IEEE Electrification - September 2021 - 73

testing inverter control, in particular. Because of this
inherent complexity, the control of microgrids remains
a major challenge.
To overcome this problem, it is necessary to rethink
the structure of models used and to take a multilayered
modeling approach and to conceptualize a microgrid as
an interconnection of physically-interacting dynamical
components in extended state space. Each component is
represented by modeling its internal state variables and
shared variables at the interfaces with other components.
By taking such an approach, it becomes possible
to model interactions between components by representing
dynamics of shared variables and, also, design
internal physical control of the component to meet the
specifications of the shared variables. Then the local primary
control of components is to meet specifications of
the shared variables. If this is done, then the interactions
between the components can interactively be controlled
through " handshaking " between the components,
exchanging the minimal information about these shared
variables and checking the feasibility conditions
amounting to " plug-and-play " self-adjusting components
in an autonomous microgrid. I have pursued in
my research this approach and have shown that the
minimal characterization of interaction variables is in
terms of a triplet of energy, power, and rate of change of
reactive power. It is quite exciting to realize that such
mathematical modeling now fully supports physical
intuition about energy dynamics of interacting components.
Most important is that this modeling of interconnected
dynamics is technology-agnostic since each
component can be represented in terms of its internal
state variables (technology specific) and the interaction
variables only enter higher-level model relevant for
system-level dynamics. This modeling approach simplifies
complex microgrid representation and enables distributed
interactive control for provable performance.
Change of Paradigm: From Today's
Hierarchical Control to Interactive
Modular Electricity Services
In conclusion, relaxing today's rigid assumptions that
require temporal and spatial separation of its different layers
as well as decoupling of real and reactive power
dynamics is essential. It is becoming important to rethink
the principles of modeling for provable control with welldefined
performance objectives. Microgrids are candidates
for intelligent balancing authorities (iBAs), much the same
way as today's balancing authorities (BAs) participate in
operations of a BPS. iBAs are embedded within the BAs
and often are not utility-owned (Figure 5). To support their
orderly integration in the existing BPS, it is critical to
establish protocols for interactive information exchange
between different industry layers, which is transparent
and visible. Much the same way as today's reliability standard
for frequency regulation is measured in terms of area
control error, it has recently been proposed that it is possible
to establish more temporally and spatially granular
measures that characterize in a unified way any iBA. Modeling
in transformed energy/power dynamic state space to
capture interactions between the iBAs in terms of energy,
power, and rate of change of reactive power is shown to be
sufficient to integrate the new entities, including
microgrids, into the legacy BPS and have orderly operation.
For this to be implemented, it is essential to have
control embedded within the iBAs, which can meet the
specifications defined in terms of ranges of energy, power,
and rate of change of reactive power. Microgrids and
Step 4: Receive dispatch
signals from the coordinator
at both energy and
regulation timescales.
Step 5: Convert regulation
signals to internal control
setpoints.
Step 6: Measure the signals
needed for feedback
control and apply the
composite control
designed using information
from steps 4 and 5.
Exogenous
Disturbances
mi (t)
Figure 5. Price-responsive control of DERs. QoS: quality of service.
IEEE Electrification Magazine / SEPTEMBER 2021
73
P∗ (kTt)
i
P∗ (nTs)
i
Anticipated
Prices
Control
Implementation
Droop
P∗ (kTt),
yref (nTs)
i
i
Bid Creation
Coefficients
Flexibility
Computation
Present
Operation
Conditions
Lower Layer Design
and Control
Step 1: Design a composite control
involving stabilization, regulation, and
feed-forward components.
Consumption
Estimations
Electrical and
Non-Electrical
Consumer
ρe [k], ρr [k]
mi (kTt), mi (kTt)
b
ymin, ymax
ii
QoS: Specifications
Step 2: With the embedded internal
control, obtain closed-loop operating
conditions-dependent droop
characteristics.
Step 3: Compute bid functions utilizing
estimations of local disturbances and
their bounds and QoS specifications.
Bid
Functions
"
"
IEEE Electrification - September 2021

Table of Contents for the Digital Edition of IEEE Electrification - September 2021
