IEEE Network - March / April 2017 - page 32

IEEE Network • March/April 2017
0890-8044/17/$25.00 © 2017 IEEE
The uncertainties brought by intermit-
tent renewable generation and uncoordinated
charging behaviors of EVs pose great challenges
to the reliable operation of power systems, which
motivates us to explore the integration of robust
optimization with energy scheduling in V2G net-
works. In this article, we first introduce V2G robust
energy scheduling problems and review the state-
of-the art contributions from the perspectives of
renewable energy integration, ancillary service
provision, and proactive demand-side participa-
tion in the electricity market. Second, for each
category of V2G applications, the corresponding
problem formulations, robust solution concepts,
and design approaches are described in detail
based on the characteristics of problem struc-
tures and uncertainty sets. Then, an adjustable
robust energy scheduling solution is proposed to
address the over-conservatism problem by explor-
ing chance-constrained methods. Results demon-
strate that the proposed algorithm not only can
efficiently shift the peak load and reduce the total
operation cost, but also provide great flexibility in
adjusting the trade-off between economic perfor-
mance and reliable operation. Finally, we present
key research challenges and opportunities.
The smart grid provides an open platform for
integrating every piece of equipment involved
in energy generation, transmission, distribution,
storage, and consumption into a network with
up-to-date information and communication tech-
nologies. As a key component of the smart grid,
the emerging vehicle-to-gird (V2G) technology
can explore the batteries of electric vehicles (EVs)
to reduce energy demand and supply imbalance
by absorbing excess energy during off-peak hours
and delivering it back to the grid when needed.
As a result, V2G networks can benefit the grid by
facilitating the integration of intermittent renew-
able energy sources, enhancing system reliability
and safety through ancillary services, and promot-
ing the demand-side liberalization of the electric-
ity market through demand response and virtual
power plant (VPP) programs [1].
However, due to the dynamic nature of EV
charging time, locations, user behavior, and load
profiles, the large-scale penetration of uncon-
trolled and uncoordinated EVs into power sys-
tems, especially distribution networks, may cause
a high level of volatility and increase potential
sources for system disturbances. Furthermore,
intermittent and fluctuating renewable generation
provides little controllability and predictability,
and poses new challenges in balancing genera-
tion and load. Hence, intelligent energy sched-
uling schemes are required to harness the full
potential of the aforementioned benefits brought
by V2G networks.
Two main methodologies, i.e., stochastic opti-
mization and robust optimization, have been
widely applied in handling data uncertainties in
optimization [2]. Stochastic optimization provides
an effective solution if the uncertain numerical
data follow a well known probability distribution.
However, considering the complex operation
details and various practical constraints, it is diffi-
cult to identify accurate probability distributions
for uncertain factors. Hence, stochastic optimi-
zation based energy scheduling approaches may
not sufficiently address the impacts of uncertain-
ties on the reliability performance.
In comparison, robust optimization can over-
come the aforementioned limitations of stochastic
optimization, and provide the following advantag-
es for V2G energy scheduling [3]:
• It allows a distribution-free model of uncer-
tainties and only requires moderate informa-
tion, which can be implemented more easily
in practical V2G networks.
• The worst-case operation scenarios of V2G
networks have been taken into consideration
during the modeling process, and the gener-
ated solution is proved to be immune against
all possible realizations of the uncertainties.
Realizing robust energy scheduling in V2G net-
works is not trivial. First of all, the computational
complexity increases exponentially with the num-
ber of optimization stages and EVs. It would be
infeasible to take every detail into consideration
as the problem size increases [4]. Second, the
robust version of a tractable energy scheduling
problem is not guaranteed to be tractable, which
mainly depends on the problem structure and
the design of uncertainty sets. Finally, it will take
an unrealistically high price to ensure robustness
when the worst case scenarios are considered
simultaneously for numerous uncertain factors of
EVs and renewable energy sources.
There are existing works that investigated the
robust optimization oriented approaches. A group
coordination-based robust charging strategy and
robust linear optimizations based energy manage-
ment scheme were proposed to solve the energy
scheduling problem in [5] and [6], respectively.
In [7], a robust optimization framework was pro-
posed to solve the frequency regulation capacity
scheduling problem with the consideration of the
performance-based compensation scheme and
the random charging and discharging behaviors.
The robust energy scheduling problem in the sce-
Robust Energy Scheduling in Vehicle-to-Grid Networks
Zhenyu Zhou, Changhao Sun, Ruifeng Shi, Zheng Chang, Sheng Zhou, and Yang Li
Zhenyu Zhou, Changhao
Sun (corresponding author),
and Ruifeng Shi are with
North China Electric Power
Zheng Chang is with the
University of Jyväskylä.
Sheng Zhou is with Tsinghua
Yang Li is with the Potevio
New Energy Co., Ltd.
Digital Object Identifier:
1...,22,23,24,25,26,27,28,29,30,31 33,34,35,36,37,38,39,40,41,42,...100
Powered by FlippingBook