Water Resources Research,
Год журнала:
2020,
Номер
56(3)
Опубликована: Янв. 29, 2020
Abstract
Addressing
challenges
in
urban
water
infrastructure
systems,
including
aging
infrastructure,
supply
uncertainty,
extreme
events,
and
security
threats,
depends
highly
on
distribution
networks
modeling
emphasizing
the
importance
of
realistic
assumptions,
complexities,
scalable
solutions.
In
this
study,
we
propose
a
derivative‐free,
linear
approximation
for
solving
network
flow
problem.
The
proposed
approach
takes
advantage
special
form
nonlinear
head
loss
equations,
and,
after
transformation
variables
constraints,
problem
reduces
to
optimization
that
can
be
efficiently
solved
by
modern
solvers.
Ultimately,
amounts
series
problems.
We
demonstrate
through
several
case
studies
show
model
arbitrary
topologies
various
types
valves
pumps,
thus
providing
flexibility.
Under
mild
conditions,
converges.
provide
sensitivity
analysis
discuss
detail
current
limitations
our
suggest
solutions
overcome
these.
All
codes,
tested
networks,
results
are
freely
available
Github
research
reproducibility.
Archives of Computational Methods in Engineering,
Год журнала:
2023,
Номер
30(7), С. 4209 - 4244
Опубликована: Май 31, 2023
Abstract
Water
distribution
networks
are
crucial
for
supplying
consumers
with
quality
and
adequate
water.
A
water
system
comprises
connected
hydraulic
components
which
ensure
supply
to
meet
demand.
Optimization
of
is
carried
out
minimize
resource
utilization
expenditure
or
maximize
the
system’s
efficiency
higher
benefits.
Genetic
algorithms
signify
an
effective
search
technique
non-linear
optimization
problems
have
gained
acceptance
among
resources
planners
managers.
This
paper
reviews
various
developments
in
systems
using
genetic
algorithms.
These
pertinent
creating
novel
distributing
expansion,
reinforcement,
rehabilitation
process
prevailing
mechanisms.
Graphical
Nature Water,
Год журнала:
2024,
Номер
2(4), С. 370 - 379
Опубликована: Март 26, 2024
Abstract
Globally,
1.6
billion
people
in
rural
regions
face
water
scarcity.
Expanding
freshwater
access
via
brackish
groundwater
desalination
can
provide
additional
resources
to
address
this
challenge.
In
study,
we
have
developed
a
time-variant
electrodialysis
reversal
(EDR)
technology
that
flexibly
uses
available
solar
energy
for
desalination.
Our
proposed
photovoltaic-powered
system
vary
pumping
and
EDR
power
match
the
availability
of
intermittent
power,
maximizing
rate.
results
show
improved
performance
with
direct
use
77%
energy—91%
more
than
conventional
systems—and
92%
reduction
battery
reliance.
village-scale
case
study
India,
these
improvements
lead
22%
cost,
making
competitive
currently
used
on-grid,
reverse
osmosis
systems
are
mainly
powered
by
fossil
fuels.
Future
advances
could
further
reduce
costs,
providing
an
improved,
sustainable
solution
scarcity
remote
areas.
IEEE Transactions on Smart Grid,
Год журнала:
2021,
Номер
12(6), С. 4799 - 4812
Опубликована: Авг. 16, 2021
A
major
outage
in
the
electricity
distribution
system
may
affect
operation
of
water
and
natural
gas
supply
systems,
leading
to
an
interruption
multiple
services
critical
customers.
Therefore,
enhancing
resilience
infrastructures
requires
joint
efforts
sectors.
In
this
paper,
a
service
restoration
method
considering
electricity-water-gas
interdependency
is
proposed.
The
objective
provide
electricity,
water,
supplies
customers
desired
ratio
according
their
needs
after
extreme
event.
operational
constraints
networks
are
considered.
characteristics
electricity-driven
coupling
components,
including
pumps
compressors,
also
modeled.
Relaxation
techniques
applied
nonconvex
posed
by
physical
laws
those
networks.
Consequently,
problem
formulated
as
mixed-integer
second-order
cone
program,
which
can
readily
be
solved
off-the-shelf
solvers.
proposed
validated
numerical
simulations
on
integrated
developed
based
benchmark
models
subsystems.
results
indicate
that
refines
allocation
limited
generation
resources
demonstrate
exactness
convex
relaxation.
IEEE Transactions on Control Systems Technology,
Год журнала:
2024,
Номер
32(3), С. 945 - 959
Опубликована: Янв. 1, 2024
A
major
challenge
in
the
transition
to
a
net-zero
energy
system
is
how
decarbonize
use
for
heating
and
cooling
via
electrification
while
ensuring
security
of
power
with
high
penetration
renewable
generation.
One
possible
way
simultaneously
address
these
disparate
objectives
model
predictive
control
(MPC)
manage
multivector
consumption
storage
individual
buildings
so
that
operational
constraints
connecting
networks
are
not
violated.
However,
large
number
building
assets
their
multiple
shared
using
standard
MPC
requires
solution
optimization
problems
large,
nonconvex,
and,
therefore,
intractable.
In
this
article,
novel
scheme
proposed
which
overall
problem
decomposed
solved
parallel
by
decentralized
agents.
Since
uncoordinated
actions
agents
could
cause
congestion
electricity
district
(DHC)
networks,
an
flow
coordinator
also
introduced.
This
checks
agent
solving
optimal
each
network
uses
price
signals
direct
search
globally
feasible
solution.
To
improve
computational
efficiency
when
determining
flows,
utilizes
reformulation
DHC
network.
An
exemplary
case
study
multienergy
demonstrates
ensures
near-optimal
economic
performance
compared
equivalent
centralized
benchmark—in
case,
reducing
maximum
computation
time
from
over
55
min
just
1
s.
The
approach
suitable
online
management
within
district,
both
minimize
costs
end
users
maintain
secure,
reliable
operation
networks.
IEEE Transactions on Control of Network Systems,
Год журнала:
2020,
Номер
7(3), С. 1283 - 1295
Опубликована: Фев. 11, 2020
The
vast
infrastructure
development,
gas
flow
(GF)
dynamics,
and
complex
interdependence
of
with
electric
power
networks
call
for
advanced
computational
tools.
Solving
the
equations
relating
injections
to
pressures
pipeline
flows
lies
at
heart
natural
network
(NGN)
operation,
yet
existing
solvers
that
require
careful
initialization
uniqueness
has
been
an
open
question.
In
this
context,
article
considers
nonlinear
steady-state
version
GF
problem.
It
first
establishes
solution
problem
is
unique
under
arbitrary
NGN
topologies,
compressor
types,
sets
specifications.
For
setups
where
pressure
specified
on
a
single
(reference)
node
compressors
do
not
appear
in
cycles,
task
posed
as
n
convex
minimization.
To
handle
more
general
setups,
solver
relying
mixed-integer
quadratically
constrained
quadratic
program
(MI-QCQP)
also
devised.
This
can
be
used
any
setup
NGN.
introduces
binary
variables
capture
directions,
relaxes
drop
inequality
constraints,
uses
carefully
selected
objective
promote
exactness
relaxation.
relaxation
probably
exact
NGNs
nonoverlapping
cycles
fixed-pressure
node.
handles
efficiently
involved
bilinear
terms
through
McCormick
linearization.
Numerical
tests
validate
our
claims,
demonstrate
MI-QCQP
scales
well,
even
when
sufficient
conditions
are
violated,
such
overlapping
multiple
nodes.
IEEE Transactions on Control of Network Systems,
Год журнала:
2020,
Номер
7(3), С. 1151 - 1163
Опубликована: Янв. 7, 2020
Optimal,
network-driven
control
of
water
distribution
networks
(WDNs)
is
very
difficult:
valve
and
pump
models
form
nontrivial,
combinatorial
logic;
hydraulic
are
nonconvex;
demand
patterns
uncertain;
WDNs
naturally
large
scale.
Prior
research
on
WDN
addressed
major
challenges,
yet
either
i)
adopted
simplified
models,
topologies,
rudimentary
valve/pump
modeling
or
ii)
used
mixed-integer,
nonconvex
optimization
to
solve
problems.
The
objective
this
article
develop
tractable
computational
algorithms
manage
operation,
while
considering
arbitrary
topology,
flow
direction,
an
abundance
types,
objectives,
operational
constraints-all
only
using
convex,
continuous
optimization.
Specifically,
we
propose
new
geometric
programming
(GP)-based
model
predictive
(MPC)
algorithms,
designed
the
equations
obtain
controls,
i.e.,
pump/valve
schedules
alongside
heads
flows.
proposed
approach
amounts
solving
a
series
convex
problems
that
graciously
scale
networks.
tested
126-node
network
with
many
valves
pumps
shown
outperform
traditional,
rule-based
control.
developed
GP-based
MPC
as
well
numerical
test
results,
all
included
Github.
IEEE Transactions on Signal Processing,
Год журнала:
2023,
Номер
71, С. 2027 - 2042
Опубликована: Янв. 1, 2023
Online
topology
estimation
of
graph-connected
time
series
is
challenging
in
practice,
particularly
because
the
dependencies
between
many
real-world
scenarios
are
nonlinear.
To
address
this
challenge,
we
introduce
a
novel
kernel-based
algorithm
for
online
graph
estimation.
Our
proposed
also
performs
Fourier-based
random
feature
approximation
to
tackle
curse
dimensionality
associated
with
kernel
representations.
Exploiting
fact
that
networks
often
exhibit
sparse
topologies,
propose
group-Lasso
based
optimization
framework,
which
solved
using
an
iterative
composite
objective
mirror
descent
method,
yielding
fixed
computational
complexity
per
iteration.
We
provide
theoretical
guarantees
our
and
prove
it
can
achieve
sublinear
dynamic
regret
under
certain
reasonable
assumptions.
In
experiments
conducted
on
both
real
synthetic
data,
method
outperforms
existing
state-of-the-art
competitors.
The
location
of
tanks
impacts
the
optimal
design
and
reliability
water
distribution
networks.
However,
contention
exists
in
literature
regarding
best
for
tanks.
aim
this
study
was
therefore
to
develop
a
tool
compute
failure
tolerance
when
pipe
occurs
network,
as
consequence,
determine
tank(s).
To
achieve
this,
five
designs
Anytown
Network
(ATN),
which
is
benchmark
network
literature,
were
selected.
These
designs,
recommended
additional
at
different
locations
hydraulically
simulated
using
pressure
driven
analysis
EPANET
2.2,
these
results
validated.
tolerance,
Microsoft
Excel
®
developed,
validated
applied
ATN
designs.
comparison
values
generated
revealed
influence
tank
on
during
i.e.,
while
each
>
0.68
(a
less
vulnerable
network),
an
tank(s)
downstream
demand
center.
Incidentally,
emerged
cheapest
points
fact
that
higher
need
not
be
more
expensive
network.
