DC connection

Applied Energy xxx (2017) xxx–xxx

Contents lists available at ScienceDirect

Applied Energy journal homepage: www.elsevier.com/locate/apenergy

A novel design of architecture and control for multiple microgrids with hybrid AC/DC connection Pan Wu a, Wentao Huang a,⇑, Nengling Tai a, Shuo Liang b a b

Department of Power Electrical Engineering, Shanghai Jiao Tong University, Shanghai, SH 200240, China Guangxi Electric Power Research Institute, Nanning, GX 530023, China

h i g h l i g h t s

g r a p h i c a l a b s t r a c t


Proposes an advanced microgrid

interface based on MMC and energy storage system for multiple microgrids.
Proposes a novel architecture for multiple microgrids with hybrid AC/DC connection.
Proposes different control schemes for multiple microgrids under various operation conditions.
The large-scale integration of distributed renewable energies in multiple microgrids is enhanced.
The optimal use of distributed generators in multiple microgrids is realized.

a r t i c l e

i n f o

Article history: Received 28 March 2017 Received in revised form 13 July 2017 Accepted 14 July 2017 Available online xxxx Keywords: Multiple microgrids Hybrid unit of common coupling (HUCC) Architecture Control scheme Distributed renewable energy

a b s t r a c t Microgrid provides an effective approach to utilize distributed renewable energies (DREs). Given the ongoing transformation of distribution system with high penetration of DREs, coordinating and consuming a large amount of distributed generators (DGs) within one single microgrid has become increasingly infeasible. Interconnecting multiple microgrids as a microgrid cluster is an effective way to improve the operation quality of large-scale DG integration. As the keys to the microgrid clusters, the flexible configurations and coordinated operation among multiple microgrids have not been adequately addressed. In order to solve this problem, a novel architecture for multiple microgrids and its coordinated control schemes are designed. Firstly, the advanced microgrid interface named hybrid unit of common coupling (HUCC) is designed and utilized in replacement of the conventional point of common coupling (PCC). The HUCC employs modular multilevel converter (MMC) as its core component and provides both AC and DC interfaces. Then, this paper puts forward a HUCC-based architecture for multiple microgrids where microgrids are grid-connected via the AC interfaces and interconnected via the DC interfaces. Based on the proposed architecture, coordinated control schemes under different operation scenarios are came up with at last. A case study of the HUCC-based multiple microgrids is performed in PSCAD/EMTDC on the basis of the demonstration project in Guangxi, China. The simulation results show that the interconnected microgrids with the proposed architecture and control schemes operates effectively and efficiently under different operation scenarios. The

⇑ Corresponding author. E-mail address: [email protected] (W. Huang). http://dx.doi.org/10.1016/j.apenergy.2017.07.023 0306-2619/Ó 2017 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Wu P et al. A novel design of architecture and control for multiple microgrids with hybrid AC/DC connection. Appl Energy (2017), http://dx.doi.org/10.1016/j.apenergy.2017.07.023

2

P. Wu et al. / Applied Energy xxx (2017) xxx–xxx

Nomenclature N M i j PiT

number of transformers number of microgrids transformer index, i 2 ½1; N microgrid index, j 2 ½1; M active power of the i-th transformer (MW)

PiTmax

maximum active power supply capacity of the i-th transformer (MW) active power of the j-th microgrid (MW)

PCj j PCmax PEj j PEmax

PGj PLj j U DC

maximum active power supply capacity of the j-th microgrid (MW) active power of the energy storage system (ESS) in the jth microgrid (MW) maximum active power supply capacity of the ESS in the j-th microgrid (MW) total power of distributed generators (DGs) in the j-th microgrid (MW) total loads of the j-th microgrid (MW) DC voltage of the hybrid unit of common coupling (HUCC) in the j-th microgrid (kV)

U j DC j

f MG j

reference DC voltage of the HUCC in the j-th microgrid (kV) frequency of the j-th microgrid (Hz)

f MG

rated frequency of the j-th microgrid (Hz)

j V MG

voltage of the j-th microgrid (kV)

V j MG

rated voltage of the j-th microgrid (kV)

j THDMG

total harmonic distortion of the j-th microgrid limit on microgrid frequency deviation limit on microgrid voltage deviation limit on DC voltage deviation limit on total harmonic distortion of microgrid normal microgrid index adjustment time of the DC voltage of the j-th microgrid (s) allowable DC voltage adjustment time (s) rapid DC voltage change limit on rapid voltage change of DC voltage

e1 e2 e3

d NMG T Uj DC

Dt U 0DC

r

proposed architecture and control schemes not only enhance the large-scale integration of DREs, but realize the optimal use of DGs as well. Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction Distributed renewable energies (DREs) have been regarded as an effective tool to deal with future energy and environment crisis worldwide [1,2]. As the concerns for environmental pollution caused by fossil fuels keep rising and the energy policies sustain improvement, the DG penetration in distribution systems is increasing rapidly [3–5]. At the same time, the configuration and operation of distribution system is also experiencing fast development and transformation [6,7]. Microgrid is an aggregation of distributed generators (DGs), energy storage systems (ESSs) and local loads [8]. It is put forward to solve the ongoing transformation of distribution power system with the integration of various distributed renewable energies [9–11]. Microgrid has both grid-connected and islanded modes [12]. Over the years, microgrid has been proved to be one of the most effective patterns to utilize DGs in medium and low voltage distribution systems [13,14]. Due to its operational flexibility and reliability, microgrid provides an excellent platform where the utility grid, DGs, ESSs and local loads interact positively with each other [15]. Even though DGs can be controlled appropriately in microgrid, a single microgrid is incapable of incorporating a large amount of DGs [16,17]. The multiple microgrids concept is a cluster of microgrids that accommodate a large number of DGs via local integration [18]. On the one hand, multiple microgrids mitigate the limitation on DG penetration of a single microgrid; on the other hand, the system reliability is enhanced due to less influence of every single DG. As a result, the focus on the organization of DGs has shifted from a single microgrid to multiple microgrids in areas that are abundant in DREs [19,20]. Compared with a single microgrid, multiple microgrids reach power balance through the cooperation among all the microgrids. So far, the research on multiple microgrids mainly focuses on sys-

tem control, optimization and management, aiming to improve the reliability and economy of the system via proper control and operation strategies. The control architecture of multiple microgrids is firstly studied in the European Research Project titled ‘‘More Microgrids: Advanced Architectures and Control Concepts for More Microgrids” within the 6th Framework Programme (2002–2006). The control architecture consists of three levels, namely the Distribution Management System (DMS), the Central Autonomous Management Controller (CAMC) and the MicroGrid System Central Controller (MGCC) [21]. Ref. [22] simplifies the control architecture of multiple microgrids to two levels and enhances the system reliability by introducing a reserve capacity mechanism. Ref. [23] introduces a bi-level control framework to reduce the loss-ofload probability in power networks that consists of several residential/commercial communities. In [24], a self-decision making method for load management utilizing multi-agent systems is proposed. The method reduces the peak load of a smart distribution network feeder and is applicable for multiple feeders. Ref. [25] decomposes the dispatching decision among multiple microgrids into inter-temporal decision and real-time scheduling decision to realize the economic operation of the system. Ref. [26] proposes a method of simultaneous allocation of electric vehicle parking lots and distributed renewable resources. Economic objectives and system loss are both considered in the method. In [27], an intelligent energy management system that considers demand response is proposed for trading and managing power in multiple microgrids. The researches above put forward coordination strategies from different perspectives to meet the operation requirements and objectives of multiple microgrids. Nevertheless, the multiple microgrids that are investigated are connected to the utility gird via parallel synchronous connection. Moreover, the coordination approaches of the multiple microgrids are not thoroughly developed, thus limiting the operational flexibility and economy of the multiple microgrids to a certain extent.

Please cite this article in press as: Wu P et al. A novel design of architecture and control for multiple microgrids with hybrid AC/DC connection. Appl Energy (2017), http://dx.doi.org/10.1016/j.apenergy.2017.07.023

P. Wu et al. / Applied Energy xxx (2017) xxx–xxx

The interconnection of multiple microgrids is an appropriate way to expand and organize the large-scale integration of DREs. It allows multiple microgrids to exchange power with each other so as to facilitate their interaction and coordination. In [28], an interconnection method based on AC tie lines is proposed to realize coordinated voltage support for normal operation, coordinated frequency control for islanded operation, and local black start for emergency. Ref. [29] utilizes two decomposition methods to solve the security constrained economic dispatch problem in two tieline-connected areas. Information privacy and convergence rate of the two methods are compared. In [30], a self-healing agent is developed within the network tertiary controller to ensure the power balance of the two microgrids that are interconnected by a static switch under overloading conditions. However, AC interconnection among multiple microgrids may lead to loop networks or even electromagnetic loop networks in distribution networks, which is undesirable to the safe and optimized operation of the system. Meanwhile, it is also hard to directly interconnect multiple microgrids that have different voltage levels and different control systems via AC tie lines [31]. Applying the multi-terminal DC architectures in multiple microgrids is a structural innovation for multiple microgrids interconnection. But there haven’t been enough related researches. Ritwik Majumder and Gargi Bag firstly proposed the application of voltage-sourced converter (VSC) in the parallel operation of grid-connected multiple microgrids [32]. However, the interaction and coordination among the multiple microgrids were not discussed thoroughly. Ritwik Majumder then proposed to interconnect multiple microgrids with DC cables [33]. But the grid-connected operation was not considered. Ref. [34] presents a framework for implementation, simulation, and evaluation of an oblivious power routing algorithm for clusters of DC microgrids that are connected together through multi-terminal DC system. The algorithm solves the optimal power flow problem while managing congestion and mitigating power losses. Md. Jahangir Hossain et al. put forward a robust distributed control strategy to improve the stability of the multiple microgrids operating in islanded mode by regulating the system power flow [35]. Nevertheless, the operational flexibility of the multiple microgrids was not taken into account. Besides, modular multilevel converter (MMC) is a better choice than VSC in the medium and low voltage distribution systems from the perspectives of efficiency and power quality [36]. Ref. [37] investigates the feasibility of microgrids interconnection and comes up with an electricity cluster-oriented

3

network based on power electronic inverters. But only the islanded operation is considered. Ref. [38] proposes an adaptive droop control strategy for the multi-terminal low voltage distribution system to improve the integration capacity of plug-in electric vehicles. The system architecture is inspiring but microgrids are not discussed. The objective of this paper is to present the architecture and control schemes for multiple microgrids in order to realize the flexible integration and the optimal use of large-scale DREs. In comparison with the previous research articles, the main contributions of this paper are as follows: (1) The advanced unified microgrid interface named hybrid unit of common coupling (HUCC) provides both AC and DC interfaces for microgrid connection; (2) The HUCCbased multiple microgrids architecture reduces the negative interplay and increases the flexible interaction among multiple microgrids so that the ability to integrate large-scale DREs is greatly enhanced; (3) The coordinated control schemes for the HUCCbased multiple microgrids greatly increase the operational flexibility and coordinated operation of multiple microgrids so that the reliability and efficiency of DGs are improved, and the optimal use of DGs is realized. The paper begins with the architecture design of the HUCC-based multiple microgrids and the comparison of it with typical multiple microgrids configuration. Following this, Section 3 presents the basic configuration, operation modes and control schemes of the HUCC. Section 4 introduces the operation of the HUCC-based multiple microgrids during three different types of operation scenarios. The schematic overview of the framework of this paper is shown in Fig. 1. This paper designs the HUCC with hybrid interfaces and the corresponding multiple microgrids architecture. The coordinated control schemes for the HUCC-based architecture under different operation scenarios are designed to achieve different operational objectives within certain operation boundaries. The presented architecture and control schemes could realize the largescale integration of DREs and the optimal use of DGs. 2. An advanced architecture of multiple microgrids with hybrid AC/DC connection 2.1. Typical multiple microgrids configuration One advantage of microgrid is the capability of balancing local loads with local DGs and ESSs. Normally, the capacity of a single microgrid is limited. Hence, there should be multiple microgrids

Fig. 1. Schematic overview of the framework.

Please cite this article in press as: Wu P et al. A novel design of architecture and control for multiple microgrids with hybrid AC/DC connection. Appl Energy (2017), http://dx.doi.org/10.1016/j.apenergy.2017.07.023

4

P. Wu et al. / Applied Energy xxx (2017) xxx–xxx

of a typical multiple microgrids configuration is shown in Fig. 2 [39]. As Fig. 2 shows, four microgrids are interconnected via AC lines in the distribution network. There are two types of relation among the three feeders. Feeder A, where MG1 and MG2 are directly connected, and Feeder B, where MG3 is directly connected, exchange power with the same power source, therefore they are referred to as same-source feeders; Feeder A (B) and Feeder C, where MG4 is directly connected, exchange power with different power sources, thus they are referred to as different-source feeders. Even though it is a natural way to organize multiple microgrids via AC connections, such a configuration has some deficiencies in further exploring the potential of multiple microgrids. The AC interconnected multiple microgrids is lacking in flexibility, especially during the switch of operation modes. Furthermore, it lacks controllability and faces with synchronization problems. Moreover, since the microgrids are coupled, the mutual electromagnetic influence is unavoidable especially under faults. 2.2. HUCC-based multiple microgrids Fig. 2. A typical multiple microgrids configuration.

in areas that are abundant in DREs. In order to make the most of DREs and acquire better operating characteristics, the microgrids are naturally interconnected via AC lines. The schematic diagram

To overcome the inherent defects of the AC interconnected multiple microgrids, a novel architecture for multiple microgrids based on the HUCC is proposed. In this architecture, multiple microgrids are interconnected via DC lines and the AC connection to the host utility grid is preserved at the same time, as is shown in Fig. 3.

Fig. 3. HUCC-based multiple microgrids.

Please cite this article in press as: Wu P et al. A novel design of architecture and control for multiple microgrids with hybrid AC/DC connection. Appl Energy (2017), http://dx.doi.org/10.1016/j.apenergy.2017.07.023

5

P. Wu et al. / Applied Energy xxx (2017) xxx–xxx

It can be seen in Fig. 3 that the highlight of the proposed architecture is the HUCC. It is utilized as the unified microgrid interface in replacement of the conventional point of common coupling (PCC) and it provides both AC and DC connection. Since the MMC is employed as the core component of the HUCC, the controllability and flexibility of the multiple microgrids are enhanced. The detailed configuration and control schemes of the HUCC will be introduced in Section 3. Although hybrid connections are provided by the HUCC, the DC interface is just for the interconnection of multiple microgrids and the AC interface is only for the connection between multiple microgrids and the host utility grid. In fact, the HUCC-based architecture is derived from the typical configuration in Fig. 2 by complying some specific rules of connection. The rules are as follows: (1) For each feeder of the system, select the microgrid that has the highest generation capacity of distributed energy from the directly connected microgrids as the Secondary Main Microgrid. Meanwhile, the other microgrids are denoted as the Backup Microgrids. (2) For same-source feeders, select the microgrid that has the highest generation capacity of distributed energy from the Secondary Main Microgrids as the Main Microgrid. Meanwhile, the other Secondary Main Microgrids are denoted as the Backup Microgrids. (3) During normal operation, the Main Microgrids are capable of providing AC connections to the host utility grid while such function of the Backup Microgrids is disabled, as the dashed lines show in Fig. 3. The Backup Microgrid will enable its AC interface and serve as a Main Microgrid only when the Main Microgrid is unable to function normally. (4) All the microgrids are connected to each other via the DC interfaces of the HUCCs. (5) It is worthwhile noting that the already existing AC lines are utilized instead of DC lines to connect the multiple microgrids to the host utility grid. Such a design avoids the high cost of system transformation. Meanwhile, the controllability of the multiple microgrids is not diminished. Compared with the AC interconnected multiple microgrids, the HUCC-based architecture possesses many advantages: (1) Asynchronous interconnection; (2) Higher control dimensions; (3) Better operational flexibility: various combinations of connection modes and control schemes of the HUCCs; (4) Less electromagnetic coupling; (5) Improved DRE integration capacity: optimized power sharing through advanced control schemes. 3. Hybrid unit of common coupling 3.1. Basic configuration As is mentioned in Section 2, the HUCC is utilized as the unified microgrid interface. It is the core of the proposed architecture and control schemes. Fig. 4 illustrates the basic configuration and components of the HUCC. As Fig. 4 shows, the HUCC consists of three buses, namely the HUCC bus, the AC bus and the DC bus. The HUCC bus is the bridge between the microgrid and the other two buses. The AC bus and the DC bus provide the AC interface and the DC interface separately. The AC interface is intended for the connection to the host utility grid while the DC interface is only for the connection to other microgrids. Between the DC bus and the HUCC bus, a MMC is employed as the rectifier. Meanwhile, an AC breaker and a DC

Fig. 4. Hybrid unit of common coupling.

breaker are deployed near the AC bus and the DC bus. Furthermore, an ESS is connected to the HUCC bus to facilitate the power exchange among the multiple microgrids. It is also worth mentioning that the superconducting magnetic energy storage device is suitable for this configuration due to the direct access to the DC bus. 3.2. Operation modes and corresponding control schemes On the basis of the basic configuration, the various operation modes of the HUCC are determined by the different combinations of switch states and MMC control schemes. In terms of switch states, the HUCC is capable of providing both single and hybrid connection modes. In the single connection mode, only the AC (DC) breaker is switched on, whereas both the AC and the DC breakers are switched on in the hybrid connection mode. According to the connection rules introduced in Section 2, the general switch states of the multiple microgrids during normal and emergency operation are listed below. As Table 1 shows, the DC switch is on during most operation scenarios except for the black start and the islanding. In order to increase the operational flexibility of the multiple microgrids, several MMC control schemes could be adopted. Meanwhile, as a supplement of the MMC control, the ESS control schemes add to the system controllability. Generally, the ESS supports the voltage stability and the power exchange of the system. The control schemes of the MMC and the ESS are listed in Table 2. It must be noted that the switch states, MMC control schemes and the ESS control schemes must be well coordinated so that the flexible operation of the multiple microgrids can be achieved. In general, a certain operation scenario determines the network architecture of the multiple microgrids, therefore deciding the switch states of all the HUCCs. At the same time, the MMC control schemes are chosen to meet the operational objectives corresponding to the certain scenario. Then the ESS control schemes are selected to coordinate with the MMC control schemes. Furthermore, the control schemes of the microsources in every microgrid are determined by the MMC in coordination with the system

Please cite this article in press as: Wu P et al. A novel design of architecture and control for multiple microgrids with hybrid AC/DC connection. Appl Energy (2017), http://dx.doi.org/10.1016/j.apenergy.2017.07.023

6

P. Wu et al. / Applied Energy xxx (2017) xxx–xxx

Table 1 Switch states of the multiple microgrids. Microgrid (MG)

Normal operation

Main MG Backup MG Islanded MG (Black Start or Islanding) Islanded MG (Connecting to the system)

Emergency operation

AC switch

DC switch

AC switch

DC switch

On/Off Off n n

On On n n

Hold Hold Off Off

Hold Hold Off On

Table 2 Control schemes of the MMC and the ESS. Control schemes

MMC ESS Microsource Control Objectives

UDC p

Set up DC voltage

PQ p p p Fixed power output

Droop P-UDC p

Vf P-f, Q-V p p

Power adjustment and balance

p p p Set up voltage and frequency

4.1. System control overview The coordination of the multiple microgrids relies on the coordination of the system control. Despite the fact that the particular control schemes under different operation scenarios may vary, the architecture of the system control remains the same. The overall control configuration of the multiple microgrids is illustrated in Fig. 6. It can be seen in Fig. 6 that the coordination of the multiple microgrids is based on the communication technology. Once the operation scenario and its operational objectives are determined, the control signals for the MMCs, ESSs and the switches are transmitted to all the HUCCs. Meanwhile, the control signals for the microsources in every microgrid are managed by the corresponding MMC controller. Based upon the universal architecture of the system control, the specific control schemes under three different operation scenarios are described as follows. 4.2. Grid-connected operation

Fig. 5. Logical relation of control.

operational objectives. The control schemes of the microsources in the microgrid are shown in Table 2. The logical relation of the above control schemes is illustrated in Fig. 5. Several particular combinations of these control schemes will be introduced in Section 4. 4. System control of the HUCC-based multiple microgrids during different operation scenarios The function of the HUCC is to increase the operational flexibility and coordination of the multiple microgrids. To realize this goal, the switch states, MMC control schemes, ESS control schemes and the microsource control schemes must be well coordinated as introduced in Section 3. Three typical operation scenarios for the multiple microgrids, namely grid-connected operation, islanded operation and emergency operation, are presented in this section. The network architecture, operational objectives and boundaries, and the control schemes corresponding to every operation scenario are explained as well.

4.2.1. Network architecture To increase the integration capacity of DREs and make full use of them, grid-connected operation is usually adopted for the multiple microgrids. On account of the connection rules presented in Section 2, the network architecture for grid-connected operation is derived. It is shown in Fig. 3 where the Main Microgrids are connected to the host utility gird via the AC interfaces of the HUCCs and all the microgrids are interconnected via the DC interfaces. During gird-connected operation, the AC connection to the host utility grid not only helps to stabilize the voltage and frequency of the Main Microgrids, but creates a bi-directional power path for DGs as well. 4.2.2. Operational objectives and boundaries The concept of interconnected multiple microgrids is proposed to reinforce the coordination among multiple microgrids so as to solve the problem of vast DG penetration. Hence, the general operational objective during grid-connected operation is to realize the flexible and optimized power distribution among the multiple microgrids. To realize this objective, the system power flow needs to be optimized with the consideration of system loss. Meanwhile the utilization of DGs should be increased as much as possible. The overall objective could be decomposed into two specific objectives, which are:

Please cite this article in press as: Wu P et al. A novel design of architecture and control for multiple microgrids with hybrid AC/DC connection. Appl Energy (2017), http://dx.doi.org/10.1016/j.apenergy.2017.07.023

P. Wu et al. / Applied Energy xxx (2017) xxx–xxx

7

Fig. 6. Overall control configuration of the multiple microgrids.

(1) Optimal power flow of the multiple microgrids (least system loss); (2) Maximum utilization of DGs. In order to achieve the optimal power flow, microgrids with abundant DRE generation utilize their DGs as much as possible and satisfy their local power demand first. Then the HUCCs are controlled to transfer the extra power to the microgrids with power deficit. The excess/deficit power of the multiple microgrids system is absorbed/provided by the host utility grid. It should be noticed that, system loss is taken into account in scheduling the DRE generation and power flow. While achieving these operational objectives, several operation boundaries should be met. The power limits of the transformers, MMCs and the ESSs should not be exceeded. At the same time, high

power quality of the system should be maintained. The boundaries are as follows:

8 i P 6 PiTmax > > > T > > j j > > > PC 6 PCmax > > j j > > < PE 6 PEmax PTj þ PCj  PLj þ PEj þ PGj ¼ 0 > >





> > j j j j j j > >

f MG f MG

;

V MG V MG

;

UDC UDC

6 e ; e ; e > 1 2 3 >
f jMG

V jMG

UjDC
> > > > : j THDMG 6 d

ð1Þ

As for the values of the constants in Eq. (1), the IEEE standard 5191992 could be referred to regarding the system harmonic distortion; the IEC standard 61000-2-2 and the EN standard 50160 could

Please cite this article in press as: Wu P et al. A novel design of architecture and control for multiple microgrids with hybrid AC/DC connection. Appl Energy (2017), http://dx.doi.org/10.1016/j.apenergy.2017.07.023

8

P. Wu et al. / Applied Energy xxx (2017) xxx–xxx

be referred to regarding the system frequency deviation; and the IEEE standard 1547-2003 and the IEC standard 60038 could be referred to regarding the system voltage deviation. 4.2.3. Control scheme With the compliance of the operational objectives and boundaries, the control scheme for grid-connected operation is proposed. The coordination of the control schemes is listed in Table 3. From the perspective of system reliability, droop control is adopted by the MMCs in the multiple microgrids, where every MMC contributes to the stability of DC voltage and shares power according to its capacity. Since the frequency and voltage of the Main Microgrids are kept constant via the connection to the host utility grid, the ESSs and microsources of the Main Microgrids should work on PQ control to supply fixed power to the microgrids. Meanwhile, the ESSs and microsources of the Backup Microgrids should work on droop control in cooperation with the corresponding MMCs so as to stabilize the voltage and frequency of the Backup Microgrids and share the power demand together. As for the slave microgrids, the MMCs should work on Vf control to set up the voltage and frequency of the microgrids. In the meantime, the ESSs and the microsources should work on PQ control. The control structure of the MMC is illustrated in Fig. 7. As is shown in Fig. 7, the reference values and the actual values of the control variables are input into the MMC controller. After the calculation of the outer-loop controller, new reference variables are sent into the inner-loop controller for further calculation. Finally, the modulation of the MMC is adjusted as demanded. 4.3. Islanded operation 4.3.1. Network architecture Islanded operation is another common operation scenario for the multiple microgrids. During islanded operation, the connection to the host utility grid is cut off, but the multiple microgrids are still interconnected via the DC interfaces of the HUCCs. Thus the network architecture is similar to that in Fig. 3 except for the disconnection of all the AC lines. 4.3.2. Operational objectives and boundaries During islanded operation, the voltage and frequency support from the host utility grid is cut off and the power is exchanged only among the multiple microgrids. Hence, the general operational objective during islanded operation is to reinforce the stability of the multiple microgrids in case of large voltage or frequency deviation. To realize this objective, the system power needs to be balanced and the power fluctuation needs to be minimized. Meanwhile, the voltage and frequency of the microgrids should remain stable. The overall objective could be decomposed into two specific objectives, namely: (1) Suppress the fluctuation of active and reactive power; (2) Stabilize the voltage and frequency of the microgrids. To stabilize the voltage and frequency of the microgrids, power balance of the system must be maintained constantly. Firstly, local

Fig. 7. Control structure of the MMC.

power demand should be met as far as possible. Then, the extra power is transferred from the microgrids with abundant DRE generation to the microgrids with power deficit via the HUCCs. At the same time, the ESSs should adjust their operation to mitigate the fluctuation of DRE generation so as to suppress the system power fluctuation. When achieving these operational objectives, some operation boundaries should be met. The power limits of the MMCs and the ESSs should not be exceeded. In the meantime, large DC voltage deviation and poor power quality of the microgrids should be avoided. The boundaries are as follows.

8 j > PCj 6 PCmax > > > > j > > PEj 6 PEmax > > > > j j j j j > < PT þ PC  PL þ PE þ PG ¼ 0
j j

U U
>
DC j DC
6 e3 > >
UDC
> > > > > j > > THDMG 6d > > :

ð2Þ

4.3.3. Control scheme With the compliance of the operational objectives and boundaries, the control scheme for islanded operation is proposed. The coordination of the control schemes is listed in Table 4. It can be seen that the designed control schemes are almost the same as those adopted during grid-connected operation. The main difference lies in the ESS and the microsource control schemes of the Main Microgrids. Since, the voltage and frequency support

Table 3 Control scheme during grid-connected operation. Microgrid (MG)

Main MG Backup MG Slave MGa a

Switch AC

DC

On Off Off

On On On

MMC control

ESS control

Microsource control

Droop Droop Vf

PQ Droop PQ

PQ Droop PQ

Microgrid whose voltage and frequency are controlled by MMC.

Please cite this article in press as: Wu P et al. A novel design of architecture and control for multiple microgrids with hybrid AC/DC connection. Appl Energy (2017), http://dx.doi.org/10.1016/j.apenergy.2017.07.023

9

P. Wu et al. / Applied Energy xxx (2017) xxx–xxx Table 4 Control scheme during islanded operation. Microgrid (MG)

Main MG Backup MG Slave MG

Switch AC

DC

Off Off Off

On On On

from the host utility grid is cut off, the ESSs and the microsources should work on droop control to set up the voltage and frequency of the Main Microgrids during islanded operation. In the meantime, the control structure of the MMC is kept the same as that during gird-connected operation as shown in Fig. 7. 4.4. Emergency operation 4.4.1. Network architecture Besides normal operation, abnormal events may occur in the multiple microgrids, such as the black start, microgrid islanding, and microgrid recovery, which are summarized as emergency operation. The network architectures during emergency operation vary greatly. As for the black start, all the microgrids are islanded and the connections to other microgrids and the host utility grid are disabled. As for the microgrid islanding and the microgrid recovery, the isolated microgrid is islanded while the other microgrids keep interconnected. 4.4.2. Operational objectives and boundaries During emergency operation, the transient stability of the system is vital. Thus the general operational objective during emergency operation is to strengthen the transient stability of the multiple microgrids in case of further accidents. To realize this objective, small system inertia should be avoided and the transient process should be mitigated. The overall objective could be decomposed into two specific objectives, which are: (1) Increase the system inertia; (2) Reduce the impact of the transient process. While achieving these operational objectives, several operation boundaries should be met. The power limits of the MMCs and the ESSs should not be exceeded. The transient behavior of the DC voltage should be within the acceptable range. In addition, the operation of the normal microgrids should not be influenced. The boundaries are as follows:

8 j j PC 6 P Cmax > > > > j j > > PE 6 P Emax > > > > < U 0DC < r T Uj DC 6 Dt > > >
j

j
> > j j > f NMG

V NMG V NMG
>

f NMG


6 e1 ; e2 > > j j > :
f NMG

V NMG


ð3Þ

where the IEC standard 61000-2-2 and the EN standard 50160 could be referred to regarding the values of Dt and r. 4.4.3. Control scheme Control schemes vary in terms of different emergency incidents. The control schemes aim to increase the transient stability of the multiple microgrids. For the black start, every microgrid is isolated. Thus the MMCs don’t work, while the ESSs and the microsources work on the black start control scheme.

MMC control

ESS control

Microsource control

Droop Droop Vf

Droop Droop PQ

Droop Droop PQ

As for the microgrid islanding caused by fault, the MMC of the isolated microgrid does not work but the ESS and the microsources are controlled to suppress the fault current. In the meantime, the other microgrids keep the normal operation with the setting values adjusted. For the microgrid recovery, the ESS and the microsources of the recovered microgrid work on islanded control schemes, namely droop control or master-slave control, while the MMC begins to start. The MMC firstly increases its DC voltage to the set point, then it raises the output power to the set value after its connection to the system. With the start process of the MMC, the recovered microgrid is connected to the system again, meanwhile the other microgrids keep the normal operation. 5. Case study The simulation model built in this paper is derived from the demonstration project of multiple microgrids in Sanli No. 1 Junior High School, Guangxi, China, where three independent microgrids are synchronously interconnected via the PCCs. On the basis of the demonstration project, the HUCC is adopted, the connection rules in Section 2 are applied and a fourth microgrid is added to the simulation model. The model is built in PSCAD/EMTD and its structure is illustrated in Fig. 8. MG1, MG2 and MG4 are the three microgrids of the demonstration project. They contain distributed generation apparatuses (PVs and wind turbines), energy storage devices and local loads. MG3 is the newly added microgrid that only contains energy storage devices and local loads. To testify the validity of the proposed architecture and control schemes, simulations are carried out under different operation scenarios. The main parameters of the model are given in Tables 5 and 6. Besides, to testify that the system operates within the operation boundaries during all operation scenarios, reference values for certain boundary constants are selected according to the standards mentioned in Section 4. The boundary for frequency deviation (e1 ) is set to 2% (absolute value), namely 1 Hz, according to the IEC standard 61000-2-2; the boundary for microgrid voltage deviation (e2 ) is set to 10% (absolute value) according to the IEC standard 60038; the boundary for DC voltage deviation (e3 ) is set equal toe2 due to the direct connection between the converters and the microgrids (if the control ability of the converter is considered, e3 could be set larger); the boundary for total voltage harmonic distortion (d) is set to 5% and the single harmonic distortion should not exceed 3% according to the IEEE standard 519-1992; and the boundary for the allowable DC voltage adjustment time (Dt) is set to 0.5 s according to the IEC standard 61000-2-2. The boundary constant values are listed in Table 7. 5.1. Grid-connected operation During grid-connected operation, MG1 and MG4 serve as the Main Microgrids while MG2 and MG3 serve as the Backup Microgrids. The extra power generated in the Main Microgrids are transmitted to the other microgrids and to the host utility grid via the HUCCs. To simplify the simulation without diminishing the

Please cite this article in press as: Wu P et al. A novel design of architecture and control for multiple microgrids with hybrid AC/DC connection. Appl Energy (2017), http://dx.doi.org/10.1016/j.apenergy.2017.07.023

10

P. Wu et al. / Applied Energy xxx (2017) xxx–xxx

Fig. 8. Simulation model structure.

Table 5 Reference values of the simulation model. Microgrid

AC voltage (MG)

Frequency (MG)

DC voltage (MMC)

Active power (MMC)

Reactive power (MMC)

MG1 MG2 MG3 MG4

10.5 10.0 10.0 10.5

50 50 50 50

5.0 5.0 5.0 5.0

0.60 MW 0.70 MW 0.25 MW 0.40 MW

0 0 0 0

kV kV kV kV

Hz Hz Hz Hz

kV kV kV kV

Table 6 Rated values of the simulation model. Microgrid

Level (MMC)

Capacity (MMC)

Capacity (Transformer)

Local loads

MG1 MG2 MG3 MG4

21 21 21 21

1.0 MW 1.0 MW 1.0 MW 1.0 MW

2.0 2.0 1.5 1.5

0.40 MW 0.90 MW 0.45 MW 0.20 MW

PV

Wind turbine

Energy storage

0.4 MW 0.1 MW n 0.2 MW

0.4 MW n n 0.3 MW

0.2 MW 0.1 MW 0.2 MW 0.1 MW

Microgrid

MG1 MG2 MG3 MG4

MVA MVA MVA MVA

Power generation

Total power

1.0 MW 0.2 MW 0.2 MW 0.6 MW

Please cite this article in press as: Wu P et al. A novel design of architecture and control for multiple microgrids with hybrid AC/DC connection. Appl Energy (2017), http://dx.doi.org/10.1016/j.apenergy.2017.07.023

11

P. Wu et al. / Applied Energy xxx (2017) xxx–xxx Table 7 Boundary constant values.

Table 8 Operation indexes during grid-connected operation.

Boundary constants

e1

e2

e3

d

Dt

Value

2%

10%

10%

5%

0.5 s

Microgrids (MG)

Frequency deviation

AC voltage deviation

DC voltage deviation

THD (Voltage)

MG1 MG2 MG3 MG4

0.3% 0.4% 0.6% 0.3%

0.6% 0.8% 0.9% 0.6%

3.9%

0.9% 1.1% 1.3% 0.9%

Fig. 9. Simulation results during grid-connected operation. (a) Output DC power of the HUCCs. (b) DC voltage of the HUCCs. (c) AC voltage and frequency of the microgrids. (d) Output AC power of the main microgrids.

Fig. 10. Simulation results during grid-connected operation with a small disturbance. (a) Output DC power of the HUCCs. (b) DC voltage of the HUCCs. (c) AC voltage and frequency of the microgrids. (d) Output AC power of the main microgrids.

Please cite this article in press as: Wu P et al. A novel design of architecture and control for multiple microgrids with hybrid AC/DC connection. Appl Energy (2017), http://dx.doi.org/10.1016/j.apenergy.2017.07.023

12

P. Wu et al. / Applied Energy xxx (2017) xxx–xxx

effectiveness, the reactive power of all the MMCs are set to 0, and the power received by the Backup Microgrids is set almost equal to the output power of the Main Microgrids considering the system transmission loss. In addition, the MMCs of MG1, MG2 and MG4 work on droop control while the MMC of MG3 works on Vf control. The output power and the voltage of every HUCC are monitored, as well as the AC voltage and the frequency of every microgrid. The simulation results are given in Fig. 9. As is illustrated in Fig. 9, the power generated by DGs is exchanged among the four microgrids as designed during gridconnected operation. Besides, the DC voltage of every HUCC is kept at 5 kV while the frequency and voltage of every microgrid are kept steady at the rated values. The maximum frequency deviation, voltage deviation and harmonic distortion of the multiple microgrids are calculated and shown in Table 8. By comparing Table 8 with Table 7, it can be seen that the operation boundaries are all satisfied. As mentioned above, the reference output power of every microgrid is designed to simplify the simulation, hence the power exchange between the multiple microgrids and the host utility grid is relatively small, as Fig. 9(d) shows. Nevertheless, numerous simulations have proved that extra power generated by the multiple microgrids can be utilized by the host utility grid effectively during grid-connected operation. In order to further test the proposed architecture and control schemes, a small disturbance is added to the simulation model during grid-connected operation. At 0.7 s, a total DRE generation of 0.2 MW in MG1 is out of service. The following operation of the multiple microgrids is monitored and the corresponding simulation results are given in Fig. 10. As Fig. 10 shows, due to the power decrease in MG1 at 0.7 s, the system DC voltage decreases. At the same time, the droop control of the MMCs starts to adjust the output power accordingly in order to balance the system power. After a short time of adjustment, a new operation point is reached where the interconnection DC voltage is lowered and the output power of the microgrids is adjusted. The simulation shows that the multiple microgrids is tolerant of small disturbances during grid-connected operation and it is able to make adjustments automatically owing to the proposed control schemes.

5.2. Islanded operation During islanded operation, the AC connections of the HUCCs are cut off, therefore the total power generated by DGs are only exchanged among the four microgrids. The MMCs work on the same control schemes as grid-connected operation, nevertheless, the ESS control schemes and the microsource control schemes of MG1 and MG4 switch to droop control so as to establish the voltage and frequency of each microgrid. The output power and the voltage of every HUCC are monitored, as well as the AC voltage and the frequency of every microgrid. The simulation results are given in Fig. 11. It can be seen in Fig. 11 that similar to grid-connected operation, the proposed control schemes during islanded operation work appropriately to distribute power among the multiple microgrids through the HUCCs. Not only is the desired power dispatch satisfied, but also the stable operation of every microgrid is guaranteed. The maximum frequency deviation, voltage deviation and harmonic distortion of the multiple microgrids are calculated and shown in Table 9. By comparing Table 9 with Table 7, it can be seen that the operation boundaries are all satisfied. Similar to grid-connected operation, a small disturbance is also added to the simulation model during islanded operation. At 0.7 s, a total DRE generation of 0.2 MW comes into service in MG1. The

Fig. 11. Simulation results during islanded operation. (a) Output DC power of the HUCCs. (b) DC voltage of the HUCCs. (c) AC voltage and frequency of the microgrids.

Table 9 Operation indexes during islanded operation. Microgrids (MG)

Frequency deviation

AC voltage deviation

DC voltage deviation

THD (Voltage)

MG1 MG2 MG3 MG4

0.5% 0.5% 0.7% 0.6%

0.8% 0.8% 0.9% 0.7%

4.5%

1.2% 1.4% 1.5% 1.2%

following operation of the multiple microgrids is monitored and the corresponding simulation results are given in Fig. 12. As is shown in Fig. 12, due to the power increase in MG1 at 0.7 s, the system DC voltage increases. Consequently, the droop control of the MMCs reacts to the DC voltage change and resets the output power. After a short dynamic process, system power is balanced and a new operation point is reached. The simulation shows that the multiple microgrids is capable of dealing with small disturbances automatically during islanded operation. 5.3. Emergency operation To examine the transient characteristics of the proposed architecture and control during emergency operation, a permanent line to ground fault is applied in MG3, which leads to its instant

Please cite this article in press as: Wu P et al. A novel design of architecture and control for multiple microgrids with hybrid AC/DC connection. Appl Energy (2017), http://dx.doi.org/10.1016/j.apenergy.2017.07.023

13

P. Wu et al. / Applied Energy xxx (2017) xxx–xxx

Fig. 12. Simulation results during islanded operation with a small disturbance. (a) Output DC power of the HUCCs. (b) DC voltage of the HUCCs. (c) AC voltage and frequency of the microgrids.

isolation. After the outage of MG3, the power among the remaining microgrids are reallocated through droop control to reach a new steady state. The output power and the voltages of the remaining HUCCs are monitored, along with the AC voltages and the frequencies of the remaining microgrids. The simulation results are given in Fig. 13. As is illustrated in Fig. 13, MG3 is isolated from the multiple microgrids at 0.7 s, after which the DC voltage rises due to the temporary power imbalance. Correspondingly, droop control of the remaining HUCCs begins to reallocate the power in the system. After a short time of adjustment, the new power balance is achieved and the remaining HUCCs reach a new steady state. It is worthwhile mentioning that the adjustment process is smooth and fast, which ensures the transient stability of the system. Furthermore, the steady operation of the normal microgrids is not influenced during emergency operation. The maximum frequency deviation and voltage deviation of the normal microgrids along with the DC voltage adjustment time are calculated and shown in Table 10. By comparing Table 10 with Table 7, it can be seen that the system operates within the operation boundaries under emergency conditions. When the fault is eliminated, the isolated microgrid recovers normal operation with the help of the ESS and prepares to resume

Fig. 13. Simulation results during fault microgrid islanding. (a) Output DC power of the HUCCs. (b) DC voltage of the HUCCs. (c) AC voltage and frequency of the microgrids.

Table 10 Operation indexes during fault microgrid islanding. Microgrids (MG)

Frequency deviation

AC voltage deviation

DC voltage adjustment

MG1 MG2 MG4

0.8% 0.7% 0.8%

1.2% 1.0% 1.3%

0.2 s

its connection to the multiple microgrids system. The microgrid recovery process is simulated and the simulation results are given in Fig. 14. As Fig. 14 shows, microgrid recovery is a reverse process to microgrid islanding. The system behavior during these two processes is the same but in opposite ways. For instance, the DC voltage increases during fault microgrid islanding but decreases during microgrid recovery. Nevertheless, the transient features remain the same, which means the system also has high transient stability during microgrid recovery. Besides the emergency operation tests listed above, numerous simulations show that the HUCC-based multiple microgrids has a high level of transient stability.

Please cite this article in press as: Wu P et al. A novel design of architecture and control for multiple microgrids with hybrid AC/DC connection. Appl Energy (2017), http://dx.doi.org/10.1016/j.apenergy.2017.07.023

14

P. Wu et al. / Applied Energy xxx (2017) xxx–xxx

The coordination among multiple microgrids is also strengthened. (2) Owing to the proposed control schemes, the HUCC-based multiple microgrids reacts fast to any small changes in the system and is able to smoothly transfer the system to a new operation point during normal operations. The system reliability and stability are improved. (3) With the proposed architecture and control schemes, negative impact of emergency events on the multiple microgrids is reduced. Meanwhile the multiple microgrids is able to adjust itself quickly to reach a new steady state during emergency operation. The transient stability of the system is increased due to the structural and control features of the HUCC. (4) Due to the technical improvements listed above, the ability of multiple microgrids to integrate large-scale DREs is enhanced and the optimal use of DGs is realized. Given the facts that the DG penetration is increasing rapidly and a growing number of microgrids are put into operation, the HUCCbased multiple microgrids is going to play an important role in integrating and utilizing DGs, as well as coordinating multiple microgrids. Further research will be carried out on the basis of the HUCC-based architecture and control schemes. It contains the following three aspects of work:

Fig. 14. Simulation results during microgrid recovery. (a) Output DC power of the HUCCs. (b) DC voltage of the HUCCs. (c) AC voltage and frequency of the microgrids.

6. Conclusion The concept of multiple microgrids is put forward to enhance the large-scale integration of DREs and realize the optimal use of DGs. In order to achieve these goals, the flexible interaction and coordinated operation among multiple microgrids must be enhanced. To solve this problem, a novel design of architecture and control for multiple microgrids is proposed in this paper. The presented architecture is based on the design of an advanced microgrid interface named HUCC. As the replacement of the conventional PCC of microgrid, the HUCC consists of an ESS, an AC interface for the connection to the host utility grid, and a DC interface for the interconnection of multiple microgrids via the MMC. Based on the structural characteristics of the HUCC-based multiple microgrids, coordinated control schemes under different operation scenarios are then proposed. Simulations are carried out on the basis of the demonstration project in Guangxi, China to testify the effectiveness of the proposed architecture and control schemes. Some key findings of our research are concluded as follows. (1) Due to the hybrid-interface design of the HUCC, the controllability of the HUCC-based multiple microgrids is significantly enhanced. The HUCC-based multiple microgrids is able to distribute the DRE generation appropriately and improve the utilization of DGs during normal operations.

(1) The development of standards and specifications regarding the HUCC. (2) Research on key technologies that facilitate the optimal use of DGs including control strategies, protection schemes, energy management and so on. Currently, we are working on the control strategies for optimized energy management and use during grid-connected operation. (3) The application and demonstration of the HUCC-based multiple microgrids. We intend to apply the HUCC-based multiple microgrids to the demonstration project in Guangxi, China so as to further improve the operation of the existing multiple microgrids and increase the types and quantities of the integrated DREs.

Acknowledgement This work was supported by the Research Project of Chinese Ministry of Education (113023A) and sponsored by Shanghai Sailing Program (17YF1410200). The authors also gratefully acknowledge the financial support of China Scholarship Council. References [1] Lv T, Ai Q. Interactive energy management of networked microgrids-based active distribution system considering large-scale integration of renewable energy resources. Appl Energy 2016;163:408–22. [2] Jin X, Mu Y, Jia H, Wu J, Jiang T, Yu X. Dynamic economic dispatch of a hybrid energy microgrid considering building based virtual energy storage system. Appl Energy 2017;194:386–98. [3] Fang X, Yang Q, Wang J, Yan W. Coordinated dispatch in multiple cooperative autonomous islanded microgrids. Appl Energy 2016;162:40–8. [4] Long C, Wu J, Thomas L, Jenkins N. Optimal operation of soft open points in medium voltage electrical distribution networks with distributed generation. Appl Energy 2016;184:427–37. [5] Zhang Y, Gatsis N, Giannakis GB. Robust energy management for microgrids with high-penetration renewables. IEEE Trans Sustain Energy 2013;4: 944–53. [6] Wang C, Song G, Li P, Ji H, Zhao J, Wu J. Optimal siting and sizing of soft open points in active electrical distribution networks. Appl Energy 2017;189:301–9. [7] Cao W, Wu J, Jenkins N, Wang C, Green T. Benefits analysis of soft open points for electrical distribution network operation. Appl Energy 2016;165:36–47. [8] Microgrids Lasseter RH, Eng Power, Soc. Winter Meet.. IEEE, vol. 1. IEEE 2002;2002:305–8.

Please cite this article in press as: Wu P et al. A novel design of architecture and control for multiple microgrids with hybrid AC/DC connection. Appl Energy (2017), http://dx.doi.org/10.1016/j.apenergy.2017.07.023

P. Wu et al. / Applied Energy xxx (2017) xxx–xxx [9] Nikmehr N, Najafi-Ravadanegh S, Khodaei A. Probabilistic optimal scheduling of networked microgrids considering time-based demand response programs under uncertainty. Appl Energy 2017;198:267–79. [10] Ren L, Qin Y, Li Y, Zhang P, Wang B, Luh PB, et al. Enabling resilient distributed power sharing in networked microgrids through software defined networking. Appl Energy 2017. [11] Kou P, Liang D, Gao L. Distributed EMPC of multiple microgrids for coordinated stochastic energy management. Appl Energy 2017;185:939–52. [12] Ouammi A, Dagdougui H, Dessaint L, Sacile R. Coordinated model predictivebased power flows control in a cooperative network of smart microgrids. IEEE Trans Smart Grid 2015;6:2233–44. [13] Parhizi S, Lotfi H, Khodaei A, Bahramirad S. State of the art in research on microgrids: a review. IEEE Access 2015;3:890–925. [14] Hatziargyriou N, Asano H, Iravani R, Marnay C. Microgrids: an overview of ongoing research development and demonstration project. Power Energy Mag 2007. [15] Fathi M, Bevrani H. Statistical cooperative power dispatching in interconnected microgrids. IEEE Trans Sustain Energy 2013;4:586–93. [16] Arefifar SA, Ordonez M, Mohamed Y. Voltage and current controllability in multi-microgrid smart distribution systems. IEEE Trans Smart Grid 2016. [17] Toro V, Mojica-Nava E. Droop-free control for networked microgrids. Control Appl (CCA), IEEE Conf IEEE 2016; 2016. p. 374–9. [18] He M, Giesselmann M. Reliability-constrained self-organization and energy management towards a resilient microgrid cluster. Innov Smart Grid Technol Conf (ISGT), IEEE Power Energy Soc IEEE 2015; 2015. p. 1–5. [19] Lasseter RH. Smart distribution: coupled microgrids. . Proc IEEE 2011;99:1074–82. [20] Haddadian H, Noroozian R. Multi-microgrids approach for design and operation of future distribution networks based on novel technical indices. Appl Energy 2017;185:650–63. [21] Papadogoula I, Oyarzabal J, Hatziargyriou N, Schwaegerl C. Advanced architectures and control concepts for more microgrids. EC Project, Tech. Rep. SES6-019864; 2008. [22] Lü TG, Ai Q, Sun SM, Cheng Y, Zhao YY. Behavioural analysis and optimal operation of active distribution system with multi-microgrids. Proc CSEE 2016;36:122–32. [23] Boroojeni KG, Amini MH, Nejadpak A, Iyengar SS, Hoseinzadeh B, Bak CL. A theoretical bilevel control scheme for power networks with large-scale penetration of distributed renewable resources. Electro Inf. Technol. (EIT), IEEE Int. Conf IEEE 2016; 2016. p. 510–5. [24] Amini MH, Nabi B, Haghifam M-R. Load management using multi-agent systems in smart distribution network. Power Energy Soc. Gen. Meet. (PES), 2013 IEEE, IEEE; 2013. p. 1–5.

15

[25] Jiang R, Qiu X, Li D, Chen G, Wang G, Liao Q. Economic operation of smart distribution network containing multi microgrids and energy storage system. Power Syst Technol 2013;37:3596–602. [26] Amini MH, Moghaddam MP, Karabasoglu O. Simultaneous allocation of electric vehicles’ parking lots and distributed renewable resources in smart power distribution networks. Sustain Cities Soc 2017;28:332–42. [27] Nunna HK, Doolla S. Demand response in smart distribution system with multiple microgrids. IEEE Trans Smart Grid 2012;3:1641–9. [28] Hatziargyriou N. Microgrids: architectures and control. Wiley Online Library; 2014. [29] Amini MH, Jaddivada R, Mishra S, Karabasoglu O. Distributed security constrained economic dispatch. Innov. Smart Grid Technol. (ISGT ASIA), 2015 IEEE, IEEE; 2015, p. 1–6. [30] Pashajavid E, Shahnia F, Ghosh A. Overloading conditions management in remote networks by coupling neighboring microgrids. Power Eng. Conf. (UPEC), 50th Int. Univ IEEE 2015; 2015. p. 1–6. [31] Pei W, Du Y, Li H. Novel solution and key technology of interconnection and interaction for large scale microgrid cluster integration. High Volt Eng 2015;41:3193–203. [32] Majumder R, Bag G. Parallel operation of converter interfaced multiple microgrids. Int J Electr Power Energy Syst 2014;55:486–96. [33] Majumder R. Aggregation of microgrids with DC system. Electr Power Syst Res 2014;108:134–43. [34] Boroojeni K, Amini MH, Nejadpak A, Dragicˇevic´ T, Iyengar SS, Blaabjerg F. A novel cloud-based platform for implementation of oblivious power routing for clusters of microgrids. IEEE Access 2017;5:607–19. [35] Hossain MJ, Mahmud MA, Milano F, Bacha S, Hably A. Design of robust distributed control for interconnected microgrids. IEEE Trans Smart Grid 2016;7:2724–35. [36] Chuco B Watanabe EHA. Comparative study of dynamic performance of HVDC system based on conventional VSC and MMC-VSC Bulk Power Syst. Dyn. Control (iREP)-VIII (iREP), iREP Symp IEEE 2010; 2010. p. 1–6. [37] Koyanagi K Hida Y, Yokoyama R, Nagata S, Nakao K, Hirai T. Electricity cluster-oriented network: a grid-independent and autonomous aggregation of micro-grids. Mod. Electr. Power Syst. (MEPS), Proc. Int. Symp IEEE 2010; 2010. p. 1–6. [38] Pei W, Deng W, Zhang X, Qu H, Sheng K. Potential of using multiterminal LVDC to improve plug-in electric vehicle integration in an existing distribution network. IEEE Trans Ind Electron 2015;62:3101–11. [39] Yuen C, Oudalov A, Timbus A. The provision of frequency control reserves from multiple microgrids. IEEE Trans Ind Electron 2011;58:173–83.

Please cite this article in press as: Wu P et al. A novel design of architecture and control for multiple microgrids with hybrid AC/DC connection. Appl Energy (2017), http://dx.doi.org/10.1016/j.apenergy.2017.07.023