Sensors,
Journal Year:
2023,
Volume and Issue:
23(11), P. 4997 - 4997
Published: May 23, 2023
This
paper
presents
a
systematic
approach
for
solving
complex
prediction
problems
with
focus
on
energy
efficiency.
The
involves
using
neural
networks,
specifically
recurrent
and
sequential
as
the
main
tool
prediction.
In
order
to
test
methodology,
case
study
was
conducted
in
telecommunications
industry
address
problem
of
efficiency
data
centers.
involved
comparing
four
including
networks
(RNNs),
long
short-term
memory
(LSTM),
gated
units
(GRUs),
online
extreme
learning
machine
(OS-ELM),
determine
best
network
terms
accuracy
computational
time.
results
show
that
OS-ELM
outperformed
other
both
simulation
applied
real
traffic
showed
potential
savings
up
12.2%
single
day.
highlights
importance
methodology
be
industries.
can
further
developed
technology
continue
advance,
making
it
promising
solution
wide
range
problems.
World Journal of Advanced Research and Reviews,
Journal Year:
2024,
Volume and Issue:
21(2), P. 1782 - 1790
Published: Feb. 28, 2024
Optimizing
energy
efficiency
in
data
center
cooling
towers
is
crucial
for
reducing
operational
costs
and
environmental
impact.
This
review
explores
the
integration
of
predictive
maintenance
project
management
to
achieve
this
goal.
By
leveraging
techniques,
operators
can
anticipate
address
potential
issues
before
they
lead
costly
downtime
or
inefficiencies.
Project
plays
a
key
role
coordinating
these
efforts,
ensuring
that
activities
are
carried
out
efficiently
effectively.
Predictive
relies
on
analytics
machine
learning
algorithms
monitor
condition
real-time.
analyzing
such
as
temperature,
pressure,
flow
rates,
detect
anomalies
predict
failures.
proactive
approach
allows
schedule
during
planned
downtime,
minimizing
disruptions
operations.
practices,
Agile
Waterfall
methodologies,
essential
efforts.
managers
oversee
planning,
execution,
monitoring
activities,
completed
time
within
budget.
They
also
facilitate
communication
between
stakeholders,
including
teams,
analysts,
management,
ensure
everyone
aligned
working
towards
same
goals.
integrating
with
significant
improvements
efficiency.
addressing
proactively,
reduce
consumption,
extend
lifespan
equipment,
minimize
downtime.
Additionally,
practices
efforts
coordinated
effective,
maximizing
benefits
maintenance.
In
conclusion,
optimizing
requires
holistic
combines
management.
techniques
efficiency,
costs,
enhance
sustainability.
The Journal of Engineering and Exact Sciences,
Journal Year:
2023,
Volume and Issue:
9(6), P. 16397 - 01e
Published: Aug. 17, 2023
The
rapid
proliferation
of
digital
services
and
the
surge
in
cloud
computing
have
significantly
increased
demand
for
data
centers
worldwide.
As
these
facilities
consume
vast
amounts
energy,
there
is
growing
concern
about
their
impact
on
environment,
particularly
relation
to
climate
change.
This
paper
reviews
extent
which
contribute
global
greenhouse
gas
emissions
examines
energy
efficient
strategies
being
employed
mitigate
environmental
footprint.
Key
findings
indicate
that
without
intervention,
centre
could
rival
those
major
industries.
Fortunately,
innovations
cooling
technologies,
architectural
design,
renewable
sourcing,
hardware
efficiency
shown
potential
reducing
consumption.
also
underscores
importance
a
holistic
approach
encompasses
both
technological
advancements
policy
measures
meaningful
progress.
In
conclusion,
while
are
indeed
source
concern,
ongoing
energy-efficient
provide
promising
pathway
towards
sustainable
future.
Sensors,
Journal Year:
2025,
Volume and Issue:
25(5), P. 1428 - 1428
Published: Feb. 26, 2025
Dynamic
workflow
scheduling
in
cloud
environments
is
a
challenging
task
due
to
dependencies,
fluctuating
workloads,
resource
variability,
and
the
need
balance
makespan
energy
consumption.
This
study
presents
novel
framework
that
integrates
Graph
Neural
Networks
(GNNs)
with
Deep
Reinforcement
Learning
(DRL)
using
Proximal
Policy
Optimization
(PPO)
algorithm
achieve
multi-objective
optimization,
focusing
on
minimizing
reducing
By
leveraging
GNNs
model
dependencies
within
workflows,
enables
adaptive
informed
allocation.
The
agent
was
evaluated
CloudSim-based
simulation
environment
synthetic
datasets.
Experimental
results
across
benchmark
datasets
demonstrate
proposed
framework’s
effectiveness,
achieving
consistent
improvements
consumption
over
traditional
heuristic
methods.
achieved
minimum
of
689.22
s
against
second
best
800.72
moderate-sized
datasets,
significantly
up
13.92%
baseline
methods
such
as
HEFT,
Min–Min,
Max–Min,
while
maintaining
competitive
10,964.45
J.
These
findings
highlight
potential
combining
DRL
for
dynamic
environments,
effectively
balancing
multiple
objectives.
Engineering Science & Technology Journal,
Journal Year:
2024,
Volume and Issue:
5(1), P. 99 - 114
Published: Jan. 19, 2024
The
digital
age
has
led
to
a
surge
in
connectivity,
innovation,
and
information
exchange,
but
it
also
escalating
energy
consumption
by
data
centers.
Green
centers
have
emerged
as
transformative
solution,
embodying
commitment
sustainability
through
eco-friendly
practices
cutting-edge
technologies.
Key
principles
of
green
include
energy-efficient
hardware,
renewable
integration,
advanced
cooling
systems,
resource
optimization
strategies.
Energy-efficient
hardware
involves
replacing
outdated
servers,
storage
network
equipment
with
alternatives,
such
virtualization
This
reduces
power
sets
the
stage
for
more
sustainable
technologically
center
infrastructure.
Renewable
integration
dependence
on
traditional
grids
fossil
fuels,
ensuring
an
eco-friendlier
supply.
Advanced
liquid
immersion,
hot
aisle
containment,
free
air
cooling,
optimize
efficiency
while
maintaining
ideal
server
temperatures.
Resource
ensures
that
every
unit
is
utilized
judiciously,
contributing
overarching
goal
sustainability.
transition
presents
challenges
upfront
investment
costs,
fluctuating
grids,
technical
complexities
associated
systems.
However,
there
are
substantial
opportunities,
including
reduced
operational
improved
brand
image,
compliance
environmental
regulations.
Emerging
trends
artificial
intelligence
edge
computing,
which
enable
prediction
peak
workloads,
dynamic
management.
By
prioritizing
efficiency,
embracing
innovative
technologies,
staying
attuned
emerging
trends,
can
play
pivotal
role
forging
future.
Keywords:
Data
Centers,
Sustainability,
Energy
Efficiency,
It
Infrastructure,
Edge
Computing,
Artificial
Intelligence.
Energies,
Journal Year:
2023,
Volume and Issue:
16(3), P. 1314 - 1314
Published: Jan. 26, 2023
This
work
presents
a
design
for
uninterruptible
power
supply
inverters
using
Pareto
front
optimization
improved
cost
and
efficiency.
Three
PWM
modulation
techniques
applied
to
the
full-bridge
inverter
are
analyzed.
As
result,
best
MOSFET
solution
in
terms
of
efficiency
is
evaluated
based
on
database
with
47
MOSFETs.
Using
front,
optimal
sub-optimal
solutions
compared,
considering
three
characteristics
MOSFETs
manufactured
different
voltage
levels.
Thermal
electrical
measurements
used
validate
models.
Applied Sciences,
Journal Year:
2024,
Volume and Issue:
14(14), P. 6046 - 6046
Published: July 11, 2024
This
study
delves
into
the
adoption
of
organic
Rankine
cycle
(ORC)
for
recovering
waste
heat
from
data
centers
(DCs).
Through
a
literature
review,
it
examines
energy
reuse
with
focus
on
electric
power
generation,
selection
working
fluids,
and
system
design
principles.
The
objective
is
to
develop
thorough
framework
analysis,
beginning
quantity
quality
investigation
available.
Air
cooling
systems,
chosen
often
their
simplicity,
account
about
70%
used
methods.
Water
demonstrates
greater
effectiveness,
albeit
less
commonly
adopted.
pays
close
attention
potential
meticulously
considering
limitations
presented
by
available
sources
cold
vaporization
condensation,
respectively.
It
reviews
an
ORC-based
setup,
incorporating
fluid
streams
internal
processes.
research
includes
conceptual
case
where
designed
simulations
are
conducted
in
DWSIM
environment.
simulation
model
considers
hot
air
or
liquid
water
returning
center
ORC
evaporation.
Ambient
serves
condensing,
pentane
isopentane
identified
as
suitable
fluids.
Pentane
assures
net
efficiencies
ranging
between
3.1
7.1%
when
operating
pressure
ratios
increase
2.8
6.4.
Isopentane
meanwhile,
achieve
3.6–7.0%
across
2.7–6.0.
Furthermore,
provides
key
performance
indicators
reference
terms
usage
effectiveness
(PUE),
factor
(ERF),
(ERE),
greenhouse
gas
(GHG)
savings.
concludes
guidelines
including
exergy
considerations,
details
sizing
process
evaporators
condensers.
