energies

Article
Power and Capacity Consensus Tracking
of Distributed Battery Storage Systems
in Modular Microgrids
Xianyong Zhang 1, *, Yaohong Huang 1 , Li Li 1 and Wei-Chang Yeh 2
1 School of Automation, Guangdong Polytechnic Normal University, Guangzhou 510665, China;
yaohongH365084741@163.com (Y.H.); liliarticle@163.com (L.L.)
2 Integration and Collaboration Laboratory, Department of Industrial Engineering and Engineering
Management, National Tsing Hua University, Hsinchu 46804804, Taiwan; yeh@ieee.org
* Correspondence: zhangfriendjun@163.com; Tel.: +86-20-3654-9341




Received: 27 April 2018; Accepted: 27 May 2018; Published: 4 June 2018


Abstract: Conventional microgrids have a specific system configuration and a complex hierarchical
control structure, which has resulted in difficulties in their economic development. A modular
microgrid based on distributed battery storage has been proposed to realize the rapid economic
development of small-to-medium microgrids. Control of modular microgrids is simplified to voltage
control within modules and exchange power control among modules. Battery power has great
influence on battery performance. Space-time complementary power characteristics among modules
help to alleviate power fluctuations, prolong the service life and realize the unified maintenance of
distributed batteries. Leader-following consensus theory of multi-agent systems is adopted to realize
the power and capacity consensus tracking of distributed battery storage in a modular microgrid.
Sufficient and necessary conditions for continuous-time and sampled-data bounded power and
capacity consensus tracking of distributed battery storages are deduced by a matrix analytical
method. Steady regions of sampling period and sampling delay for sampled-data bounded power
and capacity consensus tracking are determined by analytical or numerical solutions. Simulations and
experiments on a modular microgrid demonstration project located on DongAo Island (China) show
the effectiveness and robustness of the proposed power and capacity consensus tracking strategy
for distributed storage systems. The power and capacity consensus tracking strategy determines the
exchange power among modules and improves the control technology of modular microgrids.

Keywords: modular microgrid; leader-following consensus tracking; power and capacity; distributed
battery storages; sampling period; sampling delay

1. Introduction
Microgrids are an effective way to realize the large-scale integrated utilization of renewable energy
sources, such as wind and solar power. Microgrids can be integrated into the distribution network
as a flexible and controllable part, or run as stand-alone power systems to provide energy supply for
areas which are unreachable by the regular utility grid [1]. With the reform of the electricity market,
the application of microgrids is booming [2], and a large number of microgrid projects have been
built around the world. An off-grid microgrid with installed capacity 357 kW was developed on the
Isle of Eigg in Scotland to provide a reliable 24-h electricity supply to the islanders [3]. A 400 kW
microgrid comprising four 100 kW hybrid combined cooling heating and power (CCHP), uniquely
able to seamlessly transition between grid-tie and island-mode operation, provides power, water and
heat for the Brevoort building in New York [4]. A megawatt demonstration project located on Nanji
Island in China was also developed to provide clean energy [5]. With the development of sustainable

Energies 2018, 11, 1439; doi:10.3390/en11061439 www.mdpi.com/journal/energies

Energies 2018, 11, 1439 2 of 25

buildings, home microgrids with installed capacities of tens of kilowatts also represent an attractive
way to save building energy [6,7].
The system capacity of the abovementioned microgrid projects varies greatly and therefore their
configuration and operation mode also differ accordingly. Some microgrids can operate only in grid
connected mode and microsources usually work in power control mode [8,9]. Some microgrids can
operate only in the islanding mode and microsources work with a master-slave control or with the
droop control [10,11]. Some microgrids are able to operate in both on-grid and off-grid mode [12],
which need mode transition control. Based on the experimental microgrid built at the Prince—Electrical
Energy System Lab of the Polytechnic of Bari (Italy), the main technical issues on five operation modes
and five transitions were examined and implemented in detail [13,14]. Appropriate control is a
prerequisite for stable, economical and efficient operation of a microgrid. In order to standardize the
design of control systems, a three-layer hierarchical control scheme derived from ISA-95 is presented
in [15,16]. The primary control based on the droop method adjusts the voltage reference provided to
the inner current and voltage control loop [17]. The secondary control restores the deviations produced
by the primary control and distributed coordinated control strategy ensuring voltage tracking and
power sharing is applied [18]. The tertiary control is the optimum energy management of microgrid
and various optimum methods such as Model Predictive Control [19,20], Conditional Value at Risk [21],
Dynamic Programming [22], Flower Pollination Algorithm [23] and Particle Swarm Optimization [24]
are adopted. The conventional configuration and control of microgrids mimics a large-scale power
system AC grid, which may be applicable to large capacity microgrids, but is too complicated and
expensive for a small-to-medium microgrid.
In order to realize the rapid economic development of small-to-medium microgrids, we have
carried out a series of studies on stand-alone microgrid technology suitable for island applications.
Power supply becomes a major problem with the exploitation of islands. The traditional power system
on islands is the diesel generator, which generates noise and air pollution. Island ecology is usually
very fragile and it is urgent to use clean energy. A stand-alone modular microgrid based on distributed
batteries was proposed in [25], which can realize easy and economic networking. Each module has
the same hardware configuration, composed of wind and solar power, battery storage, loads and
converters. Modules are connected to the microgrid through a converter and so the AC bus of the
modular microgrid is segmental. Accordingly the control of the modular microgrid is simplified to the
voltage control within modules and exchange power control among modules. Voltage control within
modules has been discussed in detail [26–28], but how to determine exchange power among modules
still needs to be studied in depth according to different optimum objectives.
Battery storage systems are the energy and power balancing units in a modular microgrid.
Nowadays the most widely used storage systems are electrochemical energy storage units such as
lead-acid batteries or lithium-ion batteries. Battery lifespan is closely related to the charge and discharge
power, the depth of discharge, charge/discharge cycle times and so on [29]. The existing battery
lifespan prediction models are all established on regular charge and discharge processes [30]. Battery
power in a modular microgrid compensates for the unbalanced power between generation and load.
Because wind and solar power and loads are intermittent and random, the battery power fluctuates
violently, which deteriorates the power quality and results in difficulties for lifespan prediction [31].
When the battery power is different, some battery storage systems will be out of service due to
exceeding the capacity limit, which leads to the unexpected reconfiguration of a modular microgrid.
Modules are distributed in different locations and there exists space-time complementary power
characteristics among modules, which helps to alleviate the battery power fluctuations. Therefore
power and capacity consensus tracking of distributed battery storage is necessary to improve the
power quality and realize the unified maintenance of battery storage systems.
Each module is free to work in on-grid mode or off-grid mode, which leads to the reconfiguration
of the microgrid, and so a decentralized power and capacity consensus control scheme should
be adopted [32]. Multi-agent consensus theory based on a sparse communication network is an
Energies 2018, 11, 1439 3 of 25

appropriate tool for analyzing and designing a distributed system [33]. Chow first proposed an
incremental cost consensus algorithm to solve the conventional centralized economic dispatch
problem in a smart grid [34] and the relationship between the convergence rate and network
topology has been discussed in [35]. Yu put forward a virtual generation tribe-based consensus
algorithm for dynamic generation dispatch of smart grids [32]. A fully distributed coordination
control scheme including containment and consensus-based algorithm is proposed realizing a good
coordination between reactive power sharing and voltage bound in an AC microgrid [36]. The above-
mentioned multi-agent consensus studies on smart grids or microgrids mainly concentrate on the
design of consensus protocols, communication network topologies, and proof of convergence with
continuous-time. The actual control system is a sampled-data control system based on a digital
processor. The information is transmitted only at the sampling time. Sampling and communication
delays always exist, which has a great influence on the stability and the convergence rate of multi-agent
systems. We have carried out a series of theoretical researches on bounded consensus tracking of
multi-agent systems with sampled-data [37,38], which lays a good foundation to design a power and
capacity consensus tracking strategy for distributed battery storage in a modular microgrid.
The configuration and operation mode of a modular microgrid are introduced first. Then the
power relationships of a modular microgrid are analyzed and a multi-agent power and capacity
model of distributed battery storage with an undirected communication network can be established.
The continuous-time and sampled-data control protocols are designed based on the leader-following
consensus tracking theory. The sufficient and necessary conditions for bounded power and capacity
consensus tracking of distributed batteries are derived by a matrix analytical method and the steady
regions of sampling period and sampling delay can be determined by analytical or numerical solutions.
Simulations and experiments are then conducted on the modular microgrid demonstration project on
DongAo Island (China).
The main contribution of this paper is that power and capacity consensus tracking of distributed
battery storage is chosen as an optimum control objective, which can make full use of the space-time
complementary power characteristics among modules, to alleviate the power fluctuations and realize
the unified maintenances of distributed battery storage systems. Leader-following consensus tracking
theory of multi-agent systems is adopted to study the continuous-time and sampled-data control
protocols and stabilization ranges of parameters. The power and capacity consensus tracking strategy
of distributed battery storage determines the exchange power among modules and therefore the
control problem of the modular microgrid is solved completely. The power and capacity consensus
tracking strategy improves the control technology of a modular microgrid and promotes the economic
application of small-to-middle modular microgrids.
The organization of this paper is as follows: Section 2 illustrates the modular microgrid. Section 3
introduces leader-following consensus tracking theory of multi-agent systems. Section 4 discusses our
continuous-time power and capacity consensus tracking strategy of distributed battery storage systems.
Section 5 discusses sampled-data power and capacity consensus tracking strategy of distributed battery
storage. Section 6 presents simulations and experiments on the modular microgrid demonstration
project on DongAo Island. Section 7 discusses the conclusions.

2. Modular Microgrid
DongAo Island is a classic tourist island, located in Zhuhai City, Guangdong Province of Southern
China, with an area of 4.66 square kilometers and a permanent population of more than 600 people.
The original diesel generator power supply system on the island had a series of drawbacks such as
poor power quality, high electricity price and no guarantee of uninterruptible power supply. There are
abundant wind and solar resources and so the original diesel generator power system should be
converted to a renewable energy microgrid.
Energies 2018, 11, x FOR PEER REVIEW 4 of 26

power 2018,
Energies supply. There
11, 1439are abundant wind and solar resources and so the original diesel generator
4 of 25
power system should be converted to a renewable energy microgrid.

2.1. System Configuration

2.1. System Configuration
In
In order
order to to realize
realize power
power generation
generation and and consumption
consumption on on the
the spot
spot reducing
reducing the the power
power
transmission
transmission loss, photovoltaic (PV) arrays and wind turbines are installed in the concentrated load
loss, photovoltaic (PV) arrays and wind turbines are installed in the concentrated load
zones. Figure 11 shows
zones. Figure shows the the battery
battery storage
storage system,
system, rooftop
rooftop PV,
PV, terrestrial
terrestrial PV
PV and
and wind
wind turbines
turbines
installed
installed onon DongAo
DongAo Island. The structure
Island. The structure of of the
the modular microgrid on
modular microgrid on DongAo Island is
DongAo Island is shown
shown inin
Figure 2. There are three concentrated load zones on DongAo Island and
Figure 2. There are three concentrated load zones on DongAo Island and three modules are three modules are established.
Module 1 is the
established. power
Module 1 plant
is thezone.
power Module 2 is the
plant zone. comprehensive
Module building zone and
2 is the comprehensive module
building 3 isand
zone the
cultural center zone. Each module is an autonomous energy system and
module 3 is the cultural center zone. Each module is an autonomous energy system and has the has the same configuration,
which is composed ofwhich
same configuration, wind and solar generation,
is composed of wind battery storage,
and solar load and power
generation, batteryelectronic
storage, converter.
load and
Modules are interconnected through the 10 kV transmission network by
power electronic converter. Modules are interconnected through the 10 kV transmission networktransformers. Each moduleby
can freely access or separate from the grid. Module 0 is composed of three-porter
transformers. Each module can freely access or separate from the grid. Module 0 is composed converter, battery
of
storage and diesel
three-porter generator.
converter, Module
battery storage 0 works as V/F
and diesel voltage source
generator. Moduleestablishing
0 works astheV/F10voltage
kV voltage of
source
the transmission
establishing network.
the 10 The diesel
kV voltage of thegenerator
transmissionin Module 0 is The
network. a backup
dieselsource and works
generator in Moduleonly when
0 is a
the renewable energy is not enough.
backup source and works only when the renewable energy is not enough.

Figure 1. Modular
Figure 1. Modular microgrid
microgrid on DongAo Island.
on DongAo Island.

A three-port converter is the key equipment for the configuration and control of a modular
A three-port converter is the key equipment for the configuration and control of a modular
microgrid. It has AC-DC-AC topology. The 480 V battery pack composed of 240 2-Volt cells in
microgrid. It has AC-DC-AC topology. The 480 V battery pack composed of 240 2-Volt cells in series
series is connected to the DC port. The AC-DC part connected to the transformer works as a
is connected to the DC port. The AC-DC part connected to the transformer works as a rectifier in
rectifier in active power and reactive power (PQ) control mode and realizes the three-phase
active power and reactive power (PQ) control mode and realizes the three-phase balanced bidirectional
balanced bidirectional power flow between module and microgrid. When the module has surplus
power flow between module and microgrid. When the module has surplus energy, it releases energy
energy, it releases energy to the microgrid. When the energy is insufficient, the module absorbs
to the microgrid. When the energy is insufficient, the module absorbs energy from the microgrid.
energy from the microgrid. The other DC-AC inverter part is only responsible for establishing the
The other DC-AC inverter part is only responsible for establishing the voltage reference within the
Energies 2018,11,
Energies 11, x FOR PEER REVIEW 5 of
of25
26
2018, 11, 1439
Energies 2018, x FOR PEER REVIEW 55 of 26

voltage reference
voltage reference within
within thethe module.
module. Because
Because of of the
the isolation
isolation of
of three-port
three-port converter,
converter, the
the AC
AC bus
bus ofof
module.
the Because
modular of theisisolation
microgrid segmental, of three-port
so control converter,
of the the AC
modular bus of is
microgrid the modularto
simplified microgrid
the voltageis
the modular microgrid is segmental, so control of the modular microgrid is simplified to the voltage
segmental, so control
control within
within of theand
modules modular
exchange microgrid
poweriscontrol
simplified
control to themodules.
among voltage control within modules
The adopted
adopted topology andof
control modules and exchange power among modules. The topology of
exchange
three-port power control
converter is among in
shown modules.
Figure The
3. adopted
The invertertopology
with of three-port
three-phase converterstructure
four-wire is showncanin
three-port converter is shown in Figure 3. The inverter with three-phase four-wire structure can
Figure
realize3. The inverter
three-phase with three-phase
balanced output voltage four-wire
voltage evenstructure can realize
under nonlinear
nonlinear orthree-phase
unbalancedbalanced output
load conditions
conditions
realize three-phase balanced output even under or unbalanced load
voltage
and has even
has been under nonlinear
been discussed
discussed in or
in detail unbalanced
detail in
in [25].
[25]. load conditions and has been discussed in detail in [25].
and

Figure 2.
Figure 2. Structure
Structure of
of modular
modular microgrid
microgrid on
onDongAo
DongAoIsland.
Island.

Figure 3. Topology
Figure Topology of the
the three-port converter.
converter.
Figure 3.
3. Topology of
of the three-port
three-port converter.
2.2. Power
2.2. Power Relationship ofof Modular Microgrid
Microgrid
2.2. Power Relationship
Relationship of Modular
Modular Microgrid
Because the
Because the diesel
diesel generator
generator acts
acts only
only as
as aa backup
backup source,
source, diesel
diesel generator
generator power
power isis not
not
Becauseinthe
considered diesel
this generator
paper. Power acts only as a backup
transmission loss andsource,
power diesel generator
conversion power
loss are is not considered
neglected and the
considered
in this in Power
paper. this paper. Power transmission
transmission loss and loss conversion
power and power loss
conversion
are loss areand
neglected neglected
the andflow
power the
power flow
power flow chart
chart of
of modular
modular microgrid
microgrid is
is shown
shown in
in Figure
Figure 44 according
according to Figure 2.
chart of modular microgrid is shown in Figure 4 according to Figure 2. to Figure 2.
2.2.1. Power
2.2.1. Power balance in
in the module:
module:
2.2.1. Power balance
Balance in the
the Module

PPBati (t ) = P (t) + P (t ) + P (t) − P , i = 1, 2 , 3

Bati(t ) = P Ei(t) + P PVi(t ) + P WTi(t) − P LDi , i = 1 , 2 , 3 (1)
Ei PVi WTi LDi (1)
PBati (t) = PEi (t) + PPVi (t) + PWTi (t) − PLDi , i = 1, 2, 3 (1)
Energies 2018,11,
Energies2018, 11,1439
x FOR PEER REVIEW 66of
of25
26

where is positive for charging power and negative for discharging power of battery, is
where PBatipower,
the solar is positive for charging
is the wind power
power,and negative is the loadfor discharging
power, andpower is of battery,
the PPVi power
exchange is the
solar power,
between modulePWTi iisand
themicrogrid.
wind power, ThePLDi is the load
dynamics power,power
of battery and PEi is the exchange power between
are:
module i and microgrid. The dynamics of battery power are:
dPBati (t) / dt = dPEi (t) / dt + dPPVi (t) / dt + dPWTi (t) / dt − dPLDi / dt , i = 1, 2 , 3 (2)
dPBati (t)/dt = dPEi (t)/dt + dPPVi (t)/dt + dPWTi (t)/dt − dPLDi /dt, i = 1, 2, 3 (2)
where , and PLDi are uncontrollable random perturbation variables, so is chosen as
where PPVi , variable
the control PWTi andand PLDi are uncontrollable
can be regulated random
by perturbation
. variables, so PEi is chosen as the
control variable and
The unit of battery P can be regulated
Bati power is kW and the by P .
Ei unit of battery capacity is kWh. The relationship
The unit
between of battery
battery capacity power
and is kW and
battery the unit
power can of bebattery
expressed capacity
as: is kWh. The relationship between
battery capacity and battery power can be expressed as:
 PBati dt
SBati ( t ) =R P dt. (3)
SBati (t) = 3600 .
Bati
(3)
3600

2.2.2. Power
2.2.2. Power Balance
balance among
among Modules
modules:

P 0 (t ) = PE 0 (t ) = − PE 1 (t ) − PE 2 (t ) − PE 3 (t) (4)
PBat0 (Bat
t) = PE0 (t) = − PE1 (t) − PE2 (t) − PE3 (t) (4)
BecauseModule
Because Module00works
worksasasa avoltage source,P Pis
voltagesource, is a power slack node and is determined by
E0 E0 a power slack node and is determined by the
the exchanged
exchanged powers
powers of other
of other modules.
modules.

Figure 4. Power flow chart of the modular microgrid.

Figure 4. Power flow chart of the modular microgrid.

3. Leader-Following Consensus Tracking Theory of Multi-Agent Systems

3. Leader-Following Consensus Tracking Theory of Multi-Agent Systems
Figure 5 shows the undirected communication network of leader-following multi-agent system
Figure 5 shows the undirected communication network of leader-following multi-agent
[39]. Node set V = {1, 2,…, n} represents independent agents. Weighted adjacent matri A =  aij 
system [39]. Node set V = {1, 2, . . . , n} represents independent agents. Weighted adjacent matri  
 
A = aij describes the relationship between nodes. If there exists information communicationi
describes the relationship between nodes. If there exists information communication between agent
between j
> 0 iotherwise
and j, aiagent and j, ai j >ai0,= 0 . For anaiundirected
j otherwise j = 0. For an network, its adjacent
undirected network,matrix A is amatrix
its adjacent symmetric
A is
,
a symmetric matrix and aij = a ji . Laplacian matrix L = [lij ] is another matrix describing the network
matrix and aij = a ji . Laplacian matrix L = [ ] is another matrix describing the network topology and
topology and the element is:
the element is:
(
− aij i 6= j
1ij = (5)
∑
 −a

j 6 = i a ij
ij
i = /
i=jj
1ij =  agent 0 presents the leader. The leader adjacent matrix

If there is a leader in a multi-agent system, (5)
 aij i=j

is defined as B = diag (b1 , b2 , . . . , bn ). If there existsj ≠ i
information communication between agent i and
the leader, the leader adjacent element bi > 0, otherwise bi = 0.
 Energies 2018, 11, 1439 7 of 25

The dynamic characteristics of first-order multi-agent system can be described as [40]:

dxi (t)/dt = ui (t), i ∈ I (6)

where xi is the state of the agent and ui is the control variable. To ensure that all the agents can follow
the leader’s reference state, Wei presented a typical continuous-time consensus tracking protocol [40]:
n
∑ aij


ui (t) = bi ( x0 (t) − xi (t)) + x j (t) − xi (t) , i = 1, 2, . . . n. (7)
j =1

where x0 , xi , x j are states of the leader agent, agent i and agent j individually.
Through the periodic sampling technique, the sampled-data control model of Equation (6) is:

xi (kT + T ) = xi (kT ) + Tui (kT ), k = 0, 1, 2, . . . , i = 1, 2, . . . n. (8)

where T is the sampling period and k is the discrete time index.

When time delay is not considered, the control input can be expressed as:
n
∑ aij


ui (kT ) = −bi ( xi (kT ) − x0 (kT )) + x j (kT ) − xi (kT ) (9)
j =1
Energies 2018, 11, x FOR PEER REVIEW 8 of 26

Figure
Figure 5. 5. Undirectedcommunication
Undirected communication network
network of
of leader-following
leader-followingmulti-agent
multi-agentsystems.
systems.

4. Continuous-Time Power and Capacity Consensus Tracking Strategy of Distributed Battery

Lemma 1. Consider an undirected network topology composed of n agents, the leader is globally reachable only
Storages
when the Hermite matrix H = B + L is positive definite. That means all the eigenvalues of H have a positive real
In order to develop a universal model, n following modules are considered. According to
part Re(λi ( H )) > 0 [41].
Equations (2), (6) and (7), the continuous-time power and capacity consensus tracking protocol of
battery
Lemma storage
2. All is:
the roots of equation (1 + a + b) × t2 + 2(1 − b) × t + 1 − a + b = 0 are in the left half open
plan only when all dPEithe eigenvalues of equation s2 + as + b = 0 are located in the unit circle. a and b are real
u (
numbers [42].
i
t ) =
dt
= bi ( PBat 0
− PBati ) +
n
(a P − P )+ k ( S − S ) + k (
 j=1 ij Batj Bati i Bat 0 Bati  j=1 ij SBatj − SBati . (10)
n
)
" #
where bi and aij are adjacent coefficients A B of battery power network, ki and kij are the adjacent
Lemma 3. Suppose matrix P = is m + n order matrix, where A is m-order square matrix and D is
coefficients of battery capacity network: C D
n-order square matrix, B is m × n matrix and C is n × m matrix. ThedP determinant dP of blockdP matrix
dP P.
Let P=P −P
Bati Bati  Bat 0
, S Bati
= SBati − SBat 0 , Pxi = − Bat 0 + PVi + WTi − LDi then
dt dt dt dt

 A B  ,
 = | A| ·  D − CA−1 B when matrix A is an invertible matrix.
 
(1) | P| = 
 
   C D 
PBAT = (PBat 1 , PBat 2 ,..., PBatn
 )T S = (S ,S ,...,S )T P = ( Px 1 , Px 2 ,..., Pxn )T .
, BAT Bat 1 Bat 2 Batn , X The dynamics of
consensus variables PBati and SBati are:

 / dt 
 dP    
P P
 Bat  = G  Bat  +  x  . (11)
Energies 2018, 11, 1439 8 of 25

 
 A B 
 = | D | ·  A − BD −1 C  when matrix D is an invertible matrix.
 
(2) | P| = 
 
 C D 

4. Continuous-Time Power and Capacity Consensus Tracking Strategy of Distributed Battery Storages
In order to develop a universal model, n following modules are considered. According to
Equations (2), (6) and (7), the continuous-time power and capacity consensus tracking protocol of
battery storage is:

dPEi
= bi ( PBat0 − PBati ) + ∑nj=1 aij PBatj − PBati + k i (SBat0 − SBati ) + ∑nj=1 k ij SBatj − SBati . (10)



ui ( t ) = dt

where bi and aij are adjacent coefficients of battery power network, ki and kij are the adjacent coefficients
of battery capacity network:
ˆ
Let PBati = PBati − PBat0 , SBati ˆ = SBati − SBat0 , Pxi = − dPdtBat0 + dPdtPVi + dPdtWTi − dPdtLDi ,
ˆ = ( PBat1
then PBAT ˆ , PBat2
ˆ , . . . , PBatn T ˆ
ˆ ) , SBAT ˆ , SBat2
= (SBat1 ˆ ) T , PX = ( Px1 , Px2 , . . . , Pxn ) T .
ˆ , . . . , SBatn
The dynamics of consensus variables PBati and SBati are:
" # " # " #
ˆ /dt
d PBat ˆ
PBat P̂x
ˆ /dt =G ˆ + . (11)
dCBat CBat 0

where: " #
− H1 − H2
G= In . (12)
3600 0

H1 = B1 + L1 . (13)

H2 = B2 + L2 . (14)
 
b1 0 ... 0
 0 b ... 0 
2
B1 =  . (15)
 
 ... ... ... ... 
0 0 ... bn
 
k1 0 ... 0
 0 k ... 0 
2
B2 =  . (16)
 
 ... ... ... ... 
0 0 ... kn
 
∑ a1j − a12 ... − a1n
 j 6 =1 
 n 
 − a21 ∑ a2j ... − a2n
 

L1 = 
 j 6 =2 .
 (17)
 ... ... ... ... 
 
∑ anj
 −a − an2 ... 
n1
j6=n
 
∑ k1j −k12 ... −k1n
 j 6 =1 
 n 
 −k21 ∑ k2j ... −k2n
 

L2 = 
 j 6 =2 .
 (18)
 ... ... ... ... 
 
∑ k nj
 −k −k n1 ... 
n1
j6=n
Energies 2018, 11, 1439 9 of 25

The characteristic polynomials of matrix G is

   
 sI + H
 n 1 H2  
  2 H2 
det(sI − G ) =  In  = s In + sH1 + . (19)
 − 3600 sIn  3600 

H1 is the Hermitian matrix of multi-agent battery power network. H2 is the Hermitian matrix
of multi-agent battery capacity network. The dynamics characteristics of battery power and battery
capacity have different time constants. Battery capacity is the integral of battery power and so battery
capacity has large time constant. In order to simplify analyses, let H2 = cH1 and c is the proportional
coefficient. Then, Equation (19) is simplified as:
n  
  cλi ( H1 )
det(sI − G ) = s2 In + (s + c) H1  = ∏ s2 + sλi ( H1 ) + . (20)
 
i =1
3600

where λi ( H1 ) is eigenvalues of matrix H1 .

Bounded consistency tracking can be achieved if and only if all characteristic roots of Equation (21)
have negative real parts:
cλ ( H )
s2 + sλi ( H1 ) + i 1 = 0 (21)
3600
Because H1 is a real symmetric matrix, so eigenvalues λi ( H1 ) of matrix H1 is real number.
The characteristic roots of Equation (21) have negative real parts when λi ( H1 ) and c are both positive
real numbers.

5. Sampled-Data Power and Capacity Consensus Tracking Strategy of Distributed Battery Storages

According to Equations (8) and (10), the sampled-data control variable PEi is:

PEi (kT + t) = PEi (kT ) + Tui (kT ), k = 0, 2, 3 . . . , i = 1, 2, . . . , n. (22)

Let sampling delay be:

τ = mT + ε (23)

where τ > 0, which includes communication delay. m is non-negative integer and ε ∈ (0, T ).
The sampled-data power and capacity consensus tracking protocol of battery storage is
designed as:
n

bi ( PBat0 (kT − mT − T ) − PBati (kT − mT − T )) + ∑ aij PBatj (kT − mT − T ) − PBati (kT − mT − T )




j =1

if t ∈ [kT, kT + ε)



 n
+k i (SBat0 (kT − mT − T ) − SBati (kT − mT − T )) + ∑ k ij SBatj (kT − mT − T ) − SBati (kT − mT − T )





j =1
ui ( t ) = n . (24)
bi ( PBat0 (kT − mT ) − PBati (kT − mT )) + ∑ aij PBatj (kT − mT ) − PBati (kT − mT )




j =1

if t ∈ [kT + ε, kT + T )



 n
+k i (SBat0 (kT − mT ) − SBati (kT − mT )) + ∑ k ij SBatj (kT − mT ) − SBati (kT − mT )





j =1

where Pxi = − PLDi (kT + T ) + PLDi (kT ) + PDGi (kT + T ) − PDGi (kT ) + PBat0 (kT ) − PBat0 (kT + T ).
If Pxi is bounded, that is Pxi ≤ Pw
Let ∆PBati (kT ) = PBati (kT ) − PBat0 (kT ), ∆SBati (kT ) = SBati (kT ) − SBat0 (kT )., Then:

∆PBAT (kT ) = (∆PBat1 (kT ), ∆PBat2 (kT ), . . . , ∆PBatn (kT )) T . (25)

∆SBAT (kT ) = (∆BBat1 (kT ), ∆BBat2 (kT ), . . . , ∆BBatn (kT )) T . (26)

Energies 2018, 11, 1439 10 of 25

Substituting Equations (22) and (24) into Equation (1) leads to:

∆PBat (kT + T ) = PBat (kT ) − εH1 ∆PBat (kT − mT − T ) − ( T − ε) H1 ∆PBat (kT − mT )

−εH2 ∆SBat (kT − mT − T ) − ( T − ε) H2 ∆SBat (kT − mT ) (27)
+ P0 (kT + T ) − P0 (kT ) − PBat0 (kT + T ) + PBat0 (kT ) 1n .
 

Defining the tracking error vector:

!T
∆PBAT (kT ) T , ∆PBAT (kT − T ) T , . . . , ∆PBAT (kT − mT ) T , ∆PBAT (kT − mT − T ) T ,
E(kT ) = . (28)
∆SBAT (kT ) T , ∆SBAT (kT − T ) T , . . . , ∆SBAT (kT − mT ) T , ∆SBAT (kT − mT − T ) T

Then the difference equation about error vector can be obtained as:
" #
PX
E(kT + T ) = G1 E(kT ) + . (29)
0(2m+3)n

−( T − ε) H1 −εH1 −( T − ε) H2 −εH2
 
In 0n × n ... 0n × n 0n × n ...

 In 0n × n ... 0n × n 0n × n 0n × n 0n × n ... 0n × n 0n × n 


 ... ... ... ... ... ... ... ... ... ... 

0n × n 0n × n ... 0n × n 0n × n 0n × n 0n × n ... 0n × n 0n × n
 
 
 
 0n × n 0n × n ... In 0n × n 0n × n 0n × n ... 0n × n 0n × n 
G1 = 
 TIn
. (30)
 3600 0n × n ... 0n × n 0n × n In 0n × n ... 0n × n 0n × n 

 0n × n 0n × n ... 0n × n 0n × n In 0n × n ... 0n × n 0n × n
 

 ... ... ... ... ... ... ... ... ... ...
 

 
 0n × n 0n × n ... 0n × n 0n × n 0n × n 0n × n ... 0n × n 0n × n 
0n × n 0n × n ... 0n × n 0n × n 0n × n 0n × n ... In 0n × n
Then the error vector can be expressed as:
" #
N −1
PW
E( NT ) ≤ G1N E(0) + ∑ G1N −k−1
0(2m+3)n
. (31)
k =0

where E(0) is the initial state of the error vector.

The characteristic polynomial of matrix G1 is:

(s − 1) In ( T − ε) H1 ( T − ε) H2
 
0n × n ... εH1 0n × n 0n × n ... εH2

 − In sIn ... 0n × n 0n × n 0n × n 0n × n ... 0n × n 0n × n 


 ... ... ... ... ... ... ... ... ... ... 

0n × n 0n × n ... sIn 0n × n 0n × n 0n × n ... 0n × n 0n × n
 
 
 
0n × n 0n × n ... − In sIn 0n × n 0n × n ... 0n × n 0n × n
(32)
 
F (s) = det(sI − G1 ) = det 
TIn
 .

 − 3600 0n × n ... 0n × n 0n × n (s − 1) In 0n × n ... 0n × n 0n × n 


 0n × n 0n × n ... 0n × n 0n × n − In sIn ... 0n × n 0n × n


... ... ... ... ... ... ... ... ... ...
 
 
 
 0n × n 0n × n ... 0n × n 0n × n 0n × n 0n × n ... sIn 0n × n 
0n × n 0n × n ... 0n × n 0n × n 0n × n 0n × n ... − In sIn

5.1. Case A: c = 0
Only power consensus tracking is considered when c = 0. Equation (32) is simplified as:
 
(s − 1) In 0n × n ... ( T − ε) H1 εH1

 − In sIn ... 0n × n 0n × n 

A= ... ... ... ... ... . (33)
 
− In
 
 0n × n 0n × n sIn 0n × n 
0n × n 0n × n 0n × n − In sIn
Energies 2018, 11, 1439 11 of 25

According to Lemma 3, by using block matrix to calculate, we can get:

n n o
F (s) = ∏ sm+2 − sm+1 + ( T − ε)λi ( H )s + ελi ( H ) . (34)
i =1

Therefore, the battery power can realize bounded consensus tracking only when all eigenvalues
of Equation (34) are located within the unit circle, that is, the spectral radius of G1 is less than 1.
Specially, when the time delay is less than one period, in other words, m = 0, Equation (34) can be
simplified as follows:
s2 + [( T − τ )λi ( H ) − 1]s + τλi ( H ) = 0. (35)

According to Lemma 2, all roots of Equation (35) are located within the unit circle only when the
roots of Equation (36) are on the left-half open plane:

Tλi ( H )t2 + 2(1 − τλi ( H ))t + [2 + (2τ − T )λi ( H )] = 0. (36)

Further analysis, the roots can meet the requirements when parameters satisfy the inequality (37):
(
1 − τλi ( H ) > 0
. (37)
2 + (2τ − T )λi ( H ) > 0

Therefore, the necessary and sufficient condition for sampled-data power consensus tracking of
distributed storages can be obtained as Equation (38) by simplification:

1
(
0<τ< λmax ( H )
2 . (38)
τ < T < 2τ + λmax ( H )

Similarly, the necessary and sufficient condition for sampled-data power consensus tracking of
distributed storages can be obtained when m = 1:
 1
 0<T< H)
λmax ( √
. (39)

1− Tλmax ( H )
 T < τ < min 2T, T +
λmax ( H )

5.2. Case B: c > 0

Sampled-data power and capacity consensus tracking is considered when c > 0.
According to Lemma 3, by using block matrix to calculate, we can get:
 h i 
T
 (s − 1) In + 36001
εH2 + ( T − ε) H2 sm (Ts−1) 0n × n . . . ( T − ε) H1
 
εH1
s ( m +1) ( s −1 )

 
 − I n sIn ... 0n × n 0n × n  
= (s − 1)n sn(m+1) 
 
... ... ... ... . . .  .


0n×n − In

sIn

 0n × n 0n × n   (40)
 


h 0n × n i 0n × n 0n × n − In sIn 
 2( m +1) 2 1
= s (s − 1) In + 3600 s(m+1) TεcH1 + s(m+2) ( T − ε) TcH1 + s(m+1) (s − 1)(( T − ε)s + ε) H1 

n h i
= ∏ s2(m+1) (s − 1)2 + 3600 Tc
s(m+1) ελi + s(m+2) ( T − ε)λi + s(m+1) (s − 1)(( T − ε)s + ε)λi
i =1

Therefore, power and capacity of battery storages can achieve bounded consensus tracking only
when all the eigenvalues of the Equation (40) are within the unit circle, that is, the spectral radius of
matrix G1 is less than 1. Equation (40) shows that it is a higher-order equation. It is impossible to
determine the analytic solutions and only the numerical solutions can be obtained.
Energies 2018, 11, x FOR PEER REVIEW 13 of 26
Energies 2018, 11, 1439 12 of 25

microgrid demonstration project as shown in Figure 2. Figure 6 shows the undirected

communication
6. Simulationsnetwork topology of modular microgrid and the adjacent elements are 0.3. Agent 1,
and Experiments
2, 3 represents module 1, 2, 3, respectively and Agent 0 presents module 0. When battery capacity
To verify the performances of power and capacity consensus tracking strategy of distributed
consensus tracking is considered, let c = 0.1 and = 0.1 . According to Figure 6, we can get the
storages in modular microgrid, simulations and experiments are carried out on the modular microgrid
demonstration project as shown in Figure 2. Figure
0 06.3shows 0 the undirected communication network
 
weighted
topologyadjacent
of modular matrix A = the
microgrid and ( )
=  0.3 elements
aij adjacent 0 0.3 are 0.3.
, the leader
Agent 1, adjacent
2, 3 represents modulematrix
 0 00..3 When
1, 2, 3, respectively and Agent 0 presents module 
0  battery capacity consensus tracking is

 and H2 = 0.1H
considered, let c = 0.1 1 . According to Figure 6, we canget the weighted adjacent
 0.3 0 0  0 0.3 0  0.6 0.3
−0.30 00  

    
B1 =matrix
 0 A0.=3 aij0 =  , 0.3 , the
0.3the 0Hermitian leader H = B1 matrix
adjacent + L1 = B1−= 3 00.90.3 −00.3 
0.  ., and the
 
and matrix 1
 0 0 0.3  0 0.3 0  0 −0.3 0.6 
0 0 0.3
 
0.6 −0.3 0
 
So eigenvalues
Hermitian matrix ofH1 =H 1B1 are 
( )
+ L1 λ=max H
−0.3 ( )
= λ1 0.9
H = 1−.20.3 ( )
, λ2 ( )
.H = 0.6 and λ3 H = 0.3 . According to
Lemma 1, the eigenvalues are positive real −0.3 0.6
0 numbers and the leader agent is globally reachable.

Figure 6. Undirected
Figure 6. Undirectedcommunication networkofofmodular
communication network modular microgrid.
microgrid.

6.1. Simulations of Continuous-Time

So eigenvalues Power
of H1 are λmax and
(H) = λ1 Capacity
( H ) = 1.2,Consensus Tracking
λ2 ( H ) = 0.6 and λ3Strategy
( H ) = 0.3. According to
Lemma
In the 1, the eigenvalues
following Figureare
7 positive
to Figurereal25,
numbers and the
the cyan leader
line, blueagent
line,isorange
globallyline
reachable.
and green line
represent variables of module 0, 1, 2, 3 respectively. Only power consensus tracking is considered
6.1. Simulations of Continuous-Time Power and Capacity Consensus Tracking Strategy
first. Load power is 0, 10, 20 and 30 kW in module 0, 1, 2 and 3 individually. Exchange power
between In the following Figures 7 to 25, the cyan line, blue line, orange line and green line represent
each module and microgrid is shown in Figure 7. When the system is stable, Module 0
variables of module 0, 1, 2, 3 respectively. Only power consensus tracking is considered first. Load
outputs 15 kW to microgrid, Module 1 outputs 5 kW to microgrid, Module 2 inputs 5 kW from
power is 0, 10, 20 and 30 kW in module 0, 1, 2 and 3 individually. Exchange power between each
microgrid and Module 3 inputs 15 kW from microgrid. The distributed battery power in modules in
module and microgrid is shown in Figure 7. When the system is stable, Module 0 outputs 15 kW to
Figure 8 showsModule
microgrid, that it 1tends to be
outputs consistent
5 kW and Module
to microgrid, the discharging
2 inputs 5power
kW fromconverge to and
microgrid 15 kW after 7 s.
Module
The cyan dotted
3 inputs 15 kWline is the
from power The
microgrid. of the leader agent,
distributed batterythat
poweris, the distributed
in modules energy
in Figure storage
8 shows thatpower
it
realizes
tendsthe
tobounded consensus
be consistent and the tracking.
discharging power converge to 15 kW after 7 s. The cyan dotted line
is the power of the leader agent, that is, the distributed energy storage power realizes the bounded
consensus tracking.
 Energies 2018, 11, 1439 13 of 25
Energies 2018, 11, x FOR PEER REVIEW 14 of 26
Energies 2018, 11, x FOR PEER REVIEW 14 of 26

7. Exchange power
Figure7.
Figure of
of modules.
Figure 7.Exchange
Exchange power
power of modules.
modules.

ThenThen power
powerand andcapacity
capacityconsensus
consensus tracking
tracking areare both
both considered.
considered.The
considered. Theinitial
initialbattery
battery storage
storage
storage
capacity
capacity
capacity isisisSBat0 =
= 120
= 120 kWH,
120 kWH, kWH, SBat1 = 150
= 150
= 150kWH, kWH,
kWH,SBat2 = 180 = 180
= 180 kWH,
kWH,kWH, = 210When
SBat3 = 210 = 210
kWH. kWH.
kWH. the When
When thethe
system
system
is is
stable, stable,
Figure Figure
9 shows 9 shows
that that
Module Module
0 outputs 0 outputs
10.5 kW 10.5
to kW to
microgrid,
system is stable, Figure 9 shows that Module outputs 10.5 kW to microgrid, Module 1 outputs 3.5 microgrid,
Module Module
1 outputs 1 outputs
3.5 kW to3.5
kWkW to tomicrogrid,
microgrid, Module
microgrid, Module
2 inputs
Module 22inputs
3.5 kW 3.5
inputs 3.5 kW
from from
microgrid
kW microgrid
and Module
from microgrid and3Module
and Module
inputs 3 3inputs
10.5 inputs
kW from 10.5
10.5 kW
microgrid.
kW fromfrom
The distributed
microgrid.
microgrid. The battery
Thedistributed power
distributedbattery in modules
batterypower
power in is different
in modules
modules is and P
is different
different =and
Bat0 and − 10.5 =
kW, = −10.5
P −10.5=
Bat1 kW, −
kW, 13.5 = =
kW,
−13.5
−13.5
P Bat2 kW, −16.5 kW,
= kW, = =P −16.5
−16.5
Bat3 =− 19.5 kW as
kW,
kW, =
=shown
−19.5
−19.5inkW Figure
as 10.
as shown
shownBecause
in the time
in Figure
Figure 10.10.constant
Because
Because ofthe
battery
thetimetime
storage
constant capacity
of is
battery very large,
storage the
capacity
constant of battery storage capacity is very large,timeis to
veryreach power
large, the consensus
time to is
reach very
powerlong. The
consensus difference
time to reach power consensus is very long. is very power
long.
among
The
The batteries
difference
difference is toamong
power
power eliminate the capacity
amongbatteries
batteries isis to differences
to eliminate
eliminate among
the
the batteries.
capacity
capacity differences
differencesamongamongbatteries.
batteries.
Simulations results show that continuous-time power and capacity consensusis
Simulations
Simulations results
results showshowthat continuous-time
that continuous-time power and
power capacity
and consensus
capacity tracking
consensus reachable
tracking
trackingis is
and the
reachable leader-following
and the power
leader-following and capacity
power consensus
and tracking
capacity model
consensus
reachable and the leader-following power and capacity consensus tracking model of distributed of distributed
tracking model battery
of storages
distributed
in modular
battery microgrid
storagesininmodular is effective.
modularmicrogrid
microgridisis effective.
effective.
battery storages

Figure 8. Battery power in modules.

Figure 8. Battery power in modules.
Energies 2018, 11, 1439 14 of 25
Energies 2018, 11, x FOR PEER REVIEW 15 of 26
Energies 2018, 11, x FOR PEER REVIEW 15 of 26

Figure
Figure 9.
9. Exchange
Figure9. power of
Exchange power of modules.
modules.
modules.

Figure 10. Battery

Figure Battery power
power in
inmodules.
modules.
Figure 10.
10. Battery power in modules.

6.2. Simulations
6.2. of Sampled-DataPower Power Consensus Tracking
Tracking Strategy
6.2.Simulations
SimulationsofofSampled-Data
Sampled-Data Power Consensus Consensus Tracking StrategyStrategy
When m = 0, according to Equation (38), the range of the sampling period and sampling delay
Whenmm==0,0,according
When accordingtotoEquation
Equation(38),(38),thetherange
range ofof
the sampling
the sampling period
period and sampling
and sampling delay
delay for
for system stabilization is:
system stabilization is:
for system stabilization is: (
00<<ττ < < 00.8
.8
0 < τ < 0.8 (41)(41)
τ < T < 2τ ++1.1.6
τ < T < 2τ 6 (41)
τ < T < 2τ + 1.6
LetLet
thethe
sampling
sampling delay τ = τ0.2 =s and
delay 0.2 ssampling
and sampling T = 0.5Ts.=Loads
periodperiod 0.5 s.are unchanged.
Loads Exchange
are unchanged.
power Let the
between sampling
each moduledelay andτ = 0.2
microgrid s and
is sampling
shown in period
Figure 11.
Exchange power between each module and microgrid is shown in Figure 11. When the system T =
When 0.5 s.
the Loads
system are
is unchanged.
stable, module
is
0Exchange
outputs power
15 kW between
to microgrid, each module
module 1and microgrid
outputs 5 kW is
toshown in
microgrid, Figure
stable, module 0 outputs 15 kW to microgrid, module 1 outputs 5 kW to microgrid, module 2 inputs module11. When
2 inputsthe5system
kW fromis
stable,
microgrid module
5 kW from and 0 outputs
module and
microgrid 15 kW
3 inputs
module to microgrid,
15 kW module
from15microgrid.
3 inputs 1 outputs 5
The battery
kW from microgrid. kW to
Thepowermicrogrid,
battery inpower module
modules 2 inputs
in Figure
in modules in 12
5 Figure
kW from
shows that microgrid
12 it tendsthat
shows to beand module
to be 3consistent
consistent
it tends inputs
and the15 kWthe from
discharging
and microgrid. The battery
power converge
discharging power to 15power
converge kW in
kWmodules
after
to 15 8after 8 s.in
s. Results
Figure
show 12show
that
Results shows
the that
theit stabilization
stabilization
that tends to be
range is consistent
effective asand theasdischarging
shown
range is effective in Equation
shown power
(41). converge
in Equation (41). to 15 kW after 8 s.
ResultsWhen
showmthat the stabilization range is effective as shown in Equation (41).
= 1, according to Equation (39), the range of the sampling period and sampling delay
for When
systemmstabilization
= 1, according are: to Equation (39), the range of the sampling period and sampling delay
for system stabilization are:
Energies 2018,
Energies 2018, 11,
11, x
x FOR
FOR PEER
PEER REVIEW
REVIEW 16 of
16 of 26
26

Energies 2018, 11, 1439 15 of 25

00 < <TT<< 00..88
 

When m = 1, according to Equation < τ <
(39), range
the +
11 −
of the T  period and sampling delay
11..22T
− sampling (42)
(42)
for


T
T < τ < min
min 
 2
2T
T ,T
,T + 
 
1. 2
system stabilization are: (  1. 2
0 < T < 0.8 n √
Let
Let the
the sampling
sampling period
period T
T == 0.2 s and the the sampling −1.2T τ
delay τ == 0.3
0.3 s.
s. Loads
Loads are (42)
are unchanged.
o
0.2
T <s τand
< min 2T, T + 1delay
sampling
1.2
unchanged.
Figure 13
Figure 13 shows
shows the
the exchange
exchange power
power of of modules.
modules. TheThe battery
battery power
power in in Figure
Figure 14
14 shows
shows that
that it
it also
also
Let
realizes sampling period T = 0.2 s and the sampling delay = 0.3 s. Loads are
realizes the power consensus tracking. But the convergence rate is slower with the decrease of the
the power consensus tracking. But the convergence τ rate is slower unchanged.
with the Figure
decrease of 13
the
sampling period. According to Equations (41) and (42), the system is unstable when T = 2.4 s and ττ
shows
sampling the exchange
period. power
According oftomodules.
Equations The battery
(41) and power
(42), the in Figure
system is 14 shows
unstable that
when it also
T = realizes
2.4 s and the
power
= 0.2
= s.consensus
0.2 s. Figure 15
Figure tracking.
15 shows
shows But
that
that the
the
the convergence
battery
battery power rate
power is slower with the decrease of the sampling period.
is divergent.
is divergent.
According
Simulation results verify the effectiveness ofunstable
to
Simulation Equations
results (41) and
verify (42),
the the system
effectiveness is
of the
the stablewhen
stable T = 2.4 of
domain
domain s and
of τ = 0.2 s.
sampling
sampling Figureand
period
period 15
and
shows
sampling that
sampling delaythe battery
delay with power is
with sampled-data divergent.
sampled-data according
according to to Equation
Equation (34).
(34).

Figure
Figure11.
Figure 11.Exchange
11. Exchangepower
Exchange powerof
power of modules(((τ =
ofmodules
modules = 0.2 s，T
= 0.2 s, T ==
0.2 s，T 0.5
=0.5 s).
0.5s).
s).

Figure 12.
Figure 12. Battery
Battery power
power in
in modules (( =
modules (τ = 0.2 s,
= 0.2 s, TT
0.2 s, T === 0.5
0.5 s).
s).
 Energies 2018, 11, 1439 16 of 25

Simulation results verify the effectiveness of the stable domain of sampling period and sampling
delay2018,
Energies with11,sampled-data according to Equation (34).
x FOR PEER REVIEW 17 of 26

Figure13.
Figure Exchangepower
13.Exchange power of
of modules ( ==0.3
modules (τ 0.3 s,
s, T T= =0.2
0.2s).s).

Figure14.
Figure Batterypower
14.Battery power in
in modules ( ==0.3
modules (τ 0.3 s,
s, T T= =0.2
0.2s).s).
Energies 2018, 11, 1439 17 of 25
Energies 2018, 11, x FOR PEER REVIEW 18 of 26

Energies 2018, 11, x FOR PEER REVIEW 18 of 26

Figure15.
Figure 15.Battery
Batterypower modules((τ == 2 s,
power in modules 2 s,TT==2.4
2.4s).s).

6.3. 6.3. Simulations

Simulations of Sampled-Data
of Sampled-Data Power
FigurePower
15. andCapacity
and
Battery Capacity
power Consensus
Consensus
in modules Tracking
( = 2 s, T =Strategy
Tracking Strategy
2.4 s).

WhenWhen sampled-data
sampled-data powerand
power andcapacity
capacity consensus
consensus tracking
trackingisisconsidered,
considered, it is
it difficult to have
is difficult to have
6.3.
the Simulations of Sampled-Data
analytical solutions, Power
but easy and numerical
to have Capacity Consensus
solutionsTracking Strategy
for Equation (40).
the analytical solutions, but easy to have numerical solutions for Equation (40).
Equation
When (40) is simplified
sampled-data poweras:
and capacity consensus tracking is considered, it is difficult to have
Equation (40) is simplified as:
the analytical solutions,
3 but easy toTc have (numerical solutions for Equation (40).
F (s ) = ∏ s 2( m + 1)(s − 1)2 + s m + 1)ελi + s ( m + 2)(T − ε )λi  + s ( m + 1)(s − 1) ((T − ε )s + ε ) λi (43)
Equation
3 (40) is simplified as:3600   i
i =1 Tc h
F (s) = ∏ s 2( m +1)
(3s − 1) + 2
s(m+1) ελi + s(m+2) ( T − ε)λi + s(m+1) (s − 1)(( T − ε)s + ε)λi (43)
3600 Tc
Figure 16
i =1 F (s ) = ∏
showss 2( m +the stabilization
(s − 1) +
1) 2

srange
3600 period
( m + 1)
i
ελ of
+ ssampling(T − ε period
( m + 2)
i (
)λ  + s (and ) i (43)
m + 1) sampling delay when m = 0.
(s − 1) (T − ε )s + ε λ
The stabilizationi = 1range of sampling is from 0.1 to 3.2 s. Figure 17 shows the stabilization
Figure
ranges of16sampling
Figure shows
16 showsthe thestabilization
period and sampling
stabilization range
range of
delay sampling =period
when m period
of sampling and
1. Theand sampling
stabilization
sampling delay
range
delay of when
when m =m0.= 0.
sampling
TheThestabilization
period range
is from 0–0.5
stabilization rangeof sampling
s. Figure period
18 shows
of sampling periodis from
the is 0.1 to
stabilization3.2 s.
from 0.1 toranges Figure
3.2 s. of 17 shows
sampling
Figure the stabilization
period
17 shows theand ranges
sampling
stabilization
of sampling of period
delay when
ranges m = 2.and
sampling The sampling delay
stabilization
period range
and sampling when
ofdelaym when
= 1. period
sampling The
m =stabilization
1.isThe 0 to range
0.5 s. of
fromstabilization sampling
range period is
of sampling
from 0–0.5iss.from
period Figure 18 s.
0–0.5 shows the
Figure 18stabilization ranges of sampling
shows the stabilization ranges of period andperiod
sampling sampling delay when
and sampling
m =delay
2. The stabilization range of sampling period is from 0 to 0.5 s.
when m = 2. The stabilization range of sampling period is from 0 to 0.5 s.

Figure 16. Stabilization ranges of sampling period and sampling delay ( = 0).

Figure 16. Stabilization ranges of sampling period and sampling delay ( = 0).
Figure 16. Stabilization ranges of sampling period and sampling delay (m = 0).
 Energies 2018, 11, 1439 18 of 25
Energies 2018, 11, x FOR PEER REVIEW 19 of 26

Energies 2018, 11, x FOR PEER REVIEW 19 of 26

Figure 17.17.
Stabilization
Stabilizationranges
rangesof
ofsampling period and delay ( =1). 1).
samplingdelay
Figure
Figure 17. Stabilization ranges of samplingperiod
sampling periodand
andsampling
samplingdelay ((m =
= 1).

The stabilization
The stabilization
range
stabilizationrange
of sampling
rangeofofsampling
sampling
delay
delay
varies
varies
correspondingtotothe
corresponding
the samesampling
sampling period
delay varies corresponding to the samesame
sampling periodperiod
when
when
when m is different.
m is different.
m is different. The numerical
The numerical
The numerical solutions
solutions
solutions of of Equation
Equationof(43)
Equation (43) give
(43) give
give a guidance a guidance
foraselection
guidance forselection
for selection
of sampling of of
period
sampling
sampling period
and samplingperiod and
and
delay sampling
to sampling
realize thedelay
delay to
to realize the
realizebounded
sampled-data sampled-data
the sampled-data bounded
power and bounded
capacity powerand
power
consensus and capacity
capacity
tracking of
consensus
consensus tracking
distributedtracking of distributed battery storages.
of distributed battery storages.
battery storages.

Figure 18. Stabilization ranges of sampling period and sampling delay ( = 2).
Figure 18. Stabilization ranges of sampling period and sampling delay (m = 2).
Figure 18. Stabilization ranges of sampling period and sampling delay ( = 2).
6.4. Simulations of Topology Reconfiguration
6.4. Simulations of Topology Reconfiguration
6.4. Simulations of Topology Reconfiguration
Figure 19 shows the operation mode of module 3. Grid-connected mode and standalone mode
Figure 19 shows the operation mode of module 3. Grid-connected mode and standalone mode
areFigure 19 shows
indicated the operation
by 1 and modeModule
0 individually. of module
3 is 3. Grid-connected
connected mode andfrom
to the microgrid standalone
1 s to 10mode
s,
are indicated by 1 and 0 individually. Module 3 is connected to the microgrid from 1 s to 10 s,
areseparated
indicatedfrom
by 110and
s to 020individually. Module
s, and reconnected to 3the
is microgrid
connectedagain
to the microgrid
from fromin1modules
20 s. Loads s to 10 s,
separated from 10 s to 20 s, and reconnected to the microgrid again from 20 s. Loads in modules
separated from 10 s toWhen
remain unchanged. 20 s, module
and reconnected to the microgrid
3 is in grid-connected mode,again from 20 s.power
the consensus Loadsisin−15
modules
kW.
remain unchanged. When module 3 is in grid-connected mode, the consensus power is −15 kW. When
Whenunchanged.
remain Module 3 is in standalone
When module mode,is the consensus powermode,
is −10 kW.
Module 3 is in standalone mode, the3 consensus
in grid-connected
power is −10 kW. the consensus power is −15 kW.
When Module 3 is in standalone mode, the consensus power is −10 kW.
Energies 2018, 11, x FOR PEER REVIEW 20 of 26

Energies
Energies 11, 11,
2018,
2018, 1439PEER REVIEW
x FOR 19 of 25
20 of 26

Figure 19. Operation state of module D.

Figure 20 show that all the battery power of modules can realize consensus during the whole
process. When Module 3 is off theFigure microgrid, the load of
Operation state
Module 3 is self-supplied and the battery
19. Operation
Figure 19. stateofofmodule
moduleD. D.
power of Module 3 is −30 kW. Meanwhile, Module 0 outputs 10 kW to microgrid and Module 2
input Figure
10 kW
Figurefrom
20 20 microgrid
show
showthat
thatall asthe
all shown
the in power
battery
battery Figure
powerof 21.
ofmodules
modules can
can realize
realize consensus
consensus during
during the whole
the whole
Simulation
process.
process. When Whenresults
Module show
Module 3 is that
3 isoff themicrogrid,
offthe
the power and
microgrid, thecapacity
the load consensus
loadofofModule
Module tracking strategy
3 is3 self-supplied of distributed
and and
is self-supplied the battery
the battery
storages
power ofisof
power robust
Module3under
Module 3is −30system
is −30 reconfiguration,
kW. Meanwhile,
kW. Meanwhile, Module
Module which
0 outputs is kW
10
0 outputs beneficial
to kW
10 totomicrogrid
microgrid capacity expansion
and Module
and Moduleor2
2 input
unexpected
10 10
input kWkW exit,
from and increases
microgrid
from asthe
as shown
microgrid reliability
in
shownFigure 21.of the
in Figure 21.system.
Simulation results show that the power and capacity consensus tracking strategy of distributed
storages is robust under system reconfiguration, which is beneficial to capacity expansion or
unexpected exit, and increases the reliability of the system.

Figure20.
Figure Exchange power
20.Exchange powerininmodules.
modules.

Figure 20. Exchange power in modules.

 Energies 2018, 11, 1439 20 of 25

Simulation results show that the power and capacity consensus tracking strategy of distributed
storages is robust under system reconfiguration, which is beneficial to capacity expansion or
unexpected
Energies exit, PEER
2018, 11, x FOR and increases
REVIEW the reliability of the system. 21 of 26

Figure21.
Figure Battery power
21.Battery powerininmodules.
modules.

6.5. Experiments
6.5. Experiments of Power
of Power ConsensusTracking
Consensus Tracking of
of Distributed
Distributed Batteries
Batteriesin in
Modular Microgrid
Modular Microgrid
FigureFigure 22 shows
22 shows wind
wind powerininmodules
power modules 1, 1, 2
2 and
and3.3.Wind
Windpower
power varies
variesgreatly in the
greatly in different
the different
modules. The average power is 10.08, 12.28, and 12.98 kW, individually. Figure 23 shows solar power
modules. The average power is 10.08, 12.28, and 12.98 kW, individually. Figure 23 shows solar
in modules 1, 2 and 3. The average power is 5.28, 5.12 and 4.11 kW, individually. The solar power
power in modules 1, 2 and 3. The average power is 5.28, 5.12 and 4.11 kW, individually. The solar
has similar distribution characteristics but there are occasional drops because of partial shadows.
power has24similar
Figure shows thedistribution
load powercharacteristics but there
in different modules. are occasional
The average drops
power is 19.10, because
21.39 and 5.51 ofkWpartial
shadows. Figure The
individually. 24 shows the load power
load characteristics in different
are also different modules.
Because of The average
different power
electrical is 19.10,
uses. If each21.39
and module
5.51 kWruns individually.
in standaloneThe load
mode, thecharacteristics
difference between are power
also different
generationBecause
and powerof different
consumption electrical
is
uses.only
If each module runs in standalone mode, the difference between power generation
compensated by the local battery of each module. The battery power is shown as the dotted line and power
consumption is only25compensated
A, B, C in Figure and the batteryby the local
power variesbattery
greatly inofdifferent
each module.
module.The
Thebattery power is shown
drastic fluctuation of
battery power damages and reduces the service life of battery, which leads to difficult
as the dotted line A, B, C in Figure 25 and the battery power varies greatly in different module. The maintenance
of battery.
drastic When
fluctuation of modules are interconnected
battery power damages and andreduces
the power
the and capacity
service life ofconsensus trackingleads
battery, which is to
applied, the consensus battery power is shown as the blue line in Figure 25,
difficult maintenance of battery. When modules are interconnected and the power and capacitywhich make full use of the
space-time complementary power characteristics among modules and greatly decreases the fluctuation
consensus tracking is applied, the consensus battery power is shown as the blue line in Figure 25,
of battery power. It shows the consensus power is smoother than stand-alone individual battery
which make full use of the space-time complementary power characteristics among modules and
power, which can not only prolong the lifespan of batteries, but also realize the unified maintenance
greatly decreases the fluctuation of battery power. It shows the consensus power is smoother than
of batteries.
stand-alone individual battery power, which can not only prolong the lifespan of batteries, but also
realize the unified maintenance of batteries.
 Energies 2018, 11, 1439 21 of 25
Energies 2018,
Energies 2018, 11,
11, xx FOR
FOR PEER
PEER REVIEW
REVIEW 22 of
22 of 26
26

Figure 22.
Figure22.
Figure Wind power.
Wind power.
22. Wind power.

Figure 23.
Figure 23. Solar power.
power.
Figure 23. Solar
Solar power.
 Energies 2018, 11, 1439 22 of 25
Energies
Energies 2018,
2018, 11,
11, x
x FOR
FOR PEER
PEER REVIEW
REVIEW 23
23 of
of 26
26

Figure
Figure24.
24. Load power.
Load power.

Figure 25. Battery power.

Figure 25. Battery power.

7. Conclusions
7. Conclusions
The The
power andand
power capacity consensus
capacity consensus tracking
trackingprotocol
protocol of
of distributed batterystorage
distributed battery storage systems
systems in in a
modular microgrid is designed in this paper. The leader-following power and capacity
a modular microgrid is designed in this paper. The leader-following power and capacity consensus consensus
tracking model
tracking of distributed
model of distributedbattery
batterystorage
storage is
is established accordingto to
established according thethe modular
modular microgrid
microgrid
demonstration project on DongAo Island (China). The sufficient and necessary
demonstration project on DongAo Island (China). The sufficient and necessary conditions conditions for for
continuous-time
continuous-time andand sampled-databounded
sampled-data boundedpower
power and
and capacity
capacityconsensus
consensus tracking of distributed
tracking of distributed
battery storages are deduced by matrix analytical method. The continuous-time
battery storages are deduced by matrix analytical method. The continuous-time power and power and capacity
capacity
consensus tracking is stable when the Hermitian matrix of power communication network is positive
consensus tracking is stable when the Hermitian matrix of power communication network is
positive definite and the ratio of capacity adjacent coefficient to power adjacent coefficient is
positive real number. The steady regions of sampling period and sampling delay for sampled-data
power and capacity consensus tracking is determined by analytical or numerical solutions.
Especially two cases of m = 0 and m = 1 are discussed in detail. Simulation results verify the
validation of the power and capacity consensus tracking strategy and the stabilization ranges of
Energies 2018, 11, 1439 23 of 25

definite and the ratio of capacity adjacent coefficient to power adjacent coefficient is positive real
number. The steady regions of sampling period and sampling delay for sampled-data power
and capacity consensus tracking is determined by analytical or numerical solutions. Especially
two cases of m = 0 and m = 1 are discussed in detail. Simulation results verify the validation of
the power and capacity consensus tracking strategy and the stabilization ranges of parameters.
The power and capacity consensus tracking is robust under reconfiguration of the modular microgrid.
Experiments on a wind-solar modular microgrid show that the power and capacity consensus tracking
of distributed storages makes full use of the space-time complementary power characteristics among
modules, improving the lifespan and realizing the unified maintenance of batteries. The power and
capacity consensus tracking strategy determine the exchange power among modules. Therefore the
control problem of modular microgrid is solved completely and the economic promotion of modular
microgrids is realized.

Author Contributions: X.Z. and L.L. proposed the power and capacity consensus tracking strategy of distributed
battery storages in modular microgrid. Y.H. did the simulation research. W.-C.Y. analyzed the data. X.Z. wrote
the paper.
Acknowledgments: This work was supported by the Science and Technology Planning Project of Guangdong
Province (2017A010102013) and Guangzhou Science and Technology Planning Project (201806020025).
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Mumtaz, F.; Bayram, I.S. Planning, operation, and protection of microgrids: An overview. Energy Procedia
2017, 107, 94–100. [CrossRef]
2. Wang, J.; Zhong, H.; Tang, W.; Rajagopal, R.; Xia, Q.; Kang, C.; Wang, Y. Optimal bidding strategy for
microgrids in joint energy and ancillary service markets considering flexible ramping products. Appl. Energy
2017, 205, 294–303. [CrossRef]
3. Chmiel, Z.; Bhattacharyya, S.C. Analysis of off-grid electricity system at Isle of Eigg (Scotland): Lessons for
developing countries. Renew. Energy 2015, 81, 578–588. [CrossRef]
4. Panora, R.; Gehret, J.E.; Furse, M.M.; Lasseter, R.H. Real-world performance of a CERTS microgrid in
Manhattan. IEEE Trans. Sustain. Energy 2013, 5, 1356–1360. [CrossRef]
5. Chen, C.; Chen, J.; Cao, H.; Huang, Y.; Xiao, Y.; Li, L. Control scheme of bi-directional converter for microgrid
under the model of parallel running on Nanji island. IEEE Trans. Smart Grid 2014, 2, 30–36. (In Chinese with
English abstract).
6. Marzband, M.; Fouladfar, M.H.; Akorede, M.H.; Lightbody, G.; Pouresmaeil, E. Framework for smart
transactive energy in home-microgrids considering coalition formation and demand side management.
Sustain. Cities Soc. 2018, 40, 136–154. [CrossRef]
7. Marzband, M.; Fouladfar, M.H.; Savaghebi, M.; Pouresmaeil, E.M.; Guerrero, J.; Lightbody, G. Smart
transactive energy framework in grid-connected multiple home microgrids under independent and coalition
operations. Renew. Energy 2018, 126, 95–106. [CrossRef]
8. Marzband, M.; Sumper, A.; Ruiz-Alvarez, A.; Dominguez-Garcia, J.L.; Tomoiaga, B. Experimental evaluation
of a real time energy management system for stand-alone microgrids in day-ahead markets. Appl. Energy
2013, 106, 365–376. [CrossRef]
9. Turner, G.; Kelley, J.P.; Storm, C.L.; Wetz, D.A.; Lee, W.L. Design and active control of a microgrid testbed.
IEEE Trans. Smart Grid 2017, 6, 73–81. [CrossRef]
10. Stadler, M.; Cardoso, G.; Mashayekh, S.; Forget, T.; Deforest, N.; Agarwal, A.; Schönbein, A. Value streams in
microgrids: A literature review. Appl. Energy 2016, 162, 980–989. [CrossRef]
11. Bevrani, H.; Shokoohi, S. An intelligent droop control for simultaneous voltage and frequency regulation in
islanded microgrids. IEEE Trans. Smart Grid 2013, 4, 1505–1513. [CrossRef]
12. Ribeiro, L.A.D.S.; Saavedra, O.R.; Lima, S.L.D.; Matos, J.G.D. Isolated micro-grids with renewable hybrid
generation: The case of Lençóis Island. IEEE Trans. Sustain. Energy 2010, 2, 1–11. [CrossRef]
13. Alegria, E.; Brown, T.; Minear, E.; Lasseter, R.H. CERTS microgrid demonstration with large-scale energy
storage and renewable generation. IEEE Trans. Smart Grid 2014, 5, 937–943. [CrossRef]
Energies 2018, 11, 1439 24 of 25

14. Cagnano, A.; De Tuglie, E.; Dicorato, M.; Forte, G.; Trovato, M. PrinCE Lab experimental microgrid Planning
and operation issues. In Proceedings of the 2015 IEEE 15th International Conference on Environment and
Electrical Engineering (EEEIC), Rome, Italy, 10–13 June 2015; pp. 1671–1676. [CrossRef]
15. Cagnano, A.; De Tuglie, E.; Cicognani, L. Prince-Electrical Energy Systems Lab: A pilot project for smart
microgrids. Electr. Power Syst. Res. 2017, 148, 10–17. [CrossRef]
16. Guerrero, J.M.; Vasquez, J.C.; Matas, J.; Vicuna, L.G.D.; Castilla, M. Hierarchical control of droop-controlled
AC and DC microgrids—A general approach toward standardization. IEEE Trans. Ind. Electron. 2011,
58, 158–172. [CrossRef]
17. Bidram, A.; Davoudi, A. Hierarchical structure of microgrids control system. IEEE Trans. Smart Grid 2012,
3, 1963–1976. [CrossRef]
18. Riverso, S.; Tucci, M.; Vasquez, J.C.; Guerrero, J.M.; Ferrari-Trecate, G. Stabilizing plug-and-play regulators
and secondary coordinated control for AC islanded microgrids with bus-connected topology. Appl. Energy
2018, 210, 655–661. [CrossRef]
19. Godina, M.; Rodrigues, E.M.G.; Pouresmaeil, E.; Matias, J.C.O.; Catalao, J.P.S. Model predictive control home
energy management and optimization strategy with demand response. Appl. Sci. 2018, 8, 408. [CrossRef]
20. Godina, R.; Rodrigues, E.M.G.; Pouresmaeil, E.; Catalao, J.P.S. Optimal residential model predictive control
energy management performance with PV microgeneration. Comput. Oper. Res. 2017, 1–14. [CrossRef]
21. Tavakoli, M.; Shokridehaki, F.; Akorede, M.F.; Marzband, M.; Vechiu, I.; Pouresmaeil, E. CVaR-based energy
management scheme for optimal resilience and operational cost in commercial building microgrids. Int. J.
Electr. Power Energy Syst. 2018, 100, 1–9. [CrossRef]
22. Shuai, H.; Fang, J.; Ai, X.; Tang, Y.; Wen, J.; He, H. Stochastic optimization of economic dispatch for microgrid
based on approximate dynamic programming. IEEE Trans. Smart Grid 2018, 1. [CrossRef]
23. Wang, S.; Peng, D.; Dong, R.; Yong-He, L. Modified Flower Pollination Algorithm and Applications on
Optimization Dispatch of Microgrid. J. Northeast. Univ. 2018, 39, 334–338. (In Chinese with English abstract).
[CrossRef]
24. Kerdphol, T.; Fuji, K.; Mitani, Y.; Watanabe, M.; Qudaih, Y. Optimization of a battery energy storage system
using particle swarm optimization for stand-alone microgrids. Int. J. Electr. Power Energy Syst. 2016, 81, 32–39.
[CrossRef]
25. Zhang, X.; Yeh, W.; Jiang, Y.; Huang, Y.; Xiao, Y.; Li, L. A case study of control and improved simplified
swarm optimization for economic dispatch of a stand-alone modular microgrid. Energies 2018, 11, 793.
[CrossRef]
26. Zhang, X.; Shu, J.; Wu, C.; Zhou, L.; Song, X. Island microgrid based on distributed photovoltaic generation.
Power Syst. Prot. Control 2014, 6, 55–61. (In Chinese with English abstract). [CrossRef]
27. Zhou, L.; Shu, J.; Zhang, X.; Zhang, Z. Control strategy of voltage quality in distributed energy microgrids.
Power Syst. Technol. 2012, 36, 18–22, (In Chinese with English abstract). [CrossRef]
28. Zhang, X.; Zhang, X.; Li, L.; Luo, F.; Zhang, Y. A new cascaded control strategy for paralleled line-interactive
UPS with LCL filter. Earth Environ. Sci. 2016, 40. [CrossRef]
29. Dufo-López, R.; Bernal-Agustín, J.L. Multi-objective design of pv–wind–diesel–hydrogen–battery systems.
Renew. Energy 2008, 33, 2559–2572. [CrossRef]
30. Jenkins, D.P.; Fletcher, J.; Kane, D. Lifetime prediction and sizing of lead-acid batteries for microgeneration
storage applications. IET Renew. Power Gener. 2008, 2, 191–200. [CrossRef]
31. Fotouhi, A.; Propp, K.; Auger, D.J.; Longo, S. State of Charge and State of Health Estimation Over the Battery
Lifespan. Green Energy Technol. 2018, 267–288. [CrossRef]
32. Zhang, X.; Yu, T.; Yang, B.; Li, L. Virtual generation tribe based robust collaborative consensus algorithm for
dynamic generation command dispatch optimization of smart grid. Energy 2016, 101, 34–51. [CrossRef]
33. Miloje, S.R.; Miroslav, K. Distributed adaptive consensus and synchronization in complex networks of
dynamical systems. Automatica 2018, 91, 233–243. [CrossRef]
34. Zhang, Z.; Chow, M.-Y. Incremental cost consensus algorithm in a smart grid environment. In Proceedings
of the Power and Energy Society General Meeting, San Diego, CA, USA, 24–29 July 2011; Volume 5, pp. 1–6.
[CrossRef]
35. Zhang, Z.; Chow, M.-Y. Convergence analysis of the incremental cost consensus algorithm under different
communication network topologies in a smart grid. IEEE Trans. Power Syst. 2012, 27, 1761–1768. [CrossRef]
Energies 2018, 11, 1439 25 of 25

36. Han, R.; Meng, L.; Trecate, G.F.; Coelho, E.A.A.; Vasquez, J.C.; Guerrero, J.M. Containment and
consensus-based distributed coordination control for voltage bound and reactive power sharing in AC
microgrid. In Proceedings of the Applied Power Electronics Conference and Exposition (APEC), Tampa, FL,
USA, 26–30 March 2017; Volume 6, pp. 3549–3566. [CrossRef]
37. Li, L.; Zhang, X.; Wang, Z.; Lin, Z. Consensus of Leader-Following Multi-Agent Systems with Sampling
Information under Directed Networks. In Proceedings of the IEEE International Conference on CSE & EUC,
Guangzhou, China, 21–24 July 2017; Volume 10, pp. 63–68. [CrossRef]
38. Li, L.; Fang, H. Bounded consensus of leader-following multi-agent systems with sampling delay for
double-integrator dynamics. Int. J. Model. Identif. Control 2013, 19, 323–332. [CrossRef]
39. Shi, Y.; Yin, Y.; Wang, S.; Liu, Y.; Liu, F. Distributed Leader-following Consensus of Nonlinear Multi-agent
Systems with Nonlinear Input Dynamics. Neurocomputing 2018, 286, 193–197. [CrossRef]
40. Ren, W. Multi-vehicle consensus with a time-varying reference state. Syst. Control Lett. 2007, 56, 474–483.
[CrossRef]
41. Hu, J.; Hong, Y. Leader-following coordination of multi-agent systems with coupling time delays. Physica A
2007, 374, 853–863. [CrossRef]
42. Ren, W.; Cao, Y. Convergence of sampled-data consensus algorithms for double-integrator dynamics.
In Proceedings of the 47th IEEE Conference on Decision and Control, Cancun, Mexico, 9–11 December 2008;
Volume 16, pp. 3965–3970. [CrossRef]

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).