Combining the stator and LSC

reactive current capability results in thereactive current capability of DFIG-based WT.

Figure 3 shows the grid voltage and currentwaveform. A grid current value is 0.01 amps.The Rotor speed and angle of DFIG Generator.A Rotor speed (p.u) is 1.2 RPM

Figure 3: Grid Voltage and Current Waveform

Wind Speed 10 m/s

Stator Current 0.08 amps

Machine Side Current 0.015 amps

Load Side Current 0.01 amps

Rotor Speed (p.u) 1.2 RPM

Grid Current 0.01 amps

Generator Input Voltage 460 volts

Generator Input Current 8 amps

Frequency 60 hertz

Output Frequency 60 hertz

Output Voltage 460 volts

Table 1: Typical Current Ratings

58


Figure 4: Motor Side Current (MSC)

Figure 5: Motor Side Six SVPWM Pulses

59


Figure 6: Load Side Current (LSC)

Figure 7: Load Side Six SVPWM Pulses

60


Figure 4 waveform describes the Motorside current. A n MSC current value is 0.015amps. An MSC contains six Insulated GateBipolar Transistor (IGBT). A depends upon thepulse width modulation signals, an IGBT willturn on and turn off at a particular time interval.In this process a generator, AC output voltageconverted to DC voltage (Rectifier) in back toback converter. A PWM pulses are givenbelow.

Figure 6 shows the load side currentwaveform. This is the second procedure ofback to back converter. A load side current is0.01 amps. A LSC contains six Insulated GateBipolar Transistor. A depends upon the pulsewidth modulation signals, an IGBT will turn onand turn off a particular interval. In this processan MSC DC voltage converted into a voltage(Inverter) in back to back converter. In this backto back converter can do the both rectifier and

inverter operation. A Inverter section six typePWM pulses are given below.

Figure 8 shows the reactive power versusactive power characteristic by considering therelationship. All limitations discussed in theprevious sections are considered. It can beshown that around synchronous operatingpoint, derating in the MSC current due to thelimitation in IGBT junction temperatureimposes a limit on the overexcited reactivepower output. In other approaches, the reactivepower capability could be increased by usinga converter of higher rating or by reducing therisk margin, which in most cases is impractical.Sizing the heat sink with a lower maximumtemperature or using a forced external coolingcould also reduce this effect. Another studyalso shows that using the three-level NPCconverter improved the reactive powercapability of DFIG around synchronous speed

Figure 8: Active and Reactive Power Waveform

61


(Bruns et al., 2009). However, all thesetechniques are not preferable by the convertermanufacturer due to higher cost and spaceconstraint.

HARDWARE RESULTSThe control circuit comprises of controller anddriver ICs, the controller has a sequence ofcoding as instructions to produce necessarysignals to trigger the particular solid statedevice and these signals are generatedaccording to the requirements of the circuit byreceiving the necessary input signals from thevoltage and current feedback controllers.These signals are further modulated andprocessed to produce the control signals thenthese signals are given to the driver ICs theseICs produce high or low signal pulses to drivethe particular switch. In this prototype PIC

microcontroller PIC16F877A is used as acontroller and the driver IC IR2112 is used asa driver. These two components control theswitching sequence and drive the switchesaccording to the requirement of the user. ThePIC microcontroller dsPIC30F2010 is used asa controller in this prototype as an aim indesigning the low cost and compact structuredprototype dsPIC microcontroller is used in thisprototype though promises an efficientoperation, it has higher numbers of inputs andoutputs while the requirement of the inputs andoutputs here is very less, the cost of the dsPICis also considerably low. Hence dsPICmicrocontroller is considered to be morereliable and feasible controller for thisparticular prototype. The necessarycomponent for the implementation of thehardware is chosen according to the capacity

Figure 9: Hardware Kit

62


of the circuit and the ratings of each componentare analyzed from the datasheet thenhardware is implemented. The hardware isimplemented such that the components areconnected according to the main circuitdiagram in Figure 9.

The power supply circuit has a step-downtransformer of rating 230 V/12 V when supplyis given to this transformer the power is step-down from 230 V to 12 V and given to the fullwave rectifier. A rectifier is an electrical devicethat converts Alternating Current (AC), whichperiodically reverses direction, to DirectCurrent (DC), which is in only one direction, aprocess known as rectification the termrectifier describes a diode that is being usedto convert AC to DC here the diode IN5408 isused four diodes are used here form a fullwave rectifier four diodes arranged this wayare called a diode bridge or a bridge rectifier.A full-wave rectifier converts the whole of theinput waveform to one of constant polarity(positive or negative) at its output. Full-waverectification converts both polarities of the inputwaveform to DC (direct current), and is moreefficient. The step-down voltage 12 V AC isrectified through the rectifier to 12 V DC full-wave rectification suffice to deliver a form ofDC output, neither produces constant-voltageDC. In order to produce steady DC from arectified AC supply, a smoothing circuit or filteris required.

In its simplest form this can be just areservoir capacitor or smoothing capacitor,placed at the DC output of the rectifier. Therewill still remain an amount of AC ripple voltagewhere the voltage is not completely smoothed.A 330 F and 25 mV capacitor is used assmoothing capacitor, sizing of the capacitor

represents a tradeoff. For a given load, alarger capacitor will reduce ripple but will costmore and will create higher peak currents inthe transformer secondary and in the supplyfeeding it. The circuit has an indicator LED toindicate the supply it is connected through acurrent limiting resistor of 1 K. The 12 V supplyis given to two fixed voltage regulators to gettwo fixed voltages to control circuit; to get afixed voltage of 5 V the supply is given regulatorLM 7805 at terminal 1 and common point isconnected to the negative terminal and we getan output of 5 V. To get a fixed voltage of 12 Vsupply is given to regulator LM 7812 and inputis given to input terminal 1 and the negativeterminal is connected to the common point and12 V supply is received the output terminal.

The control circuit has two parts, onemicrocontroller and four MOSFET driver IC“s.The microcontroller is a programmable devicethat holds a sequence of codes that performsa sequence of tasks and transfers the giveninput as signals to the switches. Themicrocontroller PIC16F877A is used as acontroller and input supply for themicrocontroller is 12 V and the supply is givenbetween the pins 1 and 12 those pins areMCLR/VPP and VSS. The dsPIC30F2010microcontroller has a sequence of program

Machine Side Current 0.015 amps

Load Side Current 0.01 amps

Grid Current 5 amps

Input Voltage 230 volts

Input Current 8 amps

Frequency 50 hertz

Output Frequency 50 hertz

Output Voltage 30 volts

Table 2: Typical Hardware Values

63


codes. These codes help us to develop asequence of instructions which is burnt on themicrocontroller which in turn develops PWMsignals by comparing saw tooth referencesignal with control voltage signals fromfeedback controllers these signals are givento the MOSFET driver IC“s the driver IC“s usedhere is IR 2112.

The pins 20, 21, 26, 27, 28, 29 and 30 arethe pins which generate PWM pulses thesepins are bidirectional pins belong to PORT Dthey are parallel slave ports and they are digitalinput/ output ports which can act as input aswells as output these pins are connected tothe MOSFET driver IC IR2112. Hence PORTDis a bidirectional I/O port or Parallel Slave Port.When interfacing to a microprocessor bus 64.The control signals are also received by thesepins the pulses are fed to the IC IR2112 at the

pins 9 and 12. The IR2112 is a high voltage,high speed power MOSFET and IGBT driverwith independent high and low side referencedoutput channels. Logic inputs are compatiblewith standard CMOS or Low power SchottkyTTL outputs, down to 3.3 V logic. The outputdrivers feature a high pulse current buffer stagedesigned for minimum driver cross-conduction. Propagation delays are matchedto simplify use in high frequency applications.The floating channel can be used to drive anN-channel power MOSFET or IGBT in the highside configuration which operates up to 600volts. The pins 7 and 10 produce High and Lowoutputs respectively, these signals are givenas gate drive to the MOSFET IRF 840. Thesequence of the pulses drives the MOSFETsin a sequence that ensures the clamping ofthe energy stored in a magnetizing inductor to

Figure 10: Mode 1 Operation

64


output through synchronous rectification circuit.Figure 10 shows the mode 1 period normalpwm pulses are applied. So the result of theoutput sine waveform contains oscillation andincrease the junction temperature in convertersection.

Figure 11 shows the mode 2 operation aSVPWM pulse is used in switching operation.In here a low frequency switching is 2.5 kHz. Inthis section a constant 30 voltage is fed to thegrid section. A SVPWM reduce junctiontemperature in converter sections. A SMES isused to store energy. It contains inductor andcapacitor. It stores energy from the MSC andsupply to the LSC. It can act as a boostconverter. The following coding is V-F controlSpace Vector and install in PIC microcontrollerdsPIC30F2010.

A digital storage oscilloscope is anoscilloscope which stores and analyses thesignal digitally rather than using analoguetechniques. It is now the most common typeof oscilloscope in use because of theadvanced trigger, storage, display andmeasurement features which it typicallyprovides. The input analogue signal issampled and then converted into a digitalrecord of the amplitude of the signal at eachsample time. The sampling frequency shouldbe not less than the Nyquist rate to avoidaliasing. These digital values are then turnedback into an analogue signal for display on aCathode Ray Tube (CRT), or transformed asneeded for the various possible types ofoutput-liquid crystal display, chart recorder,plotter or network interface.

Figure 11: Mode 2 Operation

65


CONCLUSIONThis paper has provided a method to analyzethe reactive power capability of the DFIG-based WT system. IGBT/diode power lossesand thermal model have been developedusing MATLAB/Simulink environment. Thefollowing conclusion can be made. A SMES isthat the time delay during charge anddischarge is quite short and a loss of power isless than other storage methods. Nowadaysmostly used this technique in industry’s. ASVPWM type is necessary in a different rotorspeed operating mode, subsynchronous,synchronous, supersynchronous speed to havethe lowest IGBT/diode power losses andincreases in the permissible output current ofMSC and its reduce the third harmonics in theoutput. This helps increase the reactive powercapability of DFIG at such conditions. It alsoreduces the maximum junction temperatureand helps in protecting the IGBT/diode fromover temperature failures and hence couldincrease their failure.

REFERENCES1. Alemi P and Lee D-C (2011), “Power

Loss Comparison in Two- and Three LevelPWM Converters”, in Proc. Int. Conf.Power Electron.-ECCE Asia, May/June,pp. 1452-1457.

2. Bierhoff M, Brandenburg H and Fuchs FW (2007), “An Analysis on Switching LossOptimized PWM Strategies for ThreePhase PWM Voltage Sourceconverters”,in Proc. Annu. Conf. IEEE Ind. Electron.Soc., pp. 1512-1517.

3. Bruns M, Rabelo B and Hofmann W(2009), “Investigation of Doubly-FedInduction Generator Drives at

Synchronous Operating Point in WindTurbines”, in Proc. Eur. Power Electron.Conf., pp. 1-10.

4. Chung D-W and Sul S-K (1999),“Minimum-Loss Strategy for Three-PhasePWM Rectif ier”, IEEE Trans. Ind.Electron., Vol. 46, No. 3, pp. 517-526.

5. Engelhardt S, Erlich I, Feltes C,Kretschmann J and Shewarega F (2011),“Reactive Power Capability of WindTurbines Based on Doubly Fed InductionGenerators”, IEEE Trans. EnergyConvers., Vol. 26, No. 1, pp. 364-372.

6. Global Wind Energy Council (2011),available at http://www.gwec.nt

7. Hava A M, Kerkman R J and Lipo T A(1998), “A High PerformanceGeneralized Discontinuous PWMAlgorithm”, IEEE Trans. Ind. Appl.,Vol. 34, No. 5, pp. 1059-1071.

8. IEEE Guide for Operation andMaintenance of Turbine Generators(1990), IEEE Standard.

9. Jung J and Hofmann W (2011),“Investigation of Thermal Stress in theRotor of Doubly-Fed Induction Generatorsat Synchronous Operating Point”, in Proc.IEEE Int. Electric Mach. Drives Conf.,pp. 896-901.

10. Krug D, Bernet S and Dieckerhoff S(2003), “Comparison of State-of-the-ArtVoltage Source Converter Topologies forMedium Voltage Applications”, in Proc.IAS Annu. Meet. Ind. Appl. Conf., Vol. 1,pp. 168-175.

66


11. Nilsson N E and Mercuriot J (1994),“Synchronous Generator Capability CurveTesting and Evaluation”, IEEE Trans.Power Delivery, Vol. 9, No. 1, pp. 414- 424.

12. Pou J, Osorno D, Zaragoza J, Jaen C andCeballos S (2011), “Power LossesCalculation Methodology to EvaluateInverter Efficiency in Electrical Vehicles”,in Proc. Int. Conf. Compat. PowerElectron., pp. 404-409.

13. Solomon O R and Famouri P (2006), “ANovel Approach for EvaluatingPerformance of Discontinuous Pulse WidthModulation Schemes for Three PhaseVoltage Source Inverter”, in Proc. IEEE Int.Symp. Ind. Electron., pp. 870-875.

14. Sujod M Z and Erlich I (2012), “Control ofDFIG Based Wind Turbine ConverterUsing Continuous and DiscontinuousPWM: Acomparative Study”, in Proc. IFACPower Plant Power Syst. Control Conf.,September, p. 5.

15. Sujod M Z and Erlich I (2013), “ReactivePower Capability of DFIG Based WindTurbine Around Synchronous OperatingPoint with Two-Level and Three-LevelNPC Converter”, in Proc. IEEE PowerTech. Conf., June, p. 6.

16. VDN: Transmission Code (August 2007),Network and System Rules of the

German Transmission SystemOperators, available, at http://www.vde.com/de/fnn/dokumente/seiten/technrichtlinien.aspx

17. Wei L, McGuire J and Lukaszewski R A(2011), “Analysis of PWM FrequencyControl to Improve the Lifetime of PWMInverter”, IEEE Trans. Ind. Appl.,Vol. 47,No. 2, pp. 922-929.

18. Wu Y, Shao S, McMahon R A, Zhan Y andKnight A M (2008), “Power Loss Study ofInverter-Fed Machine Drives UsingDiscontinuous Pulse Width Modulation”,in Proc. IEEE Int. Conf. SustainableEnergy Technol., pp. 1172-1177.

19. Wu Y, Shafi M A, Knight A M and McMahonR A (2011), “Comparison of the Effects ofContinuous and Discontinuous PWMSchemes on Power Losses of Voltage-Sourced Inverters for Induction MotorDrives”, IEEE Trans. Power Electron.,Vol. 26, No. 1, pp. 182-191.

20. Zhao D, Pavan Kumar Hari V S,Gopalaratnam Narayanan S andRajapandian A (2010), “Space-VectorBased Hybrid Pulse Width ModulationTechniques for Reduced HarmonicDistortion and Switching Losses”, IEEETrans. Power Electron., Vol. 25, No. 3,pp. 760-774.

smes based dfig generator reactive power …

Documents