ZhongQC Slides PDF

Integrated Control and Power Electronics
for Energy and Power Systems
Qing-Chang Zhong
Distinguished Lecturer, IEEE Power Electronics Society
Max McGraw Endowed Chair Professor in Energy and Power Engineering
Dept. of Electrical and Computer Engineering
Illinois Institute of Technology
Email: zhongqc@ieee.org
Web: http://mypages.iit.edu/∼qzhong2/
Qing-Chang Zhong (IIT, Chicago, zhongqc@ieee.org) Control and Power Electronics for Energy and Power Systems
Outline
The evolution, philosophy, and current standing

of our research
Sample enabling technologies developed
Smart grid integration
Power quality control
Parallel operation of inverters
Synchronisation strategies
Removal of electrolytic capacitors
Our vision and future research plan
Evolution of our research
Research activities
Power & Energy Systems
Robust Control Theory
& Time-Delay Systems
Process Control
1998 2001 2004 2007 2010 2013

Year
Pushing boundaries and breaking limits Research philosophy / How I think
From hardware to software Focused and thorough research: Deep thinking

From applied to theoretical
Holistic approach: Down to details but keep the
From control to power big picture in mind
Cover many application areas Looking for solutions and problems as well
Looking for hidden links
Current standing
Seamless integration of
control and power electronics.
Associate Editor for

IEEE Trans. on Automatic Control
IEEE Trans. on Control Systems Technology
IEEE Trans. on Power Electronics
IEEE Trans. on Industrial Electronics
IEEE Journal of Emerging and Selected Topics in Power
Electronics
IEEE Access
European Journal of Control
Activities in control theory
Robust control of time-delay systems
(frequency-domain approaches): Solved a
series of fundamental problems in this area:
Projections
J-spectral factorisation
Delay-type Nehari problem
Standard H ∞ problem of
single-delay systems
Unified Smith predictor
Realisation of distributed delays in
controllers
Infinite-dimensional systems: applied the
generic theory of infinite-dimensional
systems to time-delay systems and solved
problems about feedback
stabilizability,approximate controllability,
passivity etc.
Uncertainty and disturbance estimator
(UDE)-based robust control: can be applied
to linear or nonlinear, time-varying or
time-invariant systems with or without
delays; attracted several groups worldwide.
Activities in control applications
Advances in Industrial Control
Control of integral processes with dead-time
Disturbance observer-based control
strategy
Dead-beat response Antonio Visioli
Qing-Chang Zhong
Stability region on the control
parameter space
Achievable specifications etc. 1 Control of
Practical experience with a production line Integral Processes
16 reactors, controlled by 3
industrial computers
with Dead Time
Effective object code > 100 KB
(Intel 8086 assembler)
Continuous Stirred Tank Reactors (CSTR)
Unmanned Aerial Vehicles (UAVs) and
quadrotors
Positioning systems with hysteresis
Activities in energy and power
Sample platform technologies
Smart grid integration: Synchronverters, self-synchronised
synchronverters, STATCOM without a PLL etc.
Parallel operation of inverters: Robust droop control,
universal droop control, harmonic droop control
Removal of phase-locked loops
Power quality in microgrids: C-inverters, bypassing
harmonic current components, H-infinity repetitive control
Provision of a neutral line: Common-mode currents,
removal of isolating transformers
Synchronisation: sinusoid-locked loops
Active capacitors
Applications
Wind power
Hybrid electric vehicles
High-speed trains
Single-phase to three-phase conversion
Battery
Qing-Chang Zhong (IIT, Chicago, cell balancing Control and Power Electronics for Energy and Power Systems
zhongqc@ieee.org)
Completely Autonomous
Power Systems (CAPS)
Next Generation Smart Grids
Qing-Chang Zhong
Nuclear Plants
Hydro Plants
Coal Plants SG
SG SG
SV SV
Transmission
Motors and Electric Vehicles
Distribution
SV
SV
SV SV
Lightings
Lighting
SV SV Solar Farms
HVDC
Electronic Apparatus Wind Farms
(to appear 2015)
IEEE Power Electronics Society Distinguished Lecturer, 2014-2015

FREEDM Systems Center Scientific Advisor, NCSU, 2014
Rolls-Royce UTP Board Member in Power Electronics, 2013-
Industrial Advisory Board
Chaired by Kevin Daffey, Global Head of

Rolls-Royce Electrical Power and Control Systems
Members:
Facilities
Largest OPAL RT Real Time Digital Simulator in EU and North
America for control, power electronics and power systems
A microgrid consisting of one 80kVA gen-set, three 200 kVA
inverters (max 400 kVA), which is highly reconfigurable for
fundamental applied research in control and energy management
strategies.
A 60kW Chroma grid simulator
OP4500
Bench 5 5-Gbits optical fiber pair
EXT CNTR
OP4500
Bench 4
Bench 3 OP4500
OP5607
VIRTEX 7
Bench 2 OP4500 EXT CNTR
PCI Express 4x
Bench 1
OP4500
Outline
The evolution, philosophy, and current standing of our research

Power systems challenges
Synchronverters: Inverters that mimic SG
Self-synchronised synchronverters: No more PLLs
Self-synchronised PWM rectifiers
Power systems
“. . . the greatest engineering achievement of

the 20th century.” (National Academy of Engineering ’2000)
“. . . the largest and most complex machine

engineered by humankind.” (P. Kundur ’94)
Power systems
“. . . the greatest engineering achievement of

the 20th century.” (National Academy of Engineering ’2000)
“. . . the largest and most complex machine

engineered by humankind.” (P. Kundur ’94)
Evolution of power systems
Centralised Generation
Distributed
Generation
Generation Smart Grid
Power Plant
500 kV Transmission Extra-High-Voltage Substation
(500/230 kV)
Commercial/ 230 kV
Industrial Transmission Transmission
Customer System
Distribution Substation
Urban (69/12 kV) 69 kV Sub-transmission
Customers Distribution
System
(12kV)
High-Voltage Substation
Distribution Line (230/69 kV)
Overhead
Distribution
Underground Cable Transformer
To Other
High-Voltage
Residential Residential
Underground Substations
Customer Customer
Distribution Transfomer
DOE: Smart Grid System Report
http://en.wikipedia.org/wiki/Electrical_grid.
Karady G and Holbert K 2004
Distributed
Generation
Power Plant
(500/230 kV)
Commercial/ 230 kV
Customer System
System
(12kV)
Overhead
Distribution
To Other
High-Voltage
Customer Customer
Distributed
Generation
Power Plant
(500/230 kV)
Commercial/ 230 kV
Customer System
System
(12kV)
Overhead
Distribution
To Other
High-Voltage
Customer Customer
What’s next?
Next-Generation Smart Grids:
Power Electronics Based Autonomous Power Systems
What’s next?
What’s next?
Challenges being faced by power systems
Ageing infrastructure (mostly over 100 years old)

Faults
Blackouts
Top 10 blackouts in the history
ͷǤǤǡʹͲͲ͵
ȋǡǡ
ͷǤǤǡʹͲͲ͵ ǡǡȌ
ȋƬȌ ͷͷ
ͷͷ ͹ǤǤǡͳͻ͸ͷ
ȋƬȌ
͵Ͳ
ͳǤ ǡʹͲͳʹ

͸͹Ͳ
ʹǤǤǡʹͲͲͷ
ͶǤǤǡʹͲͲͻ ǦȋȌ
Ƭ ͳͲͲ
ͺ͹
͵ǤǤǡͳͻͻͻ

ͻ͹
Generated from the data at http://en.wikipedia.org/wiki/List_of_power_outages

Faults
Blackouts
Fast growth of electricity consumption
Civilisation: > 30 cities with 10+ million people by
2020 (Wiki)
Digital economy: Data centres to consume 20%
electricity in the USA by 2030 (EPRI)
86 Energy Consumption Estimates by End-Use Sector, 1949-2010
Residential, By Major Source Commercial, By Major Source

12 12
10 10
Electrical
Electrical
Losses¹
Losses¹
8 8
Quadrillion Btu
Quadrillion Btu
Renewable Energy Natural
Gas
6 6
Natural Gas
Electricity²
Renewable
4 4 Energy
Electricity²
2 Petroleum 2
Petroleum
Coal Coal
0 0
1950 1960 1970 1980 1990 2000 2010 1950 1960 1970 1980 1990 2000 2010
Industrial, By Major Source Transportation, By Major Source

12 30
Natural Gas
Petroleum
10 25
Petroleum
8 20
Quadrillion Btu
Quadrillion Btu
Electrical
6 Losses¹ 15
4 Electricity² 10
Coal Renewable
2 5 Energy
Renewable Energy
Natural Gas
0 0
1950 1960 1970 1980 1990 2000 2010 1950 1960 1970 1980 1990 2000 2010
http://www.eia.gov/totalenergy/data/annual/archive/038410.pdf.

Faults
Blackouts
2020 (Wiki)
Demand of high energy efficiency
Large-scale utilisation of renewable energy, EVs
and energy storage systems etc.

Faults
Blackouts
2020 (Wiki)

Faults
Blackouts
2020 (Wiki)
How to address the challenges?
Upgrading the system, e.g. by introducing

Phase Measurement Units (PMU)
Wide-Area Monitoring Systems (WAMS)
PMUs and WAMS in China
Northeast
Shenyang
North
Xinjiang
Beijing
Northwest Zhengehou
Jiangsu
Tibet
Henan
Nanjing Shanghai
Chuan–Yu
Three Gorge
Power Plant East
Guizhou Central
PMU at Substation Yunnan
PMU at Power Plant
Kunming South Guangzhou
WAMS Central Station Guangdong
X.R. Xie, Y.Z. Xin, J.Y. Xiao, J.T. Wu, and Y.D. Han. WAMS applications in Chinese power systems. IEEE Power & Energy Magazine. Vol. 4, No. 1, pp. 54-63, Jan.-Feb. 2006.

Strengthening the system, e.g. by introducing

HVDC and HVAC links

Strengthening the system, e.g. by introducing

HVDC and HVAC links
The UK national grid
Adapted from Department of Energy & Climate Change, UK, Our electricity transmission network: A vision for 2020.
The mainland Chinese power system
800kV 6400 MW, 2015 800kV 6400 MW, 2018
800kV 6400 MW, 2018 3000 MW, 2009
Heilongjiang
800kV 6400 MW, 2011 800kV 6400

3000 MW, MW, 2016
2011 NEPG
800kV 6400 MW, Jilin
2015 NCPG
Xinjiang
Liaoning
3000 MW, 2010
Janjin
800kV 6400
Inner Mongolia
MW, 2014 Gansu
Beijing
1500 MW, 2008
Hebel
Shanxi
NWPG Ningula Shandong 800kV 6400 MW, 2015

Qinghai
Henan
Shianxi
ECPG
3000 MW, 2011
Jiangsu
800kV 6400 Shanghai
Sichuan & Hubei
MW, 2016 Xizang Anhui
Chongqing
CCPG 1200 MW, 2011
Zheijang
Jiangxi
800kV 6400 MW, 2012 1000 MW, 2012
Guizhou Hunan
800kV 6400 MW, 2019 Fujian
3000 MW, 2016
SCPG
800kV 6400 MW, 2018 Guangdong
Yunnan
Guanqxi Taiwan
800kV 5000-6000 MW, 2015 800kV 5000 MW, 2009

Bangkok 3000 MW, 2013 Hainan
X-P. Zhang, C. Rehtanz and Y. Song., ‘‘A grid for tomorrow,’’ Power Engineer , vol.20, no.5, pp.22-27, Oct.-Nov. 2006.
The US power system
Source: http://views.cira.colostate.edu/fed/Egrid/.
These actions are all important and effective.
But are we doing enough?
Let’s go one step back and recall the challenges:
Ageing infrastructure
and ESS etc.
What do these challenges really mean/what is
fundamental behind these challenges?/What will
these make future power systems look like?
⇓
Power electronics-based.
and ESS etc.
⇓
and ESS etc.
⇓
and ESS etc.
⇓
and ESS etc.
⇓
Fundamental challenge
Future power systems will be
power electronics-based,
with a huge number of heterogeneous players.
It is less of a power problem but more of a systems problem
How to guarantee system stability?
How to organically expand power systems without
jeopardising stability?
No longer able to heavily rely on communication networks
It is fine for monitoring, information systems and high-level
functions.
But for low-level control, this will cause a great concern of
reliability.
No longer manageable with human interaction
functions.
reliability.
functions.
reliability.
functions.
reliability.

!

"

#$%&'()$% *(+$%, -./01

Is there ONE simple mechanism to enable organic growth and

autonomous operation of power systems?
Is it possible for new add-ons to play an equal role as conventional
generators in regulating the system stability?
Is it possible for the majority of loads to play the same role too?
If yes, can these happen regardless of size and capacity?
OM;=3PE5@F8=>F
654789
2345 654789 N<F5@4= 654789
A@;><K L>Q@;
63?@= 654789
:7;<98=>45 63?@= 654789
A383=9 E5@F8=>F R@D>F5@9
E7@=CM L83=4C@ LM98@K9

B>CD8>7C
STUVWXYZ[VW \Z]VW^ _`abc
E5@F8=37>F GHH4=48<9 I>7; J4=K9 L354= J4=K9

moewgrxjopx
hgfijk
defg hgfijk nxgrfo hgfijk
srmpn} ~prm
heqro hgfijk
limnkjopfg heqro hgfijk
sejeok wgrxjopx rvpxgrk
wirou ~jeofur ~kjr}k

tpuvjpiu

wgrxjoeipx yzzfofjnk {pim |fo}k ~egfo |fo}k

³±¡´©¤ª¡¢ª

² ª¤¡
¥¤¢ ¯ °¢µ¤
£¤¡
¡¢ £¤¡
¥¡ ©¤ª¡¢ª ¶¤¨¢ª¤
©¤¡§± °¡§¤ °±¤¯

¦¢§¨¢§
·¸¹º»¼½¾¿º» À¾Áº»Â ÃÄÅÆÇ
©¤ª¡¢ª «¬¬¡ ¢ ®¡¯ °¡ ®¡¯

åãÑÓÉæÛËÖÜÎÓÔÜ
ÌËÊÍÎÏ
ÈÉÊË ÌËÊÍÎÏ äÒÜËÖÊÓ ÌËÊÍÎÏ
×ÖÑÔÒá âÔçÖÑ
ÌÉÕÖÓ ÌËÊÍÎÏ
ÐÍÑÒÏÎÓÔÊË ÌÉÕÖÓ ÌËÊÍÎÏ
×ÉÎÉÓÏ ÛËÖÜÎÓÔÜ èÖÚÔÜËÖÏ
ÛÍÖÓÙã âÎÉÓÊÙÖ âãÏÎÖáÏ

ØÔÙÚÎÔÍÙ
éêëìíîïðñìí òðóìíô õö÷øù
ÛËÖÜÎÓÉÍÔÜ ÝÞÞÊÓÊÎÒÏ ßÔÍÑ àÊÓáÏ âÉËÊÓ àÊÓáÏ

Conventional electricity generation
(θ = 0 )
Rotor field axis
Rs , L
Rotation
M M
N
Field voltage
Rs , L Rs , L
Conventionally, the generation of electricity is dominated by synchronous

generators.
Why synchronous generators (SG)?
The real power P flowing out of an SG is
VE
P= sin (θ − θg )
Xs
where E and V are the RMS values of the generated voltage and the
terminal voltage. Moreover, an SG obeys the swing equation
J θ̈ = Tm − Te − Dp θ̇
and a power system can be regarded as a system of coupled oscillators.

Because of the sin term, an SG can synchronise with the grid or an SG.
The underlying principle that holds a power system is

the synchronisation mechanism of SG
We are going to adopt this to facilitate smart grid integration.
VE
Xs


VE
Xs


New add-ons of generation
Renewable energy
Wind
Solar
Tide
Wave etc.
Electric vehicles
Energy storage systems
It is a real mess.
Is there anything in common?
Renewable energy
Wind
Solar
Tide
Wave etc.
Electric vehicles
It is a real mess.
Renewable energy
Wind
Solar
Tide
Wave etc.
Electric vehicles
It is a real mess.
Inverters: —
Common devices for smart grid integration
Rs, Ls
ia va
ea
VDC ib vb
eb
ic vc
ec
C
-
Are we able to make inverters have the vital synchronisation mechanism?
Inverters: —
Common devices for smart grid integration
Rs, Ls
ia va
ea
VDC ib vb
eb
ic vc
ec
C
-
Are we able to make inverters have the vital synchronisation mechanism?
Our solution: Synchronverters
Synchronverters are inverters

that mimic synchronous
generators (SG).
Dynamically behave like SG and
hence possess the inherent
synchronisation mechanism.
Can operate autonomously
without communication.
The basic idea
Taking the mathematical model of a synchronous generator as the
core of the controller for an inverter.
Converting the generated voltage e to PWM signals to drive the
switches so that the average values of ea , eb and ec over a
switching period is equal to e.
Feeding back the phase current i to the mathematical model as
the stator current.
+
Dp
Circuit
Ls , R s va Lg , R g Breaker
ia vga
- ea
Tm 1 θ& 1 θ vb
VDC ib vgb
Js s eb
- vc
ec ic vgc
Te
Formulas
Q of Te, Q, e C
e
-
Mf if i
The basic idea
the stator current.
+
Dp
Circuit
ia vga
- ea
Tm 1 θ& 1 θ vb
VDC ib vgb
Js s eb
- vc
ec ic vgc
Te
Formulas
Q of Te, Q, e C
e
-
Mf if i
The basic idea
the stator current.
+
Dp
Circuit
ia vga
- ea
Tm 1 θ& 1 θ vb
VDC ib vgb
Js s eb
- vc
ec ic vgc
Te
Formulas
Q of Te, Q, e C
e
-
Mf if i
The complete controller
Dp θ&r
- PLL
Reset θg
Pset p Tm 1 θ& 1 θ
θ&n Js s
-
θc
From\to the power part

Te
Formulas
Q of Te, Q, e PWM
e generation
- Mf if
Qset 1 i
Ks
Dq
- Amplitude v fb
vm detection
vr
Four parameters Frequency regulation via frequency droop control
No conventional PI control Voltage regulation via voltage droop control

No dq transformation etc. Real power and reactive power control
Experimental results
Frequency regulation
1060.0 50.15
1040.0 50.10
Grid frequency [Hz]

1020.0 50.05
P [W]
1000.0 50.00
980.0 49.95
960.0 49.90
940.0 49.85
0 100 200 300 400 500 600 700
Time [sec]
So, all the generators can have the vital
synchronisation mechanism and take part in the grid
regulation.
How about the loads?
So, all the generators can have the vital
synchronisation mechanism and take part in the grid
regulation.
How about the loads?
Load types
Many different types of loads exist in a power system:
Home appliances
Lighting devices
Elevators
Computers/servers
Air-conditioners
Machines
...
Load types
Many different types of loads exist in a power system:
Home appliances
Lighting devices
Elevators
Computers/servers
Air-conditioners
Machines
...
Internet
>10%
Motor
50% Lighting
20%
Other
20%
(EPRI)
The majority of loads (will) have a front-end rectifier because

Motors are often equipped with AC drives to improve efficiency
and performance
Internet devices consume DC electricity
Light bulbs are being replaced with energy-efficient devices, e.g.
LED
If these loads (rectifiers) are made to behave like synchronous motors

then the majority of loads in a power system will have the
synchronisation mechanism we are looking for.
Internet
>10%
Motor
50% Lighting
20%
Other
20%
(EPRI)

and performance
LED

Internet
>10%
Motor
50% Lighting
20%
Other
20%
(EPRI)

and performance
LED

Internet
>10%
Motor
50% Lighting
20%
Other
20%
(EPRI)

and performance
LED

Internet
>10%
Motor
50% Lighting
20%
Other
20%
(EPRI)

and performance
LED

Running rectifiers as synchronous motors
Angular frequency
Dp v
Reset STA
Vo
Vref Ki Tm 1
θ& 1 ș
Kp +
s Js s
șc
Te
Formulas
To/from the power part

of Te, Q
Q and e e PWM
generation
v
Qref 1 Mfif i
Ks
320 150
318 θ˙g θ̇
Speed [rad/s]
100
P [W]
316
50
314
312 0
310 −50
0 10 20 30 40 50 0 10 20 30 40 50
Time [s] Time [s]
60 50
0
40
Vdc [V]
Q [Var]
−50
20
−100
0 −150
0 10 20 30 40 50 0 10 20 30 40 50
Time [s] Time [s]
Voltage/Current[V/A]
30 va ia
20 1) Circuit breaker turned on at t=2s;
2) Load R=50Ω connected at t=4s;
10 3) PWM signals enabled at t=10s with
0 Vref =40 V and the Q-loop disabled;
4) The Q-loop enabled at t=20s;
−10 5) Vref changed to 50 V at t=30s;
−20 6) The load changed to R=30Ω at t=41s.
38.9 38.92 38.94 38.96 38.98 39
Time [s]
So, we have made
inverters to have the synchronisation mechanism
of synchronous generators
the majority of loads to have the same
synchronisation mechanism
Is there any problem left?
— There is a dedicated synchronisation unit.
So, we have made
So, we have made
Inverters Problems with
Dp
-
θ&r
θg
PLL
dedicated
Reset
Pset p Tm 1 θ& 1 θ synchronisation

θ&n Js s
-
θc units (PLL etc.)

Te
Formulas
Q of Te, Q, e PWM Fighting with each
e generation
Qset
-
1 Mf if i other
Ks
Dq
-
vm
Amplitude
detection
v fb Causing instability
vr
Degraded
performance
Rectifiers
Angular frequency
Vo
Dp
Reset STA
v
?
Vref
Kp +
Ki Tm 1
θ& 1 ș Is it possible to get rid of
s Js s
șc the dedicated
Te
Formulas synchronisation unit,
of Te, Q
Q and e e PWM
generation
although it is believed to
Qref Mfif i
v be a must-have
1
Ks component?
Dp
-
θ&r
θg
PLL
dedicated
Reset

θ&n Js s
-

Te
Formulas
e generation
Qset
-
1 Mf if i other
Ks
Dq
-
vm
Amplitude
detection
vr
Degraded
performance
Rectifiers
Angular frequency
Vo
Dp
Reset STA
v
?
Vref
Kp +
Ki Tm 1
s Js s
șc the dedicated
Te
of Te, Q
Q and e e PWM
generation
Qref Mfif i
v be a must-have
1
Ks component?
Dp
-
θ&r
θg
PLL
dedicated
Reset

θ&n Js s
-

Te
Formulas
e generation
Qset
-
1 Mf if i other
Ks
Dq
-
vm
Amplitude
detection
vr
Degraded
performance
Rectifiers
Angular frequency
Vo
Dp
Reset STA
v
?
Vref
Kp +
Ki Tm 1
s Js s
șc the dedicated
Te
of Te, Q
Q and e e PWM
generation
Qref Mfif i
v be a must-have
1
Ks component?
Dp
-
θ&r
θg
PLL
dedicated
Reset

θ&n Js s
-

Te
Formulas
e generation
Qset
-
1 Mf if i other
Ks
Dq
-
vm
Amplitude
detection
vr
Degraded
performance
Rectifiers
Angular frequency
Vo
Dp
Reset STA
v
?
Vref
Kp +
Ki Tm 1
s Js s
șc the dedicated
Te
of Te, Q
Q and e e PWM
generation
Qref Mfif i
v be a must-have
1
Ks component?
Self-synchronised synchronverters
SP
.
PI θn
.
θr
∆T −
Dp
Pset 1 Tm − 1 θ
.
1 θ
θ&n
− Js s
Te
Formulas
Q of e, Q, Te e
Qset − i is 1 vg
1 Mfif SC
1 Ls+R
Ks
2 ig
SQ
Dq Vn
−
Vg
A mechanism is introduced to generate the reference frequency
A mechanism is introduced to synchronise with the grid before connection
50.2 100
Synchronverter frequency
80
Frequency [Hz]
50.1 60
P [W]
40
50 20
0 Real power
49.9 −20
0 10 20 30 40 50 60 0 10 20 30 40 50 60
Time [s] Time [s]
50.2 80
Grid frequency measured by a PLL
60
Frequency [Hz]
50.1 40
Q [Var]
20
50 0
−20 Reactive power
49.9
0 10 20 30 40 50 60 0 10 20 30 40 50 60
Time [s] Time [s]
Kif ∆θûg θún
K pf +
s
∆θý θúr
Vo
Vref K iv Tm 1 θü 1 ș
K pv +
s Js s
șc
Te Formulas
of Te, Q
Q and e e
is 1 v
S
Qref 1 Mfif Ls + Rs
− v 1
Ks 2
i
A mechanism is introduced to generate the reference frequency
A mechanism is introduced to synchronise with the grid before connection
316 100
θ̇g
Frequency [rad/s]
315.5 θ̇ 75
315 50
P [W]
314.5 25
314 0
313.5 −25
313 −50
0 10 20 30 40 50 0 10 20 30 40 50
60 Time [s] 50 Time [s]
50 25
40 0
Q [Var]
Vo [V]
30 −25
20 −50
10 −75
0 −100
0 10 20 30 40 50 0 5 10 15 20 25 30 35 40 45 50
Time [s] Time [s]
30
Voltage/Current[V/A]
va ia
20 1) Circuit breaker turned on at t=3s;
2) Load R=50Ω connected at t=5s;
10
3) PWM signals enabled at t=10s with
0 Vref =40 V and the Q-loop disabled;
4) The Q-loop enabled at t=20s;
−10 5) Vref changed to 50 V at t=31s;
6) The load changed to R=30Ω at t=42s.
−20
29.8 29.85 29.9 29.95 30
Time [s]
So, we have indeed made it.
All new add-ons of generation can behave like

synchronous generators.
The majority of loads can behave like
synchronous motors.
They all possess the inherent synchronisation
mechanism, without a dedicated synchronisation
unit, so they are naturally held together.

synchronous motors.

synchronous motors.

synchronous motors.
Architecture for next-generation smart grid
In the news
About IEEE Smart Grid Conferences Publications Standards Newsletter Resources
Search Smart Grid Share this | Join the Smart Grid Technical Community
Ge t Sm a r t Grid N e w s U pda
Twitter | RSS | Facebook
IEEE Smart Grid September 2013 How to Achieve Completely Autonomous Power in the Next Generation of Smart Grids
Home News Projects Business T&D End Use Technologies
How to Achieve Completely Autonomous

Power in the Next Generation of Smart Grids
Written by Qing-Chang Zhong
The paradigm of future power systems described here offers a method of standardizing the interface of all electrical supplies,
including conventional power plants and new add-ons, such as wind/solar farms, electrical vehicles and energy storage
systems, and a majority of loads with the transmission and distribution networks, by exploiting the synchronisation principle of
synchronous machines. This model opens the prospect of achieving completely autonomous operation of power systems. Smart Grid Technology
Due to civilisation and economic development, demand for electricity is constantly growing, leading directly to supply issues
What's next after self-healing? How about a
self-directed smart grid
Sep 17, 2013 Talk Back Free Alerts More On This Topic
1
Quick Take: Back in 2010, I told you about an
Electric Power Research Institute (EPRI)
paper that predicted the demise of centralized
control for the smart grid. EPRI described it as
the move from "command and control" to "inform
and motivate." That same year I pointed out S&C
Electric's vision for distributed intelligence
that would coordinate with centralized control.
Then in June of 2013, I suggested that transactive energy might be the "language" we
should use for that coordination. Transactive energy, among other things, sends "value
signals" so that devices can "negotiate" how much power they will use and how much they will
pay for it.
Now the IEEE is reporting on an alternative approach for a self-directed,

semi-autonomous grid. Qing-Chang Zhong, a professor in the Department of Automatic
Control and Systems Engineering at the University of Sheffield in the U.K., proposes a scheme
that could accommodate millions of active players. Click the link above to review the post, or
skim the short summary below. - By Jesse Berst
The number of devices connected to the grid is exploding. That number will rise even more
At conferences: 12 plenary talks
IEEE PEDG, USA, July 2013
Annual Conference of Chinese Universities in Power Systems and Automation,
China, Nov. 2013
3rd Int. Conf. on Advances in Control and Optimization of Dynamical
Systems, IIT Kanpur, India, Mar. 2014.
IEEE GreenTech, USA, April 2014
Annual Conference of Chinese Universities in Power Electronics, Wuhan,
China, April 2014
IEEE International Energy Conference, Croatia, May 2014.
Delta Power Electronics Forum, Suzhou, China, May 2014.
The 26th Chinese Control and Decision Conference (CCDC), Changsha,
China, May 2014.
OPAL RT Annual Conference, Montreal, Canada June 2014.
20th International Conference on Automation and Computing (ICAC2014),
Cranfield, UK, Sept., 2014
The 1st Indian Systems and Controls Conference, Chennai, India, Jan. 2015.
The 5th Int. Conference on Power Engineering, Energy and Electrical Drives,
Riga, Latvia, May 2015.
Outline

Principles to improve power quality
C-inverters
Bypassing harmonic currents
Power quality issues
Harmonics is becoming an important problem because future
power systems are becoming converters-dominated.
Harmonics can cause overheating, increased losses, decreased
power capacity, neutral line overloading, distorted voltage and
current waveforms.
Converters are sources of voltage harmonics:
PWM and switching effects: can be handled with LC filters,
PWM strategies etc.
Voltage reference vr
Voltage drop on the output impedance
S = P + jQ
Z o ∠θ
Vo ∠0 o
~ E∠δ
( vo )
( vr )
vo = vr − Zo (s) · i ,
Principles to improve power quality
S = P + jQ
Z o ∠θ
Vo ∠0 o
~ E∠δ
( vo )
( vr )
vo = vr − Zo (s) · i ,
to make the reference voltage vr provide the right amount of

harmonic voltages to compensate the harmonic voltage dropped
on the output impedance (harmonic droop control).
to bypass the harmonic components of load current so that the
current i is clean; and
to keep vr clean and maintain a small output impedance Zo over
the frequency range of the major harmonic current components
(C-inverters);
In addition to these, other control strategies, e.g. H ∞ repetitive
control, can be adopted.
C-inverters
The output impedance of an inverter is normally inductive (L-inverters)
but can be made capacitive (C-inverters) via the controller below.
+ VDC - i
uf i vo
L CB 1
u
PWM IGBT sCo
C AC bus
bridge
u - vr
60
C−inverter
40
L−inverter
Magnitude (dB)
20
−20
−40
THD of vo (%)
30 vo 10
i [A], vo [V]
i 8
20
10 6
0 4
−10 2
−20 0
0.9 0.91 0.92 0.93 0.94 0.95 0.96 0.9 0.91 0.92 0.93 0.94 0.95 0.96
Time [s] Time [s]
(a) C-inverter with Co = 4500µF
THD of vo (%)
30 vo 10
i [A], vo [V]
i 8
20
10 6
0 4
−10 2
−20 0
0.9 0.91 0.92 0.93 0.94 0.95 0.96 0.9 0.91 0.92 0.93 0.94 0.95 0.96
Time [s] Time [s]
(b) C-inverter with Co = 3100µF
THD of vo (%)
30 vo 10
i [A], vo [V]
i 8
20
10 6
0 4
−10 2
−20 0
0.9 0.91 0.92 0.93 0.94 0.95 0.96 0.9 0.91 0.92 0.93 0.94 0.95 0.96
Time [s] Time [s]
(c) R-inverter with Ki = 0.4
THD of vo (%)
30 vo 10
i [A], vo [V]
i 8
20
10 6
0 4
−10 2
−20 0
0.9 0.91 0.92 0.93 0.94 0.95 0.96 0.9 0.91 0.92 0.93 0.94 0.95 0.96
Time [s] Time [s]
(d) L-inverter
+ VDC - vo
i
uf i vo
L -
CB
u Ki KR(s)
PWM IGBT
C AC bus
bridge
u - vr
The voltage controller introduced above is equivalent

to putting a parallel branch with C .
+ VDC -
uf iL i io vo
L CB
sL + Ki
u
PWM IGBT Admittance K R
C
Zo (s) = .
H-bridge of resonant K
i
AC bus 1 + KR (s)
filters
Then KR can be chosen as a series resonant circuit
at given harmonic frequencies to short-circuit
harmonic current components.
2ξ h ω s
KR (s) = Σ × Kh .
h=3,5,··· s 2 + 2ξ h ω s + (h ω )2
16
14 1+K3
Gain (abs)
12
10 1+K5
8
6
4 1+K7
2
0
90
Phase (deg)
45
0
−45
−90 3ω 5ω 7ω
2 3 4
10 10 10
Frequency (rad/sec)
Experimental results with a nonlinear Load
32 10
28 9
THD of vo [%]
THD of vo [%]
24 8
7
20 6
16 5
12 4
8 3
2
4 1
0 0
0 1 2 3 4 5 6 7 8 9 10 11 12 0 1 2 3 4 5 6 7 8 9 10 11 12
Time [s] Time [s]
24 24
Output Voltage [V]
Output Voltage [V]

16 vo 16 vo
8 8
0 0
−8 −8
−16 −16
−24 −24
7 7.01 7.02 7.03 7.04 7.05 7.06 7 7.01 7.02 7.03 7.04 7.05 7.06
Time [s] Time [s]
4 4
i1 i2 i1 i2
2 2
Current [A]
Current [A]
0 0
−2 −2
−4 −4
7 7.01 7.02 7.03 7.04 7.05 7.06 7 7.01 7.02 7.03 7.04 7.05 7.06
Time [s] Time [s]
(a) without bypassing (b) with bypassing
Outline

Limitations of conventional droop controller
Robust droop controller
Universal droop controller
Harmonic droop controller
Droop control is intrinsically an EPLL
S1 = P1 + jQ1 S 2 = P2 + jQ2
Vo ∠0 o
Ro1 Ro 2
~ E ∠δ
1 1 Z E 2 ∠δ 2 ~
Inverters would be inevitably operated in parallel because of

inverters naturally operated in parallel in a large-scale system
the limited capacity of power semiconductor devices
the requirement of redundancy and reliability
They are required to provide
Accurate sharing of real power and reactive power at the same time
Excellent voltage regulation capability
Good power quality
Common control strategies
with the help of a communication: master-slave or supervisory control
without the help of a communication network: droop control.
Power delivery to a voltage source
S = P + jQ
An inverter can be i
modelled as a reference Z o ∠θ
Vo ∠0 o
voltage source with an ~ E∠δ
( vo )
output impedance Zo . ( vr )
The real power and reactive power dispatched to the

terminal via the output impedance Zo are
EVo V2 EVo
P =( cos δ − o ) cos θ + sin δ sin θ ,
Zo Zo Zo
EVo V2 EVo
Q =( cos δ − o ) sin θ − sin δ cos θ ,
Zo Zo Zo
where δ is the power angle.
Conventional Droop Control: L-inverters
For inductive output impedance, θ = 90◦ , then
EVo EVo V2
P= sin δ , Q= cos δ − o
Zo Zo Zo
⇓ when δ is small
EVo E − Vo
P≈ δ, Q≈ Vo =⇒ P ∼ δ , Q ∼ Vo
Zo Zo
conventional control strategy ⇓
ωi Ei
ω* E*
Ei = E ∗ − ni Qi , ωi = ω * − mi Pi Ei = E * − ni Qi
ωi = ω ∗ − mi Pi , Capacitive Inductive
0 Pi* Pi 0 Qi* Qi
Conventional Droop Control: R-inverters
For resistive output impedance, θ = 0◦ , then
EVo V2 EVo
P= cos δ − o , Q =− sin δ
Zo Zo Zo
E − Vo EVo
P≈ Vo , Q ≈− δ =⇒ P ∼ Vo , Q ∼ −δ
Zo Zo

Ei ωi
E* ω*
Ei = E * − ni Pi
Ei = E ∗ − ni Pi , ωi = ω * + mi Qi
ωi = ω ∗ + mi Qi , Capacitive Inductive
0 Pi* Pi 0 Qi* Qi
Conventional Droop Control: C-inverters
For capacitive output impedance, θ = −90◦ , then
EVo EVo V2
P =− sin δ and Q = − cos δ + o .
Zo Zo Zo
EVo E − Vo
P ≈− δ and Q ≈ − Vo , =⇒ P ∼ −δ , Q ∼ −Vo .
Zo Zo

ωi Ei
E*
ω*
Ei = E ∗ + ni Qi , ωi = ω * + mi Pi Ei = E * + ni Qi
ωi = ω ∗ + mi Pi , Capacitive Inductive
0 Pi* Pi 0 Qi* Qi
The theory seems perfect but it does not work well in
practice. What went wrong?
Revisit of the droop controller (for R-inverters)
E*
Ei - Pi vo
ni
vr
Ei = E ∗ − ni Pi ,
ωi = ω ∗ + mi Qi . 1
s
mi
Qi i
ω it+δ i
ω*
Fundamental limitations:
The same Ei } =⇒ Not robust at all !
The same per-unit output impedance
Inherent trade-off between power sharing and voltage regulation
E*
Ei - Pi vo
ni
vr
Ei = E ∗ − ni Pi ,
ωi = ω ∗ + mi Qi . 1
s
mi
Qi i
ω it+δ i
ω*
E*
Ei - Pi vo
ni
vr
Ei = E ∗ − ni Pi ,
ωi = ω ∗ + mi Qi . 1
s
mi
Qi i
ω it+δ i
ω*
E*
Ei - Pi vo
ni
vr
Ei = E ∗ − ni Pi ,
ωi = ω ∗ + mi Qi . 1
s
mi
Qi i
ω it+δ i
ω*
E*
Ei - Pi vo
ni
vr
Ei = E ∗ − ni Pi ,
ωi = ω ∗ + mi Qi . 1
s
mi
Qi i
ω it+δ i
ω*
Robust droop controller (patented)
E*
Ke
-
RMS
Ei 1 - ni Pi
s vo
vri
1 Qi i
mi
s
ω it+δ i
ω*
Accurate sharing of both real power and reactive power

Excellent voltage regulation
Low THD
Fast response
Reactive Power [Var]

24 2
20 P1 P2 Q1 Q2
Real Power [W]
16 1
12
0
8
4 −1
0
−4 −2
0 1 2 3 4 5 6 7 8 9 10 11 12 0 1 2 3 4 5 6 7 8 9 10 11 12
Time [s] Time [s]
24 24
Output Voltage [V]

20 E1 E2 16 vo
Voltage [V]
16 8
12 0
8 −8
4 −16
0 −24
0 1 2 3 4 5 6 7 8 9 10 11 12 7 7.01 7.02 7.03 7.04 7.05 7.06
Time [s] Time [s]
4 32
i1 i2 28
THD of vo [%]
2 24
Current [A]
20
0 16
12
−2 8
4
−4 0
7 7.01 7.02 7.03 7.04 7.05 7.06 0 1 2 3 4 5 6 7 8 9 10 11 12
Time [s] Time [s]
Droop controllers take different forms for inverters
with different output impedance. What happens if
the inverters in parallel operation have different types
of output impedance?
Is it possible to find a universal droop controller that

works for inverters having different types of output
impedance?
Droop controllers take different forms for inverters
with different output impedance. What happens if
the inverters in parallel operation have different types
of output impedance?
Is it possible to find a universal droop controller that

works for inverters having different types of output
impedance?
Revisit of the power delivery
S = P + jQ
Z o ∠θ
Vo ∠0 o
~ E∠δ
( vo )
( vr )
EVo V2 EVo
P =( cos δ − o ) cos θ + sin δ sin θ
Zo Zo Zo
EVo V2 EVo
Q =( cos δ − o ) sin θ − sin δ cos θ
Zo Zo Zo
which can be re-written as
" Vo2
#
P cos θ − sin θ EVo
Zo cos δ − Zo
=
sin θ cos θ
Zo sin δ
Q − EV o
.
Define " 2
#
P̃ EVo
Zo cos δ − VZoo
=
Zo sin δ
− EV
.
Q̃ o
Then,
P̃ ∼ E and Q̃ ∼ −δ
when |δ | is small. As a result,

P cos θ − sin θ P̃
=
Q sin θ cos θ
.
Q̃
This means that the vector P + jQ is obtained by rotatingthe
cos θ − sin θ
vector P̃ + j Q̃ by θ . The eigenvalues of are
sin θ cos θ
cos θ ± j sin θ , of which the real part cos θ is positive for any
output impedance with θ ∈ (− π2 , π2 ). Hence,
P ∼ P̃ and Q ∼ Q̃
=⇒ P ∼ E and Q ∼ −δ .
Since P ∼ E and Q ∼ −δ . the conventional droop
controller takes the form of
Ei = E ∗ − ni Pi
ωi = ω ∗ + mi Qi .
In practice, there are often small resistance in series with the

output impedance even if the output impedance is designed to
be inductive or capacitive. Hence, θ ∈ (− π2 , π2 ) and this droop
control strategy is universal and does not dependent on the
type of the output impedance.
Coincidently, this is the same as the droop control strategy for

R-inverters.
We have proven that the droop controller for
R-inverters is universal for inverters having
impedance with θ ∈ (− π2 , π2 ).
E*
Ke
-
RMS
Ei 1 - ni Pi
s vo
vri
1 Qi i
mi
s
ω it+δ i
ω*
OPAL-RT real-time simulation results
L-, C- and R-inverters, one each, share a load in the
ratio of 1:2:3.
P: [500 W/div] Q: [100 Var/div]

0 Var
L-inverter
R-inverter
C-inverter
C-inverter
L-inverter
0W R-inverter
t: [10 s/div] t: [10 s/div]
(a) Real power (b) Reactive power
R-, L-, C-inverters in parallel
P: [5 W/div] vo: [15V/div]

R-
C- 0V
L-
0W
Q: [1.5 Var/div] R- C- i: [1 A/div]
L-
R- 0A
C- L-
0 Var t: [3 s/div]
t: [3 s/div]
Injecting harmonic voltage components
Zo Load/grid
i
~ v o1
i1 ih
…
~ vr
vo ↓ … ↓ …
~ voh
…
(a) One circuit including all harmonics
S h = Ph + Qh voh = 0 if vrh is the

ih
same as the voltage
Z o ( jhω * ) dropped on the output
~ ~ ↓
v rh voh ih
impedance Zo by the
harmonic current com-
(b) The circuit at the h-th harmonic frequency
ponent ih .
Power delivery
S = P + jQ
i
Z o ∠θ
Vo ∠0 o
~ E∠δ ~ ( vo )
( vr )
(a) To a voltage source
S = P + jQ
i
Z o ∠θ
I∠0°
~ E∠δ ↓ (i )
( vr )
(b) To a current source
vo
Fourier Analysis to
Eh
the h-th harmonics

-nh Ph
calculate P, Q at
vrh
hω t+δ h 1 Qh i
-mh
s
ωt
h
It does not depend on the impedance type at the

harmonic frequency concerned.
In order to reduce multiple
harmonics in the output Fundamental E*
voltage, several harmonic droop controller
-
droop controllers corres- Ke RMS
ponding to the harmonic or-
ders can be included in the
E 1 -
controller to generate the n1 P vo
s
Calculation
required Σh vrh .
Power
vr vr1
The difficulty in defining ωt 1 Q

m1 i
the reactive power for the s
conventional droop control- ω *
ler has been avoided be- ωt

vo
cause the reactive power vr3 3rd-harmonic
droop controller i
in this strategy is defined
at the corresponding fre-
…
…
quency. Here, the fun- ωt

damental droop controller vrh th
h -harmonic
vo
is designed for R-inverters, droop controller i
which can also be used for
C-inverters and R-inverters.
Without With 3rd and 5th harmonics droop controller
6 6
i1 i2 i1 i2
Current [A]
Current [A]
4 4
2 2
0 0
−2 −2
−4 −4
7 7.01 7.02 7.03 7.04 7.05 7.06 7 7.01 7.02 7.03 7.04 7.05 7.06
Time [s] Time [s]
(a) Currents
20 20
10 10
vo [V]
vo [V]
0 0
−10 −10
−20 −20
7 7.01 7.02 7.03 7.04 7.05 7.06 7 7.01 7.02 7.03 7.04 7.05 7.06
Time [s] Time [s]
(b) Output voltage
20 20
16 16
Mag (%)
Mag (%)
THD=15.92% THD=8.57%
12 12
8 8
4 4
0 0
1 3 5 7 9 11 13 15 17 19 1 3 5 7 9 11 13 15 17 19
Harmonic order Harmonic order
(c) Harmonic voltage components
Droop control v.s. PLL
Droop control Phase-locked loops

(PLL)
fundamental to the widely adopted in
operation of power modern electrical
systems and now the engineering
parallel operation of
inverters
> 700 papers found > 5600 papers found
from ieeexplore from ieeexplore
We have found that they are intrinsically the same!

Droop control v.s. PLL
Droop control Phase-locked loops

(PLL)
fundamental to the widely adopted in
operation of power modern electrical
systems and now the engineering
parallel operation of
inverters
> 700 papers found > 5600 papers found
from ieeexplore from ieeexplore
We have found that they are intrinsically the same!

A droop controller is intrinsically an EPLL
e e
X
-
E Ke 1
2 s H (s ) X
R v
eq
X
sin cos
1 Kf
s
H (s ) X
θ ω s
Droop control
-
E 1
X µ1 X
2 s v
sin cos
1 1
µ 2 ( µ3 + ) X
θ s s
EPLL
What does this mean?
No more phase-locked loops in grid-tied converters!
What does this mean?
No more phase-locked loops in grid-tied converters!
Outline

Sinusoid-locked loops (SLL)
Phase-locked loops (PLL)
input Phase error Loop output

VCO
Detection Filter
(a) Operating concept
PD LF VCO
~ ⋅
v × v LPF d PI
θ 1 θ
s
y
sin
(b) Control structure

Qing-Chang Zhong (IIT, Chicago, zhongqc@ieee.org)of aandsimple
Control PLL for Energy and Power Systems
Power Electronics
SOGI-based PLL
v − 1 e
k ×
− . s
θ
eq 1
×
SOGI s
Quadrature-signal generator based on a second-order-generalised

integrator (SOGI-QSG)
e
αβ vd = E .
v SOGI θn
vq .
QSG eq θ 1 θ
dq PI
. s
θ
A single phase PLL equipped with the SOGI-QSG

Sinusoidal Tracking Algorithm (STA)
Also called the enhanced PLL (EPLL)
µ1 1 e
× E
×
s
v − sin
. cos
θn
.
1 θ 1 θ
× µ2
s s
µ3
What is the problem?
Response speed is low: often takes more than

one cycle
Robustness against frequency variations
Harmonics in the recovered signal
Solution:
A small synchronous machine can quickly synchronise
with the terminal voltage.
Synchr. machine connected to the grid
v = vm sin θ v i Xs e = E sin θ
~ SSM model ~
The real power P and reactive power Q flowing out of the machine are
vm E vm
P= sin (θ − θv ) , Q= [E cos (θ − θv ) − vm ] .
2Xs 2Xs
where E and vm are peak amplitude values instead of RMS values.
The machine is considered to be synchronised and floating on the grid when
E = vm , θ = θv .
In this case, P = 0 and Q = 0. Alternatively, if P and Q could be driven to zero then the
generated voltage e is the same as the input (terminal) voltage v. In other words, e can track
v if P and Q are controlled to be zero.
Mathematical model of a SG
f θ −Mf dif cos

e = Mf if θ̇ sin f θ,
dt
Dp
D E
Te = pMf if i , sin θ ,
f Tm
-
1 θ& 1 θ
Js s
-
Te
Eqn. (7)
Q Eqn. (8)
Q = −θ̇ Mf if hi , cos
f θi,
Eqn. (9) e
Mf if i
J θ̈ = Tm − Te − Dp θ̇ ,
Revisit of the synchronverter
Dp θ&r
-
Reset θg
Pset p Tm 1 θ& 1 θ
θ&n Js s
-
θc

Te
Eqn. (7)
Q Eqn. (8)
PWM
Eqn. (9)
e generation
- Mf if
Qset 1 i
Ks
Dq
- Amplitude v fb
vm detection
vr
Implementation of the SLL
Ki
.
s θr
−
Dp
Pset = 0 1 Tm − 1 θ
.
1 θ
θ&n
− Js s
Te
Formulas
Q of e, Q, Te e
Qset = 0 − 1 Mfif i 1 v
Ks Ls+R
The idea of the proposed SLL is to operate a virtual

(single-phase) synchronous generator with P = 0 and Q = 0 so
that the generated voltage e is the same as the fundamental
component of the terminal voltage v .
In single-phase systems, average values are used for
Z t
1
Te = Mf if i sin θ dt,
T t−T
Z t
1
Q =− θ̇ Mf if i cos θ dt,
T t−T
2π
where T = θ̇
is the period of voltage v .
The recovered voltage e is
e = θ̇ Mf if sin θ = E sin θ ,
and the amplitude is

E = θ̇ Mf if .
The stator current i is generated internally as the voltage
difference between e and v divided by the virtual synchronous
reactance Xs (s) = sL + R, i.e.,
e −v
i= .
sL + R
Tracking the frequency and phase
The frequency θ̇ should be the same as θ̇v in the steady state,
which means the reference frequency θ̇r should be the same as θ̇ .
This is achieved by the integrator Ksi .
After setting the desired real power to 0, the electromagnetic
torque Te is driven to zero when θ̇r = θ̇v .
At the steady state, Te is driven to zero and
P = θ̇ Te = 0
As a result, θ̇ is the same as the (angular) frequency θ̇v of

voltage v and θ is the same as the phase θv of voltage v ,
i.e., (
θ̇ = θ̇v ,
θ = θv .
The SLL locks to the frequency and phase of v .
Tracking the voltage amplitude
The voltage droop control in synchronverters is not
needed for the SLL because the generated voltage is
expected to be the same as the voltage v . The desired
reactive power is set to 0 and the loop to drive the
reactive power Q to zero is kept. At the steady state,
Q = 0 in addition to P = 0. Hence,
E = vm .
The SLL is synchronised with the voltage v because
voltage amplitude, frequency, and phase are well tracked
at the steady state.
It is worth noting that one advantage of the SLL is that
the frequency, phase, voltage amplitude and the
recovered signal are all directly available internally
without any extra calculation.
Tuning of the parameters (I)
The time constant of the frequency loop is
J
τf = .
Dp
A large τf is equivalent to having a large J, which makes the SLL
less sensitive to the variations in the grid frequency and improves
the stability. However, the response is slow. A small τf is
equivalent to having a small J, which leads to fast frequency
tracking. As a general rule of thumb, τf can be chosen much
smaller than the period of the voltage v so that the frequency can
be tracked very quickly.
The time constant of the amplitude loop is proportional to
K
τq =
θ̇n
Generally, τq ≫ τf , the frequency loop should be tuned much faster
than the amplitude loop. However, if a very large τq is chosen, it
would take long time for the voltage amplitude E to track vm .
Tuning of the parameters (II)
The inductance L and resistance R of the virtual synchronous

reactance Xs can be chosen small to enable a large transient
current i, which helps speed up the synchronisation process.
The ratio RL is the cut-off frequency of the filter sL+
1
R , which
determines the capability of filtering out the harmonics from the
voltage v .
The loop to generate the reference frequency θ̇r is an outer loop
for the frequency loop so it should be tuned much slower than the
frequency loop. Its time constant is
1
τfn =
Dp K i
and can be tuned as τfn = (10 ∼ 100)τf .
Parameters of the SLL for experiments
Parameters Values Parameters Values

fn 50 Hz J 2.0264 × 10−5
τf 0.0005 s K 4809.6
τfn 0.049 s Ki 100
τq 18.37 s L 0.3 mH
Dp 0.2026 R 0.01 Ω
Tracking a grid voltage
70 70 70
60 60 60
50 50 50
f [Hz]
f [Hz]
f [Hz]
40 40 40
30 30 30
20 20 20
10 10 10
0 0 0
0 0.04 0.08 0.12 0.16 0.2 0 0.04 0.08 0.12 0.16 0.2 0 0.04 0.08 0.12 0.16 0.2
Time [s] Time [s] Time [s]
(b) Frequency tracking
30 30 30
25 25 25
20 20 20
E [V]
E [V]
E [V]
15 15 15
10 10 10
5 5 5
0 0 0
0 0.04 0.08 0.12 0.16 0.2 0 0.04 0.08 0.12 0.16 0.2 0 0.04 0.08 0.12 0.16 0.2
(c) Detection of the voltage amplitude
30 30 30
15 15 15
e [V]
e [V]
e [V]
0 0 0
−15 −15 −15
−30 −30 −30

0 0.04 0.08 0.12 0.16 0.2 0 0.04 0.08 0.12 0.16 0.2 0 0.04 0.08 0.12 0.16 0.2
(d) Voltage tracking
10 10 10
8 8 8
THD [%]
THD [%]
THD [%]
6 6 6
4 4 4
2 2 2
0 0 0
0.1 0.12 0.14 0.16 0.18 0.2 0.1 0.12 0.14 0.16 0.18 0.2 0.1 0.12 0.14 0.16 0.18 0.2
(e) THD of e
8 8 8
6 6 6
θ [rad]
θ [rad]
θ [rad]
4 4 4
2 2 2
0 0 0
0 0.04 0.08 0.12 0.16 0.2 0 0.04 0.08 0.12 0.16 0.2 0 0.04 0.08 0.12 0.16 0.2
(e) Phase tracking
Tracking a voltage with a varying freq.
With the proposed SLL With the SOGI-based PLL With the STA
70 70 70
f fv f fv f fv
Frequency [Hz]
Frequency [Hz]
Frequency [Hz]
60 60 60
50 50 50
40 40 40
30 30 30
0 2 4 6 8 10 0 2 4 6 8 10 0 2 4 6 8 10
(a) Frequency tracking

45 45 45
E vm E vm E vm
Amplitude [V]
Amplitude [V]
Amplitude [V]
35 35 35
25 25 25
15 15 15
0 2 4 6 8 10 0 2 4 6 8 10 0 2 4 6 8 10
(b) Amplitude tracking
Tracking a square wave With the proposed SLL With the SOGI-based PLL With the STA
40 40 40
20 20 20
v [V]
v [V]
v [V]
0 0 0
−20 −20 −20
−40 −40 −40

0.1 0.12 0.14 0.16 0.18 0.2 0.1 0.12 0.14 0.16 0.18 0.2 0.1 0.12 0.14 0.16 0.18 0.2
(a) Input signal
100 100 100

f fv f fv f fv
Frequency [Hz]
Frequency [Hz]
Frequency [Hz]
80 80 80
60 60 60
40 40 40
20 20 20
0 0 0
0.1 0.12 0.14 0.16 0.18 0.2 0.1 0.12 0.14 0.16 0.18 0.2 0.1 0.12 0.14 0.16 0.18 0.2
(b) Frequency tracking
50 50 50
E vm E vm E vm
Amplitude [V]
Amplitude [V]
Amplitude [V]
45 45 45
40 40 40
35 35 35
30 30 30
0.1 0.12 0.14 0.16 0.18 0.2 0.1 0.12 0.14 0.16 0.18 0.2 0.1 0.12 0.14 0.16 0.18 0.2
(c) Amplitude tracking
60 60 60
30 30 30
e [V]
e [V]
e [V]
0 0 0
−30 −30 −30
−60 −60 −60

0.1 0.12 0.14 0.16 0.18 0.2 0.1 0.12 0.14 0.16 0.18 0.2 0.1 0.12 0.14 0.16 0.18 0.2
(d) Recovered voltage e
15 15 15
THD [%]
THD [%]
THD [%]
10 10 10
5 5 5
0 0 0
0.1 0.12 0.14 0.16 0.18 0.2 0.1 0.12 0.14 0.16 0.18 0.2 0.1 0.12 0.14 0.16 0.18 0.2
(e) THD of e
8 8 8
θe v θe v θe v
6 6 6
θ [rad]
θ [rad]
θ [rad]
4 4 4
2 2 2
0 0 0
0.1 0.12 0.14 0.16 0.18 0.2 0.1 0.12 0.14 0.16 0.18 0.2 0.1 0.12 0.14 0.16 0.18 0.2
(f) Phase tracking
Outline

Ripples in DC microgrids
DC microgrids are becoming
more and more popular
Electrolytic capacitors are
used to reduce voltage ripples
But
Fuel cells, batteries etc. do
not like ripple currents
bulky
heavy and
with limited lifetime
⇓
Is it possible to obtain low voltage ripples with small
capacitors?
Ripples in DC microgrids
DC microgrids are becoming
more and more popular
used to reduce voltage ripples
But
Fuel cells, batteries etc. do
not like ripple currents
bulky
heavy and
with limited lifetime
⇓
Is it possible to obtain low voltage ripples with small
capacitors?
Ripple eliminators
i I
S1 S3 ir
is
Ripple
vs C VDC Load
Eliminator
S2 S4
VDC: [25V/div]
VDC: [25 V/div]
2.5 V
90V
i: [8.33 A/div]
Va: [25 V/div] 42 V

610 V
vs :[320V/div]
ir: [8.33A/div]
is : [16A/div]
t: [5ms/div] t: [5 ms/div]
(a) without (b) with
Outline

Our Vision
To achieve seamless integration of

control, power electronics and power systems.
Future research plan
Fundamental Understanding
General
of Power and Energy
Control Theory
Systems
Advanced Control Strategies

for Power and Energy
Systems
High-speed trains
Marine Power
More-electric
Wind Power
Solar Power
Smart Grids
Heavy-duty
All-electric
Hybrid EV
Vehicles
Aircraft
ships
Summary — Enabling technologies
Synchronverters: Inverters that mimic SG
Self-synchronised synchronverters: No more PLLs

C-inverters

Limitations of conventional droop controller
Robust droop controller
Droop control is intrinsically an EPLL
Sinusoid-locked loops (SLL)

Acknowledgements
Research team and collaborators: Fred C. Lee, George
Weiss, Dushan Boroyevich, Tomas Hornik, Long Nguyen, Zhenyu
Ma, Wen-Long Ming, Yu Zeng.
Funding agencies:
EPSRC, Royal Academy of Engineering, TSB
Industrial partners
Further reading
Completely Autonomous
Power Systems (CAPS)
Next Generation Smart Grids
Qing-Chang Zhong
Nuclear Plants
Hydro Plants
Coal Plants SG
SG SG
SV SV
Transmission
Motors and Electric Vehicles
Distribution
SV
SV
SV SV
Lightings
Lighting
SV SV Solar Farms
HVDC
Electronic Apparatus Wind Farms
(to appear in 2015)
References
[1] Q.-C. Zhong and T. Hornik, Control of Power Inverters in Renewable Energy and
Smart Grid Integration. Wiley-IEEE Press, 2013.
[2] Q.-C. Zhong, Completely Autonomous Power Systems (CAPS): Next Generation
Smart Grids, Wiley-IEEE Press, to appear in 2015.
[3] Q.-C. Zhong and G. Weiss, “Synchronverters: Inverters that mimic synchronous
generators,” IEEE Trans. Ind. Electron., vol. 58, no. 4, pp. 1259–1267, Apr. 2011.
[4] Q.-C. Zhong, P.-L. Nguyen, Z. Ma, and W. Sheng, “Self-synchronised
synchronverters: Inverters without a dedicated synchronisation unit,” IEEE Trans.
Power Electron., vol. 29, no. 2, pp. 617–630, Feb., 2014.
[5] Q.-C. Zhong, “Robust droop controller for accurate proportional load sharing
among inverters operated in parallel,” IEEE Trans. Ind. Electron., vol. 60, no. 4, pp.
1281–1290, Apr. 2013.
[6] Q.-C. Zhong, “Harmonic droop controller to reduce the voltage harmonics of
inverters,” IEEE Trans. Ind. Electron., vol. 60, no. 3, pp. 936–945, Mar. 2013.
[7] Q.-C. Zhong and T. Hornik, “Cascaded current-voltage control to improve the
power quality for a grid-connected inverter with a local load,” IEEE Trans. Ind.
Electron., vol. 60, no. 4, pp. 1344–1355, Apr. 2013.
[8] Q.-C. Zhong and Y. Zeng, Control of Inverters via a Virtual Capacitor to Achieve
Capacitive Output Impedance, IEEE Trans. Power Electron., vol. 29, no. x, pp. x–x,
2014, to appear.
[9] T. Hornik and Q.-C. Zhong, “A current control strategy for voltage source inverters
in microgrids based on H ∞ and repetitive control,” IEEE Trans. Power Electron., vol.
26, no. 3, pp. 943–952, Mar. 2011.
Thank you.
zhongqc@ieee.org

ZhongQC Slides PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ZhongQC Slides PDF

Uploaded by

Copyright:

Available Formats

Integrated Control and Power Electronics

for Energy and Power Systems

The evolution, philosophy, and current standing

1998 2001 2004 2007 2010 2013

Pushing boundaries and breaking limits Research philosophy / How I think

From hardware to software Focused and thorough research: Deep thinking

Looking for hidden links

Associate Editor for

Electronic Apparatus Wind Farms

(to appear 2015)

IEEE Power Electronics Society Distinguished Lecturer, 2014-2015

Chaired by Kevin Daffey, Global Head of

The evolution, philosophy, and current standing of our research

“. . . the greatest engineering achievement of

“. . . the largest and most complex machine

“. . . the greatest engineering achievement of

“. . . the largest and most complex machine

DOE: Smart Grid System Report

Karady G and Holbert K 2004

DOE: Smart Grid System Report

Karady G and Holbert K 2004

DOE: Smart Grid System Report

Karady G and Holbert K 2004

Next-Generation Smart Grids:

Power Electronics Based Autonomous Power Systems

Next-Generation Smart Grids:

Power Electronics Based Autonomous Power Systems

Next-Generation Smart Grids:

Power Electronics Based Autonomous Power Systems

Ageing infrastructure (mostly over 100 years old)

Generated from the data at http://en.wikipedia.org/wiki/List_of_power_outages

Ageing infrastructure (mostly over 100 years old)

Residential, By Major Source Commercial, By Major Source

Industrial, By Major Source Transportation, By Major Source

Ageing infrastructure (mostly over 100 years old)

Ageing infrastructure (mostly over 100 years old)

Ageing infrastructure (mostly over 100 years old)

Upgrading the system, e.g. by introducing

Upgrading the system, e.g. by introducing

Strengthening the system, e.g. by introducing

Upgrading the system, e.g. by introducing

Strengthening the system, e.g. by introducing

800kV 6400 MW, 2011 800kV 6400

NWPG Ningula Shandong 800kV 6400 MW, 2015

800kV 5000-6000 MW, 2015 800kV 5000 MW, 2009

   "

     

#$%&'()$% *(+$%, -./01

         

Is there ONE simple mechanism to enable organic growth and

A383=9 E5@F8=>F R@D>F5@9

E7@=CM L83=4C@ LM98@K9

STUVWXYZ[VW \Z]VW^ _`abc

E5@F8=37>F GHH4=48<9 I>7; J4=K9 L354= J4=K9

Is there ONE simple mechanism to enable organic growth and

sejeok wgrxjopx rvpxgrk

wirou ~jeofur ~kjr}k

  

wgrxjoeipx yzzfofjnk {pim |fo}k ~egfo |fo}k

Is there ONE simple mechanism to enable organic growth and

¥¡ ©¤ª¡¢ª ¶¤¨¢ª¤

©¤¡§± °¡§¤ °±¤¯

·¸¹º»¼½¾¿º» À¾Áº»Â ÃÄÅÆÇ

"

#$%&'()$% *(+$%, -./01

sejeok wgrxjopx rvpxgrk

wirou ~jeofur ~kjr}k

¥¡ ©¤ª¡¢ª ¶¤¨¢ª¤

©¤¡§± °¡§¤ °±¤¯

©¤ª¡¢ª «¬¬¡ ¢ ®¡¯ °¡ ®¡¯