KEMBAR78
Multi string PV array | PDF
Real-time Analysis and Simulation of Multi-string
Grid-connected Photovoltaic inverter using FPGA
Presented by-
Satabdy Jena
MTech (Power & Energy Systems)
Department of Electrical Engineering
NIT Meghalaya
Satabdy Jena, Gayadhar Panda and Rangababu Peesapati
Contents
1. Introduction
2. Proposed System Structure
3. Control Algorithm
4. Field Programmable Gate Array
5. Xilinx System Generator
6. Hardware Co-Simulation
7. Conclusion
8. References
3/5/2016 2
1.Introduction
3/5/2016 3
Dearth of energy, trend of rising prices and limited resources.
Abundant availability of solar energy has lead to installation of large-scale Photovoltaic
systems(standalone or grid-connected).
• Grid synchronization
• Grid disturbances
 Frequency
 Voltage
• Industry design standards
2.Proposed System Structure
3/5/2016 4
PV2
MPPT1
MPPT2
PWM1
PWM2
SPWM PLL
Modulating
Signal
Generation
CURRENT
CONTROLLER
VOLTAGE
CONTROLLER
PV1
+-
V*dc
Vdc
3-ф
3-Level
Central
Inverter
Filter
Inductance
Isolation
Transformer
ia,b,c ua,b,c
iq*
id* idq
udq
Utilty grid
abc
dq
ud* uq*
DC/DC
Boost Converter1
DC/DC
Boost Converter2
LfRf
Cf LtRt
ϴ
ϴ
Reference
Signal
Generation
mdq
FPGA Based Control Circuit
MULTI-STRING PV ARRAY
ipv1
ipv2
vpv1
vpv2
LOAD
Fig.1. Structure of proposed system
Terminology Description Terminology Description
Vpv, Ipv PV voltage and current Lt, Rt Transformer inductance and
resistance
Vdc Capacitor voltage Ls, RS Equivalent inductance and
resistance
ua, ub, uc Grid line voltages Lt, Rt Transformer inductance and
resistance
ia, ib, ic Grid line currents id, iq d and q-axes currents
Lf, Rf Filter inductance and
resistance
ud, uq d and q-axes voltages
Lg, Rg Grid inductance and
resistance
mdq Modulating signals
3/5/2016 5
Table 1. System Terminology description
3.Control Algorithm
(a) MPP Tracking
• Incremental Conductance (Inc)
-by comparing instantaneous (
Ipv
Vpv
) and incremental conductance (
dIpv
dVpv
), the MPP voltage of the PV
array is obtained.
dPpv
dVpv
=
d(VpvIpv)
dVpv
= 0 =>
Ipv
Vpv
+
dIpv
dVpv
= 0. (1)
3/5/2016 6
Fig.2. P-V characteristics of PV array
(b) Inner and Outer Control loop design for VSI
-Decoupling Technique
• ud
∗
= Kpd id
∗
− id + Kid (id
∗
− id)dt − ωL 𝑠 𝑖 𝑞 + ud (2)
• uq
∗ = Kpq iq
∗ − iq + Kiq (iq
∗ − iq)dt + ωL 𝑠 𝑖 𝑑 + uq
3/5/2016 7
Switching
State
Switch
configuration
Voltage
+1 1100 Van
0 0110 0
-1 0011 -Van
Fig.3. Outer and Inner Loop Design
Fig.4. Three-level Inverter Structure
Table 2. Switching states for the inverter
(c) Grid Synchronization
-Phase Locked Loop
A feedback system with a PI regulator tracking the phase angle .
3/5/2016 8
𝐺 𝑜𝑙 = (𝐾 𝑃
1+𝑠𝜏
𝑠𝜏
)(
1
1+𝑠𝜏
)(
𝑉𝑚
𝑠
) (3)
-100
-50
0
50
100
Magnitude(dB)
System: sys1
Frequency (rad/s): 314
Magnitude (dB): 0.0119
10
0
10
1
10
2
10
3
10
4
10
5
-180
-150
-120
-90
System: sys1
Frequency (rad/s): 314
Phase (deg): -108
Phase(deg)
Bode Diagram
Frequency (rad/s)
-100
-50
0
50
Magnitude(dB)
10
1
10
2
10
3
10
4
10
5
-180
-135
-90
-45
0
Phase(deg)
Bode Diagram
Frequency (rad/s)
Fig.5. Block diagram of PLL
Fig.6. Bode plot for Open loop system Fig.7. Bode plot for closed loop system
4.Field Programmable Gate Array
3/5/2016 9
 An FPGA is a programmable breadboard for digital circuits-on-chip.
 It consists of:
 Programmable Logic Elements
 Programmable Interconnects
 Custom Circuitry
Fig.8.FPGA Configuration
3/5/2016 10
Fig.9. Structure of each Logic Element
5.Xilinx System Generator
 A DSP-design tool.
 Provides users with Xilinx block-set in Matlab/Simulink.
 Automatic synthesis and place-and-route.
 Bit-accurate and Cycle-accurate.
 Handle VHDL, C or m-code.
 Handle multiple sampling rates.
 Generates a run-time hardware block (JTAG).
3/5/2016 11
• Precision and sampling time.
• Fixed and Floating-point representation.
• Clock period=40 ns.
• 75 MHz.
• Compatibility between different data-point types.
3/5/2016 12
6.Implementation in FPGA
Implementation in FPGA (contd.)
3/5/2016 13
Fig.10. Voltage controller in Xilinx blockset
Fig.11. SPWM in Xilinx blockset
Fig.12. Current controller in Xilinx blockset
Fig.13. PLL in Xilinx blockset
3/5/2016 14
Fig.14. Reference Signal Generation in Xilinx blockset
3/5/2016 15
Fig.15. Experimental Set-up
7.Hardware Co-Simulation
System parameters Value System parameters Value
Series connected modules 7(PV1) (PV2) DC Link voltage 750 V
Parallel connected modules 20(PV1) 30 (PV2) Filter inductance and resistance 1 mH, 3 mΩ
MPP Voltage 54.7 V Grid inductance and resistance 16 mH, 0.8929 Ω
MPP Current 5.58 A Transformer inductance and resistance 0.06 mH, 0.002 mΩ
Open-circuit voltage 64.2 V Inverter switching frequency 5 .5 kHz
Short-circuit current 5.96 A Boost switching frequency 5 kHz
Maximum PV power 85.5 kW Grid voltage 22 kV
DC Link capacitor 5000 μF Grid frequency 50 Hz
3/5/2016 16
Table 2. System Parameters
A test system consisting of a multi-string PV array connected to a 400 V three phase source via
a multilevel inverter has been considered. A local load consuming 10 kW of active power is
connected at the PCC.
3/5/2016 17
Fig.16. Irradiation Fig.17. Power consumed by load and grid
Fig.18. Power generation by the multi-string PV array
Case 1: Varying irradiance condition
3/5/2016 18
0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8
-200
-150
-100
-50
0
50
100
150
200
X= 1.1846
Y= 77.2305
Time (sec)
i_a(A)
X= 0.80466
Y= 187.9812
1 1.05 1.1 1.15 1.2 1.25 1.3 1.35 1.4 1.45 1.5
-400
-300
-200
-100
0
100
200
300
400
i_a/u_a(A/V)
Time (sec)
ia
ua
Fig.19. Grid injected current
Fig.20. Grid voltage and current
Fig.21. Frequency
Fig.22. THD of grid injected currents
Case 2: Dynamic loading condition
3/5/2016 19
Fig.23. Grid injected current Fig.24. Load current
Fig.25. Real power exchange
P=10 kW, 1 kVAr to at t=0.8-1.5 s,P=100 kW, Q=15 kVAr
3/5/2016 20
Fig.26. Reactive power exchange
Fig.27. THD of grid injected currents
3/5/2016 21
Fig.28. Simulation and hardware results of SPWM Fig.29. Simulation and hardware results of PLL
Fig.30. Simulation and hardware results of d-axis current Fig.31. Simulation and hardware results of q-axis current
Fig.32. Simulation and hardware results of reference signal
Design Slice
Registers
(301440)
Slice
LUTs
(150720)
LUT FF
Pairs
Slices
(37680)
IOBS
(600)
Memory
(58,400)
Critical Path delay ns
PLL 139(1%) 962(1%) 123(12%) 283(1%) 59(10%) 33 (1%) 1.695ns
SPWM 0(0%) 57(1%) 0(0%) 20(1%) 92(15%) 0(0%) 3.220ns
Current
Controller
96(1%) 6410(4%) 62(1%) 1947(5%) 257(42%) 0(0%) 0.559ns
Voltage
Controller
48(1%) 1905(1%) 31(1%) 568(1%) 65(10%) 0(0%) 0.552ns
Reference
Signal
2306(1%) 4765 (3%) 2250 (46%) 1374 (3%) 307 (51%) 25 (1%) 9.484 ns
3/5/2016 22
Table 4. Resource Utilization
7.Conclusion
The proposed system was thus implemented in Matlab/Simulink and its real-time performance was
validated using FPGA. The performance of the system was analyzed and the correlation with the
theoretical evaluation was established. It is observed that the critical path delay is 9.5 ns.
3/5/2016 23
8.References
[1] TJ Hammons, “Integrating renewable energy sources into Europeangrids,” International Journal of Electrical Power & Energy Systems,vol.
30,no. 8, pp. 462-475,2008.[2] Fen Lui and Yun Chen,“Design and control for three phase grid connected inverter with LCL filter,” Proceedings
of IEEE Circuits and Systems International conference,pp.1-4,2009.
[3] Mitra Mirhosseini, Josep Pou and Vassilios G. Agelidis,“Single- and Two-Stage Inverter Based Grid Connected Photovoltaic Power Plants
With Ride-Through Capability Under Grid Faults,” IEEE Transcations on Sustainable Energy, vol.6.,No.3.,July 2015.
[4] S.J.Huang and F.S.Pai,“Design and operation of grid connected phototvoltaicsystem with power factor control and active islanding
detection,” IEEE Proceedings on Generation, Transmission, Distribution,vol.48,No.2,2001.
[5] S.J.Huang and F.S.Pai,“Design and operation of grid connected phototvoltaic system with power factor control and active islanding
detection,” IEEE Proceedings on Generation, Transmission, Distribution, vol.48,No.2,2001.
[6] M. Castilla,J.Miret, A.Camacho, Jos´e Matas,De Vicuna and Luis Garc´ıa,“Reduction of current harmonic distortion in three-phase grid-
connected photovoltaic inverters via resonant current control,”IEEE Transactions on Industrial Electronics,VOL.60, No.4, pp.1464-1472,2013.
[7] T.Hornik and Qing-Chang Zhong ,“A Current-Control Strategy for Voltage-Source Inverters in Microgrids Based on and Repetitive
Control,”IEEE Transactions on Power Electronics,vol.26, No.3, pp.943-952,2011.
[8] T Q Zheng ,“Synchronous PI control for three-phase grid-connected photovoltaic inverter,” Proceedings of 2010 Chinese Control and
Decision Conference,2010.
[9] F. Ruza, A. Reyb, J.M. Torreloc, A. Nietob, F.J. Cnovasa,“Real time test benchmark design for photovoltaic grid-connected control systems,”
Electric Power Systems Research, Elseveir, vol.81,no, 4,pp. 907-914,2011.
[10] Mohammed A. Elgendy and Bashar Zahawi,“Assessment of the incremental conductance maximum power point tracking algorithm,” IEEE
Transactions on Sustainable Energy, vol. 4,no. 1, pp. 108-117,2013.
[11] Kaura and V. Blasko,“Operation of a Phase Locked Loop System Under Distorted Utility Conditions,” Eleventh Annual Proceedings of
Applied Power Electronics Conference and Exposition,vol.2,pp.703-708,1996.
[12] L. Hadjidemetriou, E. Kyriakides, and F. Blaabjerg,“A new hybrid PLL for interconnecting Renewable Energy Systems to the grid,” IEEE
Energy Conversion Congress and Exposition,pp.20752082,2012.
[13] Dean Banerjee,“PLL Performance, Simulation and Design,” 4th Edition,Dog Ear Publishing
[14] “System Generator User Guide,”Documentation, XILINX
3/5/2016 24
Thank you
3/5/2016 25

Multi string PV array

  • 1.
    Real-time Analysis andSimulation of Multi-string Grid-connected Photovoltaic inverter using FPGA Presented by- Satabdy Jena MTech (Power & Energy Systems) Department of Electrical Engineering NIT Meghalaya Satabdy Jena, Gayadhar Panda and Rangababu Peesapati
  • 2.
    Contents 1. Introduction 2. ProposedSystem Structure 3. Control Algorithm 4. Field Programmable Gate Array 5. Xilinx System Generator 6. Hardware Co-Simulation 7. Conclusion 8. References 3/5/2016 2
  • 3.
    1.Introduction 3/5/2016 3 Dearth ofenergy, trend of rising prices and limited resources. Abundant availability of solar energy has lead to installation of large-scale Photovoltaic systems(standalone or grid-connected). • Grid synchronization • Grid disturbances  Frequency  Voltage • Industry design standards
  • 4.
    2.Proposed System Structure 3/5/20164 PV2 MPPT1 MPPT2 PWM1 PWM2 SPWM PLL Modulating Signal Generation CURRENT CONTROLLER VOLTAGE CONTROLLER PV1 +- V*dc Vdc 3-ф 3-Level Central Inverter Filter Inductance Isolation Transformer ia,b,c ua,b,c iq* id* idq udq Utilty grid abc dq ud* uq* DC/DC Boost Converter1 DC/DC Boost Converter2 LfRf Cf LtRt ϴ ϴ Reference Signal Generation mdq FPGA Based Control Circuit MULTI-STRING PV ARRAY ipv1 ipv2 vpv1 vpv2 LOAD Fig.1. Structure of proposed system
  • 5.
    Terminology Description TerminologyDescription Vpv, Ipv PV voltage and current Lt, Rt Transformer inductance and resistance Vdc Capacitor voltage Ls, RS Equivalent inductance and resistance ua, ub, uc Grid line voltages Lt, Rt Transformer inductance and resistance ia, ib, ic Grid line currents id, iq d and q-axes currents Lf, Rf Filter inductance and resistance ud, uq d and q-axes voltages Lg, Rg Grid inductance and resistance mdq Modulating signals 3/5/2016 5 Table 1. System Terminology description
  • 6.
    3.Control Algorithm (a) MPPTracking • Incremental Conductance (Inc) -by comparing instantaneous ( Ipv Vpv ) and incremental conductance ( dIpv dVpv ), the MPP voltage of the PV array is obtained. dPpv dVpv = d(VpvIpv) dVpv = 0 => Ipv Vpv + dIpv dVpv = 0. (1) 3/5/2016 6 Fig.2. P-V characteristics of PV array
  • 7.
    (b) Inner andOuter Control loop design for VSI -Decoupling Technique • ud ∗ = Kpd id ∗ − id + Kid (id ∗ − id)dt − ωL 𝑠 𝑖 𝑞 + ud (2) • uq ∗ = Kpq iq ∗ − iq + Kiq (iq ∗ − iq)dt + ωL 𝑠 𝑖 𝑑 + uq 3/5/2016 7 Switching State Switch configuration Voltage +1 1100 Van 0 0110 0 -1 0011 -Van Fig.3. Outer and Inner Loop Design Fig.4. Three-level Inverter Structure Table 2. Switching states for the inverter
  • 8.
    (c) Grid Synchronization -PhaseLocked Loop A feedback system with a PI regulator tracking the phase angle . 3/5/2016 8 𝐺 𝑜𝑙 = (𝐾 𝑃 1+𝑠𝜏 𝑠𝜏 )( 1 1+𝑠𝜏 )( 𝑉𝑚 𝑠 ) (3) -100 -50 0 50 100 Magnitude(dB) System: sys1 Frequency (rad/s): 314 Magnitude (dB): 0.0119 10 0 10 1 10 2 10 3 10 4 10 5 -180 -150 -120 -90 System: sys1 Frequency (rad/s): 314 Phase (deg): -108 Phase(deg) Bode Diagram Frequency (rad/s) -100 -50 0 50 Magnitude(dB) 10 1 10 2 10 3 10 4 10 5 -180 -135 -90 -45 0 Phase(deg) Bode Diagram Frequency (rad/s) Fig.5. Block diagram of PLL Fig.6. Bode plot for Open loop system Fig.7. Bode plot for closed loop system
  • 9.
    4.Field Programmable GateArray 3/5/2016 9  An FPGA is a programmable breadboard for digital circuits-on-chip.  It consists of:  Programmable Logic Elements  Programmable Interconnects  Custom Circuitry Fig.8.FPGA Configuration
  • 10.
    3/5/2016 10 Fig.9. Structureof each Logic Element
  • 11.
    5.Xilinx System Generator A DSP-design tool.  Provides users with Xilinx block-set in Matlab/Simulink.  Automatic synthesis and place-and-route.  Bit-accurate and Cycle-accurate.  Handle VHDL, C or m-code.  Handle multiple sampling rates.  Generates a run-time hardware block (JTAG). 3/5/2016 11
  • 12.
    • Precision andsampling time. • Fixed and Floating-point representation. • Clock period=40 ns. • 75 MHz. • Compatibility between different data-point types. 3/5/2016 12 6.Implementation in FPGA
  • 13.
    Implementation in FPGA(contd.) 3/5/2016 13 Fig.10. Voltage controller in Xilinx blockset Fig.11. SPWM in Xilinx blockset Fig.12. Current controller in Xilinx blockset Fig.13. PLL in Xilinx blockset
  • 14.
    3/5/2016 14 Fig.14. ReferenceSignal Generation in Xilinx blockset
  • 15.
  • 16.
    7.Hardware Co-Simulation System parametersValue System parameters Value Series connected modules 7(PV1) (PV2) DC Link voltage 750 V Parallel connected modules 20(PV1) 30 (PV2) Filter inductance and resistance 1 mH, 3 mΩ MPP Voltage 54.7 V Grid inductance and resistance 16 mH, 0.8929 Ω MPP Current 5.58 A Transformer inductance and resistance 0.06 mH, 0.002 mΩ Open-circuit voltage 64.2 V Inverter switching frequency 5 .5 kHz Short-circuit current 5.96 A Boost switching frequency 5 kHz Maximum PV power 85.5 kW Grid voltage 22 kV DC Link capacitor 5000 μF Grid frequency 50 Hz 3/5/2016 16 Table 2. System Parameters A test system consisting of a multi-string PV array connected to a 400 V three phase source via a multilevel inverter has been considered. A local load consuming 10 kW of active power is connected at the PCC.
  • 17.
    3/5/2016 17 Fig.16. IrradiationFig.17. Power consumed by load and grid Fig.18. Power generation by the multi-string PV array Case 1: Varying irradiance condition
  • 18.
    3/5/2016 18 0.8 0.91 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 -200 -150 -100 -50 0 50 100 150 200 X= 1.1846 Y= 77.2305 Time (sec) i_a(A) X= 0.80466 Y= 187.9812 1 1.05 1.1 1.15 1.2 1.25 1.3 1.35 1.4 1.45 1.5 -400 -300 -200 -100 0 100 200 300 400 i_a/u_a(A/V) Time (sec) ia ua Fig.19. Grid injected current Fig.20. Grid voltage and current Fig.21. Frequency Fig.22. THD of grid injected currents
  • 19.
    Case 2: Dynamicloading condition 3/5/2016 19 Fig.23. Grid injected current Fig.24. Load current Fig.25. Real power exchange P=10 kW, 1 kVAr to at t=0.8-1.5 s,P=100 kW, Q=15 kVAr
  • 20.
    3/5/2016 20 Fig.26. Reactivepower exchange Fig.27. THD of grid injected currents
  • 21.
    3/5/2016 21 Fig.28. Simulationand hardware results of SPWM Fig.29. Simulation and hardware results of PLL Fig.30. Simulation and hardware results of d-axis current Fig.31. Simulation and hardware results of q-axis current Fig.32. Simulation and hardware results of reference signal
  • 22.
    Design Slice Registers (301440) Slice LUTs (150720) LUT FF Pairs Slices (37680) IOBS (600) Memory (58,400) CriticalPath delay ns PLL 139(1%) 962(1%) 123(12%) 283(1%) 59(10%) 33 (1%) 1.695ns SPWM 0(0%) 57(1%) 0(0%) 20(1%) 92(15%) 0(0%) 3.220ns Current Controller 96(1%) 6410(4%) 62(1%) 1947(5%) 257(42%) 0(0%) 0.559ns Voltage Controller 48(1%) 1905(1%) 31(1%) 568(1%) 65(10%) 0(0%) 0.552ns Reference Signal 2306(1%) 4765 (3%) 2250 (46%) 1374 (3%) 307 (51%) 25 (1%) 9.484 ns 3/5/2016 22 Table 4. Resource Utilization
  • 23.
    7.Conclusion The proposed systemwas thus implemented in Matlab/Simulink and its real-time performance was validated using FPGA. The performance of the system was analyzed and the correlation with the theoretical evaluation was established. It is observed that the critical path delay is 9.5 ns. 3/5/2016 23
  • 24.
    8.References [1] TJ Hammons,“Integrating renewable energy sources into Europeangrids,” International Journal of Electrical Power & Energy Systems,vol. 30,no. 8, pp. 462-475,2008.[2] Fen Lui and Yun Chen,“Design and control for three phase grid connected inverter with LCL filter,” Proceedings of IEEE Circuits and Systems International conference,pp.1-4,2009. [3] Mitra Mirhosseini, Josep Pou and Vassilios G. Agelidis,“Single- and Two-Stage Inverter Based Grid Connected Photovoltaic Power Plants With Ride-Through Capability Under Grid Faults,” IEEE Transcations on Sustainable Energy, vol.6.,No.3.,July 2015. [4] S.J.Huang and F.S.Pai,“Design and operation of grid connected phototvoltaicsystem with power factor control and active islanding detection,” IEEE Proceedings on Generation, Transmission, Distribution,vol.48,No.2,2001. [5] S.J.Huang and F.S.Pai,“Design and operation of grid connected phototvoltaic system with power factor control and active islanding detection,” IEEE Proceedings on Generation, Transmission, Distribution, vol.48,No.2,2001. [6] M. Castilla,J.Miret, A.Camacho, Jos´e Matas,De Vicuna and Luis Garc´ıa,“Reduction of current harmonic distortion in three-phase grid- connected photovoltaic inverters via resonant current control,”IEEE Transactions on Industrial Electronics,VOL.60, No.4, pp.1464-1472,2013. [7] T.Hornik and Qing-Chang Zhong ,“A Current-Control Strategy for Voltage-Source Inverters in Microgrids Based on and Repetitive Control,”IEEE Transactions on Power Electronics,vol.26, No.3, pp.943-952,2011. [8] T Q Zheng ,“Synchronous PI control for three-phase grid-connected photovoltaic inverter,” Proceedings of 2010 Chinese Control and Decision Conference,2010. [9] F. Ruza, A. Reyb, J.M. Torreloc, A. Nietob, F.J. Cnovasa,“Real time test benchmark design for photovoltaic grid-connected control systems,” Electric Power Systems Research, Elseveir, vol.81,no, 4,pp. 907-914,2011. [10] Mohammed A. Elgendy and Bashar Zahawi,“Assessment of the incremental conductance maximum power point tracking algorithm,” IEEE Transactions on Sustainable Energy, vol. 4,no. 1, pp. 108-117,2013. [11] Kaura and V. Blasko,“Operation of a Phase Locked Loop System Under Distorted Utility Conditions,” Eleventh Annual Proceedings of Applied Power Electronics Conference and Exposition,vol.2,pp.703-708,1996. [12] L. Hadjidemetriou, E. Kyriakides, and F. Blaabjerg,“A new hybrid PLL for interconnecting Renewable Energy Systems to the grid,” IEEE Energy Conversion Congress and Exposition,pp.20752082,2012. [13] Dean Banerjee,“PLL Performance, Simulation and Design,” 4th Edition,Dog Ear Publishing [14] “System Generator User Guide,”Documentation, XILINX 3/5/2016 24
  • 25.