Role of Proactive Behaviour Enabled by Advanced Computational
Intelligence and ICT in Smart Energy Grids
Phuong H. Nguyen
1
, Wil L. Kling
1
, Paulo F. Ribeiro
1
,
Ganesh K. Venayagamoorthy
2
and Roel Croes
3
1
Eindhoven University of Technology, Den Dolech 2, Eindhoven, The Netherlands
2
Clemson University, 303-D Riggs Hall, Clemson, U.S.A.
3
Green ICT Foundation, Aldenhof 6711, Nijmegen, The Netherlands
Keywords: Smart Energy Grids, Computational Intelligence, Game Theory, Multi-Agent System, Proactive Behaviour.
Abstract: Significant increase in renewable energy production and new forms of consumption has enormous impact to
the electrical power grid operation. A Smart Energy Grid (SEG) is needed to overcome the challenge of a
sustainable and reliable energy supply by merging advanced ICT and control techniques to interact with the
power grid. In SEG, distributed intelligence plays an important role to alleviate significantly consequences
from uncertain power supply and changing load demand. This paper presents the state-of-the-art utilisations
of distributed intelligence in SEG. Insufficient consideration of so-called proactive behaviour limits such
SEG only to near real-time control functions and local optimisation of particular problems. This paper
addresses a need for having a comprehensive research on anticipatory, change-oriented and self-initiated
capabilities of SEG.
1 INTRODUCTION
As one of the largest and most complex engineering
systems, electrical power grids spread everywhere in
countries to supply electricity for myriad consumers
from hundreds of thousands of producers. Secure
operation of the power grid is crucial as unreliable
performance can lead to grid blackouts such as the
2003 event in Europe (UCTE, 2004) and the 2012
event in India (Enquiry Committee, 2012). However,
significant increase in renewable energy production
(e.g. solar photovoltaic, hydro and wind power) and
new forms of consumption (e.g. heat pumps, electric
vehicles) has enormous impact to the electrical
power grid operation. The accommodation of these
so-called distributed energy resources (DER), with
widely dispersed and highly stochastic natures,
challenges the current power system in processing
burden information, controlling the power flows
properly at the right moments, and especially in
balancing power supply and demand at all times.
Technically, the deviations in the power balance are
caused by two main reasons including imperfect
market participation and insufficient control
capability. The former is the difference between the
pre-scheduled values of power production and
consumption, and the real contributions. In the past,
conventional power plants were dispatched on the
day ahead pretty well based on a sufficient
knowledge of load consumption. The massive
integration of DER makes the power production
being hardly predictable or quite unpredictable while
the load consumption is more active and flexible
(Lopes, Hatziargyriou, Mutale, Djapic, and Jenkins,
2007). The latter is the limitation of the current
control system arranged top-down to maintain grid
stability and adequate bus voltages (Kundur, 1994).
Increasing amounts of distributed, non-dispatchable,
and fluctuating renewable sources reveal serious
grid problems such as overloading, voltage
excursions, and even instability. This classic control
system is insufficient to response timely and adapt
properly to the grid expansion and the significant
participation of DER (Wu, Moslehi, and Bose,
2005).
Smart Energy Grid (SEG) is needed to overcome
the mentioned challenges by merging advanced ICT
and control techniques to interact with the power
grid. This interoperability presents a chance to
optimise system performance by improving the
involvement of and synergy among actors, i.e.
producers, consumers, and network operators. This
trend takes place intensively in both US and Europe
82
H. Nguyen P., L. Kling W., F. Ribeiro P., K. Venayagamoorthy G. and Croes R..
Role of Proactive Behaviour Enabled by Advanced Computational Intelligence and ICT in Smart Energy Grids.
DOI: 10.5220/0004408900820087
In Proceedings of the 2nd International Conference on Smart Grids and Green IT Systems (SMARTGREENS-2013), pages 82-87
ISBN: 978-989-8565-55-6
Copyright
c
 2013 SCITEPRESS (Science and Technology Publications, Lda.)

(Simoes et al., 2012). In SEG, distributed
intelligence plays an important role to alleviate
significantly consequences from uncertain power
supply and changing load demand. However, the
state-of-the-art utilisations of distributed intelligence
in SEG have not been focussed comprehensively in
anticipatory, change-oriented and self-initiated
capabilities. Insufficient consideration of this so-call
proactive behaviour limits SEG to achieve only near
real-time control functions and local optimisation of
particular problems.
This paper discusses generally research trends in
SEG as well as related limitations of the current
network functions and market services. Proactive
behaviour is addressed as a key to improve
performance of the grid and attribute fairly benefits
to involved actors.
2 SEG’S RESEARCH TRENDS
2.1 Network Optimisation
2.1.1 State Estimation and Prediction
Real-time control and operation are playing an
important role to reduce consequences of
intermittency and uncertainty in such new context of
smart grids. These functions require advanced
techniques to not only estimate system state
variables but also predict their trends steps ahead
(Venayagamoorthy et al., 2012). By improving the
monitoring capability of the grid, control action will
be trigger in real-time thus improve system
reliability and stability. While Static State
Estimation (SSE) based on Weighted Least Square
(WLS) provides only a snapshot of the current state
vector (Schweppe and Wildes, 1970),
Figure 1: Illustrates the basic concept of the signals and parameters can that be processed and derived.
Voltage and Current Signals
Analog Conditioning (transducers, low-pass filters) - ADC
SP 1st step Results SP 2nd step Results
Curve Fitting techniques
Sum, LSM, derivatives
Fourier series
Fourier Transform
STFT (DFT, SWDFT, FFT)
Model system
Differential equation
Travelling Waves
Techniques
Time-Frequency
techniques (Wavelet…)
Filter banks
Special filters
Denoising, Notch,
Kalman…
Making decision:
Protection, Control,
Supervision, Planning.
Making decision:
Protection, Control,
Supervision, Planning.
Curve Fitting techniques
Sum, LSM, derivatives
Fourier series
Fourier Transform
STFT (DFT, SWDFT, FFT)
Model system
Differential equation
Travelling Waves
Techniques
Time-Frequency
techniques (Wavelet, …)
Other filter banks
Special filters
Denoising, Notch,
Kalman…
RMS
phase
phasor
frequency
harmonics orders
THD
Frequency Spectrum
Stead
y
-State Components
Time-Varying Components
Fast Transients
Unbalances,
Asymmetries
Probabilistic parameters
Histogram
Scalegram
Impedance
Power, power factor,
energy
Time-Varying Phasors
PQ identification and
classification
Fault location,
incipient defects
Other transformations
(Walsh, Hilbert…)
Probabilistic Analysis
Symmetrical Components
Pattern Recognition
Collective RMS
Other time-frequency
techniques
Normalization
Wavelet Transform
1
2
3
5
6
7
1
2
5
6
2
1
2
7
2
5
7
2
4
5
6
1
1
1
2
3
5
4
5
1
4
5
4
5
2
4
6
4
5
1
2
3
4
5
6
7
1
2
3
4
5
6
1
2
3
4
5
6
7
8
1
1
4
5
6
2
3
2
2
2
4
7
8
2
3
2
3
4
5
8
RoleofProactiveBehaviourEnabledbyAdvancedComputationalIntelligenceandICTinSmartEnergyGrids
83

Dynamic State Estimation (DSE) aims to provide
not only time-varying solutions but also predict the
future operating points of the system (Debs and
Larson, 1970). However, DSE approaches normally
based on Extended Kalman Filter (EKF) need to
collect recursively time-historic data, to update
covariance vectors, and to treat heavy computation
matrices. Computation burden mitigates the state-of-
the-art utilizations of DSE in real large-scale
networks although DSE was introduced several
decades ago. Recent improvement of DSE by using
Unscented Kalman Filter (UKF) can alleviate
significant computation burden while outperform the
former EKF-based method in terms of accuracy and
robustness. This application opens a possibility to
improve monitoring capability in SEF, especially at
the distribution system level in a wide-range of
dynamic conditions.
A fine-grained solution can be obtained by
support from advanced signal processing technique.
By measuring and analysing the signals at different
points of the system the condition of the grid can be
fully assessed. Figure 1 illustrates the basic concept
of signals and parameters that can be processed and
derived in steps. First, three-phase signals are
decomposed into time-varying harmonics and then
these are processed by symmetrical components.
The result allows the engineer to have a unique
means to visualize the nature of time varying
unbalances and asymmetries in power systems.
Improvement of dynamic state estimation and
prediction is important for situation awareness (SA)
that needs for secure and efficient operation of SEG.
SA is defined in (G. K. Venayagamoorthy, 2011) as
the perception of environmental elements within a
volume of time and space, the understanding of their
meaning, and the prediction of their states in the
near future. In near future, SA with its capability of
state estimation and prediction is critical in
distribution networks to provide distribution network
operators insight about their grids.
2.1.2 Stability of SEG
SEG is facing an increasing replacement of
conventional rotating-machine based power
production by decentralised power-electronic based
renewable energy production. Due to the decrement
of available rotating generators, instability in the
grid can be increased. Virtual Synchronous
Generator (VSG) gives an opportunity for inverter-
interfaced units to emulate virtual inertia (“VSYNC
project,” n.d.). Figure 2 illustrates a simplified
model of a VSG.
Emulation of rotational inertia by VSG can be
implemented with adjustment of active power in the
way similar to a synchronous machine in the swing
equation. However, normal capacity of storage is
small compared with the large moment of inertia of
the system. To have and solid effect on the inertia of
the grid, one needs to use multiple units of VSG
together. Figure 3 illustrates relatively impact of
numbers of VSG units on grid stability.
Grid
50Hz
Inertial
emulation
DC link
Electrical
storage
Virutal Synchronous Generator
(VSG)
State of Chargef
PWM
Figure 2: A simplified model of a Virtual Synchronous
Generator.
Figure 3: Contribution of virtual synchronous generators
of damping oscillation of the grid.
2.1.3 Decentralised Control and Operation
Decentralisation becomes an important trend for
control and operation at distribution system level,
although essential centralised systems are still be
used. In this transition, distributed intelligence plays
an important role to handle the consequences from
uncertain and variable power supply and changing
load demand. As an evolution from conventional
artificial intelligence into the mainstream of
distributed systems, distributed intelligence in SEG
is formed by pieces of software with communication
capabilities (Wooldridge and Jennings, 1995). A so-
called agent can simplify the way in which local
entities interact with the power system as they can
bring together reactive, proactive, and social
SMARTGREENS2013-2ndInternationalConferenceonSmartGridsandGreenITSystems
84

behaviour (Li, Poulton and James, 2010). Reactivity
is the agent’s capability to react timely to change
within its environment that needs for real-time
control functions in smart grids such as coordinated
voltage regulation (Nguyen, Myrzik, and Kling,
2008), or power flow management (Nguyen, Kling,
and Myrzik, 2009). Sociality is the agent’s capability
to interact with other intelligent agents that is
preferred for Microgrid applications (Dimeas and
Hatziargyriou, 2005), balancing a local area network
(Kok, 2011), or large-scale integration of DER units
in Virtual Power Plants (Hommelberg, Warmer,
Kamphuis, Kok, and Schaeffer, 2007). The state-of-
the-art utilisations of this technology in SEG exploit
mainly reactive and social capabilities of the
intelligence agent.
Though proactivity is the most important
behaviour to be driven by a set of tendencies, it has
not been considered thoroughly in SEG’s
applications. This fact limits distributed intelligence
to achieve only near real-time control functions and
local optimisation.
2.2 Market Optimisation
2.2.1 Very Short-term Forecasting
Very short-term forecasting is expected to alleviate
the consequences of uncertainty from active small-
scale producers and consumers. Over- or under-
forecasting the wind power generation has different
consequences on the value of wind power generation
in power system (Ortega-Vazquez and Kirschen,
2010). Forecasting tool, therefore, is crucial
functional block to predict the stochastic behaviour
of involved actors. With capability of mapping non-
linear input-output relations, artificial neural
network (ANN) based models are widely accepted
for enabling very-short forecasting tools (Peng,
Hubele and Karady, 1992). With high integration of
renewable energy, on-line training with mutual
information of input data selection is desired to
reduce forecasting error.
2.2.2 Real-time and Scalable Market
Power supply and demand matching is a continuous
system-wide problem that must be solved at all
times. Until now this matching process has been
centrally organised, based on generators which
follow the passive loads in a coordinated way. With
a massive amount of intermittent power sources and
more stochastic patterns of new forms of load, the
uncertainty increases significantly. The
PowerMatcher concept, for instance, is an
application of agent-based technology for power
matching via a bottom-up market approach (ECN,
n.d.). Similar approach with additional capability for
taking also network congestion into account has
been presented in (Greunsven, Veldman, Nguyen,
Slootweg and Kamphuis, 2012). The principle of
this application is that software agents connected to
electricity generating or consuming devices are
involved in an electronic market that also contains
an aggregator, who determines the market
equilibrium between demand and supply via pooling
or an event based mechanism.
Figure 4: Example of bidding curves for power supply-
demand matching.
2.3 System Optimisation
Consideration of the interrelated problems of both
network and market optimisation in a dual way leads
to conflicting interests between the actors involved.
The elaboration of these issues is novel and not
considered in most research works. To achieve an
overall optimal performance of the supply system,
priori knowledge about system states and market
trends will be exploited and suitable strategies for
allocating capacities from the resources will be
formulated via the agent-based platform. One
possibility to achieve optimisation of dual-objectives
is the utilisation of game theory. A combination of a
multi-agent model and game theory to resolve
coalition formation in multilateral trade was
mentioned (Yeung, Poon and Wu, 1999) - (Saad,
Han and Poor, 2011). In a preliminary study, mainly
cooperative game theory to dedicated conflicting
problems in SEG has been developed.
A more generic method needs to be developed
for dealing with complete aspects to help system
overcome local optimisation traps to achieve global
ones. Developed algorithms must take dynamic
characteristics and constraints of the physical grid
into account. Performances of these algorithms will
be compared with conventional optimal power flow
results in term of dynamic and fast event-based
responses.
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
-2000
-1500
-1000
-500
0
500
1000
1500
2000
2500
Price (
λ
) [p.u.]
Power (P) [W]
Device agents bid curves
Fossil Fuel Agent
Freezer agent
CHP agent
Heat Pump Agent
Wind turbine agent
PV agent
RoleofProactiveBehaviourEnabledbyAdvancedComputationalIntelligenceandICTinSmartEnergyGrids
85

3 PROACTIVE BEHAVIOUR OF
SEG
Proactive behaviour with its anticipatory capability
is expected to provide the best strategy to integrate
and attribute conflicting interests among actors. In
the past, this feature has been hardly obtained in
distribution networks due to lack of information and
monitoring capabilities. In near future, SEG will be
enforced with robust, redundant ICT infrastructure
to enrich information thus monitoring capability of
the distribution network will be improved.
On the one hand, distributed control systems
need proactive behaviour to perform distributed
control in real-time or even before a possible event
occurs (before event occurring time). Conventional
approach to the sequence of measuring, detecting
and responding causes always time delays.
Distributed control with very fast communication
can significantly reduce this time delay to enable
near real-time functions. Nevertheless, the remaining
time delay is still a barrier that limits technical
solutions (Wu et al., 2005). Distributed intelligence
with its proactive behaviour can predict unsecure
tendencies from learning its historic data and
exchange information to enrich its priori knowledge
about network situations.
On the other, distributed intelligence was
considered particularly for either technical or
commercial aspects of the power grid. Naturally,
those agent-based applications can achieve only
local optimisation due to its self-interests, as
illustrated in dash-bold lines in Figure 5.
Figure 5: Limitations in SEG.
It needs to bring together cross-cutting issues related
to different technical, economic and social
disciplines, including advance control strategies and
information and communication techniques. More
specifically, by enabling proactive behaviour,
distributed intelligence in SEG can mimic human
intelligence by number of intelligent agents to cope
with complexity, uncertainty, and variety of
circumstances.
The first focus of proactive behaviour’s
utilisation is about a local support tool to react in
real-time or on event occurring in time. Priori
knowledge as a core of the tool will be obtained by
using set of advanced computational intelligent
techniques, such as recurrent ANN or other machine
learning techniques. Via suitable Kalman Filter-
based models, dynamic system states can be
estimated and predicted adequately in steps ahead.
In addition, proactive behaviour can enhance system
optimisation to solve better conflicting interests of
involved actors in an emerging multidisciplinary
environment. As a classic but rich mathematic
applied technique, the game theory and especially its
cooperative branch could reveal innovative solutions
for such complex behaviour of SEG. Novel
algorithm supported by priori knowledge will yield
unique optimal strategy for entities to allocate their
local resources. Research’s innovation is illustrated
in Figure 6 that is extended from two local optimal
points depicted in Figure 5.
Figure 6: Innovative solutions of the research.
4 CONCLUSIONS
This paper addresses on-going research activities
related to technical aspects of Smart Energy Grids. It
aims to show a need of having coordinating
framework in which integrated functions can be
harmonised. Priori knowledge about system and
perdition of system states will be a key to enable the
framework, as so-called proactive behaviour. This
feature could help SEG to achieve global optimal
performance while system reliability and security
are ensured.
Besides technical challenges, SEG needs to
concern also on societal challenges as consumer
comfort, mondial economic changes, (absence of)
human resources, legal matters and, financial
possibilities and barriers. Societal issues might turn
into showstoppers. Co-operation is necessary to
Dual-objective
optimisation
Acting before
event occurring time
SMARTGREENS2013-2ndInternationalConferenceonSmartGridsandGreenITSystems
86

stand the world-wide economic changes. This is an
concern for all involved stakeholders with massive
knowledge and strength in the field of ICT and
power systems.
REFERENCES
Debs, A. S., & Larson, R. E. (1970). A Dynamic Estimator
for Tracking the State of a Power System. IEEE
Transactions on Power Apparatus and Systems, PAS-
89(7), 1670–1678.
Dimeas, A. L., & Hatziargyriou, N. D. (2005). Operation
of a Multiagent System for Microgrid Control. IEEE
Transactions on Power Systems, 20(3), 1447–1455.
ECN. (n.d.). PowerMatcher: Smartgrid technology.
Retrieved from http://www.powermatcher.net/
Enquiry Committee. (2012). Report of the enquiry
committee on grid disturbance in Nothern region on
30 th July 2012 and in Nothern, Eastern & North-
Eastern region on 31 st July 2012 - NEW DELHI (p.
81). Retrieved from http://www.powermin.nic.in/
pdf/GRID_ENQ_REP_16_8_12.pdf
Greunsven, J. A. W., Veldman, E., Nguyen, P. H.,
Slootweg, J., & Kamphuis, I. G. (2012). Capacity
management within a multi-agent market-based active
distribution network. Proceedings of the 3rd IEEE
PES Innovative Smart Grid Technologies Europe
(ISGT Europe). Berlin.
Hommelberg, M. P. F., Warmer, C. J., Kamphuis, I. G.,
Kok, J. ., & Schaeffer, G. J. (2007). Distributed
Control Concepts using Multi-Agent technology and
Automatic Markets: An indispensable feature of smart
power grids. IEEE Power Engineering Society
General Meeting.
Kok, K. (2011). Dynamic pricing as control mechanism.
Power and Energy Society General Meeting, 2011 (pp.
1–8). Ieee. doi:10.1109/PES.2011.6039604
Kundur, P. (1994). Power System Stability and Control (p.
1176). McGraw-Hill Professional.
Li, J., Poulton, G., & James, G. (2010). Coordination of
Distributed Energy Resource Agents. Applied
Artificial Intelligence, 24(5), 351–380.
Lopes, J. A. P., Hatziargyriou, N., Mutale, J., Djapic, P., &
Jenkins, N. (2007). Integrating distributed generation
into electric power systems: A review of drivers,
challenges and opportunities. Electric Power Systems
Research, 77(9), 1189–1203. doi:DOI: 10.1016/
j.epsr.2006.08.016
Nguyen, P. H., Kling, W. L., & Myrzik, J. M. A. (2009).
Power flow management in active networks.
PowerTech, 2009 IEEE Bucharest (pp. 1–6).
Retrieved from 10.1109/PTC.2009.5282094
Nguyen, P. H., Myrzik, J. M. A., & Kling, W. L. (2008).
Coordination of voltage regulation in Active
Networks. IEEE Transmission and Distribution
Conference and Exposition (pp. 1–6). Retrieved from
10.1109/TDC.2008.4517041
Ortega-Vazquez, M. A., & Kirschen, D. S. (2010).
Assessing the Impact of Wind Power Generation on
Operating Costs. Smart Grid, IEEE Transactions on.
doi:10.1109/TSG.2010.2081386
Peng, T. M., Hubele, N. F., & Karady, G. G. (1992).
Advancement in the application of neural networks for
short-term load forecasting. Power Systems, IEEE
Transactions on. doi:10.1109/59.141711
Saad, W., Han, Z., & Poor, H. V. (2011). Coalitional
Game Theory for Cooperative Micro-Grid Distribution
Networks. Communications Workshops (ICC), 2011
IEEE International Conference on (pp. 1–5).
Schweppe, F. C., & Wildes, J. (1970). Power System
Static-State Estimation, Part I: Exact Model. IEEE
Transactions on Power Apparatus and Systems, PAS-
89(1), 120–125.
Simoes, M. G., Roche, R., Kyriakides, E.,
Suryanarayanan, S., Blunier, B., McBee, K. D.,
Nguyen, P. H., et al. (2012). A Comparison of Smart
Grid Technologies and Progresses in Europe and the
U.S. Industry Applications, IEEE Transactions on.
doi:10.1109/TIA.2012.2199730
UCTE. (2004). Final report of the investigation committee
on the 28 September 2003 blackout Italy (p. 128).
Retrieved from http://www.rae.gr/old/cases/C13/
italy/UCTE_rept.pdf
VSYNC project. (n.d.). Retrieved from
http://www.vsync.eu/
Venayagamoorthy, G. K. (2011). Dynamic, Stochastic,
Computational, and Scalable Technologies for Smart
Grids. Computational Intelligence Magazine, IEEE.
doi:10.1109/MCI.2011.941588
Venayagamoorthy, G., Rohrig, K., & Erlich, I. (2012).
One Step Ahead: Short-Term Wind Power Forecasting
and Intelligent Predictive Control Based on Data
Analytics. IEEE Power and Energy Magazine, 10(5),
70–78.
Wooldridge, M., & Jennings, N. R. (1995). Intelligent
Agents: Theory and Practice. Knowledge Engineering
Review, 10(2), 115–152. doi:10.1.1.55.2702
Wu, F. F., Moslehi, K., & Bose, A. (2005). Power System
Control Centers: Past, Present, and Future.
Proceedings of the IEEE, 93(11), 1890–1908.
Retrieved from 10.1109/JPROC.2005.857499
Yeung, C. S. K., Poon, A. S. Y., & Wu, F. F. (1999).
Game theoretical multi-agent modelling of coalition
formation for multilateral trades. Power Systems, IEEE
Transactions on, 14(3), 929–934.
RoleofProactiveBehaviourEnabledbyAdvancedComputationalIntelligenceandICTinSmartEnergyGrids
87