IEEE Electrification Magazine - June 2016 - 53

Utility Grid
To
Loads
Source-Side Converter C

Local Control

Source-Side Converter A
Control Signals

Source-Side Converter B

System Status

Voltage and Current Regulation
Droop Control
Primary Control
MPPT

Scenario-Based Control

Local Control
Converter to
Converter
Communication

SoC
Estimation

Source Depending Functions
dc Bus
Signaling

Adaptive
Droop Control

Power Line
Communication

Decentralized Coordination

System Status

Control Signals

Centralized
Communication

Secondary
Control

Tertiary
Control

Operation
Modes

Real-Time
Optimization

Centralized Control

Local Control
Figure 8. Hiearchical control in a practical dc microgrid.

Thus, not only the controller itself but also decision-making methods have been proposed to achieve specific optimization objectives.

Centralized, Decentralized, or Distributed
Coordination: Scenario-Based Choices
For large-scale dc microgrids, hierarchical control is often
a preferred choice since it offers decoupled behavior
between different control layers. However, hierarchical
control is achieved by simultaneously using local control
of the power electronic interfaces and the coordinated
control of all these components. The secondary and tertiary control levels rely on the cooperation of several or all
local controllers. For this reason, the coordination in the
microgrid will also impact on system stability, reliability,
and performance. According to their different communication modes, coordination methods can be divided into
three categories: centralized, decentralized, and distributed. Figure 9 shows the different operating principles of
these three coordination methods.
Centralized coordination control can be implemented
in dc microgrids by employing a central controller and a

communication network, as shown in Figure 9(a). In smallscale dc microgrids, each unit can be directly regulated by
the central controller via high-bandwidth communication
using the master/slave method. It should be noted that
centralized control provides the best foundation for the
advanced control functionalities and system-level optimization, since all relevant data can be collected and processed within a single controller. However, the cost and
difficulty of implementing centralized control increases
nonlinearly with the increasing number of accessed components. Moreover, the most obvious drawback is that the
control architecture has to face the potential failure of the
central controller and/or key communication links, which
may block the transmission of the commands and result in
a systemic failure. In addition, the emerging issue of cyberattack also needs to be considered, especially for some
mission-oriented applications.
Decentralized coordination control is achieved exclusively by the local controllers, as shown in Figure 9(b). The
obvious advantage of decentralized coordination is its independence from the communication and central controller,
allowing this architecture to offer higher flexibility and
IEEE Electrific ation Magazine / j une 2 0 1 6

Table of Contents for the Digital Edition of IEEE Electrification Magazine - June 2016