Tikrit Journal of Engineering Sciences,
Год журнала:
2023,
Номер
30(4), С. 74 - 87
Опубликована: Ноя. 25, 2023
The
evolving
character
of
the
environment
makes
it
challenging
to
predict
water
levels
in
advance.
Despite
being
most
common
approach
for
defining
hydrologic
processes
and
implementing
physical
system
changes,
physics-based
model
has
some
practical
limitations.
Multiple
studies
have
shown
that
machine
learning,
a
data-driven
forecast
hydrological
processes,
brings
about
more
reliable
data
is
efficient
than
traditional
models.
In
this
study,
seven
learning
algorithms
were
developed
dam
level
daily
based
on
historical
level.
input
combinations
investigated
improve
model’s
sensitivity,
statistical
indicators
used
assess
reliability
model.
study
multiple
models
with
scenarios
suggested
bagged
trees
trained
days
lagged
provided
highest
accuracy.
tree
achieved
an
RMSE
0.13953,
taking
less
10
seconds
train.
Its
efficiency
accuracy
made
stand
out
from
rest
With
deployment
field,
predictions
can
be
help
mitigate
issues
relating
supply.
Results in Engineering,
Год журнала:
2023,
Номер
21, С. 101712 - 101712
Опубликована: Дек. 30, 2023
The
current
worldwide
effects
of
soil
erosion
are
a
result
natural
and
human
activities,
it
had
serious
consequences
on
ecosystems,
agriculture,
water
quality
in
the
watershed.
As
result,
quantifying
physical
characteristics
watershed
can
be
used
to
identify
areas
that
more
susceptible
require
immediate
mitigation
measures.
This
study
prioritizes
fourteen
sub-watersheds
Dabus
Watershed,
Ethiopia
using
Geographical
Information
System
based
variety
parameters
for
implementing
short
long-term
effective
management
practices.
For
each
sub-watershed
(SW),
compound
parameter
was
calculated
from
different
morphometric
rank
areas.
findings
show
SW1,
SW7,
SW10
contributing
very
high
area
with
its
2390.75
km2,
2555.77
1642.71
km2
covering
area;
however,
SW2,
SW8,
SW9,
SW14
lowly
degraded
due
is
low.
Thus,
measures
such
as
contouring,
terracing,
filter
strips,
other
structural/non-structural
approaches
should
implemented
where
contributed.
Agricultural Water Management,
Год журнала:
2024,
Номер
293, С. 108698 - 108698
Опубликована: Фев. 2, 2024
Optimizing
reservoir
operation
is
a
complex
problem
with
non-linearities,
numerous
decision
variables,
and
challenging
constraints
to
simulate
solve.
Researchers
have
tested
various
metaheuristics
algorithms
(MHAs)
reduce
water
deficit
in
reservoirs
presented
them
decision-makers
for
adoption.
Optimization
methods
vary
depending
on
objectives,
type,
used.
The
paper
utilizes
the
CSS
algorithm
study
impact
of
scenarios
optimal
Mujib
Jordan
deficits
using
historical
date
between
2004
2019.
explores
different
scenarios,
including
sediment
impact,
demand
management,
increasing
storage
volume
reservoir,
identify
reservoir.
compares
results
these
current
Risk
analysis
(volumetric
reliability,
shortage
index
(SI),
resilience,
vulnerability)
error
indexes
(correlation
coefficient
R2,
root
mean
square
(RMSE),
absolute
(MAE))
were
used
compare
addition
annual
values
from
each
scenario.
simulation
monthly
showed
that
accumulation
accounts
14.6%
reservoir's
at
end
Removing
sediments
retained
by
dam
can
19.42%
when
algorithm.
Additionally,
reducing
agricultural
11%
removing
reduced
42.40%.
also
examined
capacity
10%,
20%,
30%,
revealing
decrease
35.44%
removal
was
included
analysis.
scenario
11%,
sediment.
This
resulted
53.59%
deficit,
providing
viable
solutions
address
Hydrology,
Год журнала:
2024,
Номер
11(5), С. 63 - 63
Опубликована: Апрель 27, 2024
Morphological
changes
in
canals
are
greatly
influenced
by
sediment
load
dynamics,
whose
estimation
is
a
challenging
task
because
of
the
non-linear
behavior
concentration
variables.
This
study
aims
to
compare
different
techniques
including
Artificial
Intelligence
Models
(AIM)
and
empirical
equations
for
estimating
Upper
Chenab
Canal
based
on
10
years
data
from
2012
2022.
The
methodology
involves
utilization
newly
developed
equation,
Ackers
White
formula
AIM
20
neural
networks
with
training
functions
both
Double
Triple
Layers,
two
Neuro-Fuzzy
Inference
System
(ANFIS),
Particle
Swarm
Optimization,
Ensemble
Learning
Random
Forest
models.
Sensitivity
analysis
variables
has
also
been
performed
using
various
scenarios
input
combinations
AIM.
A
state-of-the-art
optimization
technique
used
identify
parameters
its
performance
tested
against
equation.
To
models,
four
types
errors—correlation
coefficient
(R),
T-Test,
Analysis
Variance
(ANOVA),
Taylor’s
Diagram—have
used.
results
show
successful
application
(AI)
capture
indicate
that,
among
all
ANFIS
outperformed
simulating
total
high
R-value
0.958.
models
was
assessed,
notable
accuracy
achieved
AIM11
AIM21.
Moreover,
equation
better
(R
=
0.92)
compared
0.88).
In
conclusion,
provides
valuable
insights
into
dynamics
canals,
highlighting
effectiveness
AI
techniques.
It
suggested
incorporate
other
use
multiple
modeling
future.
World Electric Vehicle Journal,
Год журнала:
2024,
Номер
15(7), С. 308 - 308
Опубликована: Июль 14, 2024
Machine
learning
techniques
have
advanced
rapidly,
leading
to
better
prediction
accuracy
within
a
short
computational
time.
Such
advancement
encourages
various
novel
applications,
including
in
the
field
of
operations
research.
This
study
introduces
way
utilize
regression
machine
models
predict
objectives
vehicle
routing
problems
that
are
solved
using
genetic
algorithm.
Previous
studies
generally
discussed
how
(1)
research
methods
used
independently
generate
optimized
solutions
and
(2)
values
from
given
dataset.
Some
collaborations
between
fields
as
follows:
input
data
for
problems,
optimize
hyper-parameters
models,
(3)
improve
quality
algorithms.
differs
types
collaborative
listed
above.
focuses
on
objective
problem
directly
output
data,
without
optimizing
straightforward
framework
captures
characteristics
problem.
The
proposed
is
applied
by
generating
algorithm
then
obtained
values.
numerical
experiments
show
best
random
forest
regression,
generalized
linear
model
with
Poisson
distribution,
ridge
cross-validation.
Results in Engineering,
Год журнала:
2024,
Номер
23, С. 102420 - 102420
Опубликована: Июнь 15, 2024
Seepage
is
a
critical
problem
in
earthfill
dams
which
threatens
the
dam's
stability
and
safety
owing
to
extreme
shifts
climate
change
with
rise
water
intake
dams.
To
cope
this
challenge
dam
monitoring
essential
for
structural
rehabilitation
enhancement
earth
integration
of
deep
learning
approach.
This
research
presents
novel
approach
evaluating
seepage
by
using
Recurrent
Neural
Network
its
associated
co-variant
predict
at
multiple
location
fill
Tarbela
dam.
Short-term
peak
seasonal
hydraulic
hydro
climatological
data
was
used
from
Pakistan's
Earth
Rockfill
over
period
2014
2020.
The
results
demonstrate
that,
compared
other
models,
proposed
model
efficiently
predicts
extent
study
highlights
importance
considering
historical
correlations
analysis,
providing
significant
insights
stakeholders
regarding
most
effective
utilization
resources
purposes,
provide
idea
digitization
system
Pakistan
integrated
DL
algorithms.
Indian Journal of Anaesthesia,
Год журнала:
2024,
Номер
68(12), С. 1081 - 1091
Опубликована: Дек. 1, 2024
Remifentanil
is
a
powerful
synthetic
opioid
drug
with
short
initiation
and
period
of
action,
making
it
an
ultra-short-acting
opioid.
It
delivered
as
intravenous
infusion
during
surgical
procedures
for
pain
management.
However,
deciding
on
suitable
dosage
depends
various
aspects
specific
to
each
individual.
Conventional
pharmacokinetic
pharmacodynamic
(PK-PD)
models
mainly
rely
manually
choosing
the
parameters.
Target-controlled
delivery
systems
need
precise
predictions
drug's
analgesic
effects.
This
work
investigates
supervised
machine
learning
(ML)
methods
analyse
characteristics
remifentanil,
imitating
measured
data.
From
Kaggle
database,
features
such
age,
gender,
rate,
body
surface
area,
lean
mass
are
extracted
determine
concentration
at
instant
time.
The
show
that
prediction
algorithms
perform
better
over
traditional
PK-PD
greater
accuracy
minimum
mean
squared
error
(MSE).
By
optimising
hyperparameters
Bayesian
methods,
performance
these
significantly
improved,
attaining
MSE
value.
Applying
ML
in
can
reduce
resource
costs
time
effort
essential
laboratory
experiments
pharmaceutical
industry.
Tikrit Journal of Engineering Sciences,
Год журнала:
2023,
Номер
30(4), С. 74 - 87
Опубликована: Ноя. 25, 2023
The
evolving
character
of
the
environment
makes
it
challenging
to
predict
water
levels
in
advance.
Despite
being
most
common
approach
for
defining
hydrologic
processes
and
implementing
physical
system
changes,
physics-based
model
has
some
practical
limitations.
Multiple
studies
have
shown
that
machine
learning,
a
data-driven
forecast
hydrological
processes,
brings
about
more
reliable
data
is
efficient
than
traditional
models.
In
this
study,
seven
learning
algorithms
were
developed
dam
level
daily
based
on
historical
level.
input
combinations
investigated
improve
model’s
sensitivity,
statistical
indicators
used
assess
reliability
model.
study
multiple
models
with
scenarios
suggested
bagged
trees
trained
days
lagged
provided
highest
accuracy.
tree
achieved
an
RMSE
0.13953,
taking
less
10
seconds
train.
Its
efficiency
accuracy
made
stand
out
from
rest
With
deployment
field,
predictions
can
be
help
mitigate
issues
relating
supply.
