Mining of Mineral Deposits,
Journal Year:
2023,
Volume and Issue:
17(2), P. 9 - 19
Published: June 30, 2023
Purpose.
Creating
a
generalized
algorithm
to
account
for
factors
(coal
seam
thickness,
enclosed
rock
mechanical
properties,
the
dimension
and
bearing
capacity
of
artificial
support
patterns)
causing
gateroad
state
under
effect
longwall
face
goaf.
Methods.
The
assessment
stability
is
based
on
numerical
simulation
stress-strain
(SSS).
finite
element
method
used
find
out
changes
in
SSS
surrounding
rocks
at
various
stages
mining.
elastic-plastic
constitutive
model
Hoek-Brown
failure
criterion
implemented
codes
RS2
RS3
(Rocscience)
are
applied
determine
displacements
dependently
coal
strength,
width
strength
(a
packwall
comprised
hardening
mixture
“BI-lining”).
To
specify
properties
material
series
experimental
tests
were
conducted.
A
computational
experiment
dealing
with
81
combinations
affecting
was
carried
estimate
roof
slag
floor
heaving
behind
face.
group
data
handling
(GMDH
)
employed
generalize
relationships
between
factors.
Findings.
roof-to-floor
closure
has
been
determined
intersection
goaf
packwall,
material.
It
revealed
that
gains
value
30
MPa
3rd
day
from
its
beginning
use
which
fully
corresponding
requirements
protective
capacity.
possibility
using
untreated
mine
water
liquefy
proved,
allows
simplifying
optimizing
solute
mixing
pumping
technology.
Originality.
This
study
contributes
improving
understanding
influence
underground
mining
operations
highlights
importance
utilizing
simulations
designs.
impact
each
factor
resulting
variable
(decrease
cross-section
gate
road
by
height)
combinatorial
structural
identification
estimated
as
follows:
48%,
thickness
25%,
enclosing
23%,
4%.
Practical
implications.
findings
provide
stakeholders
technique
reasonable
parameters
systems,
predictive
developed
can
be
mitigate
potential
instability
issues
excavations.
results
have
implications
similar
geological
settings
valuable
design
optimization
other
regions.
Geoscience Frontiers,
Journal Year:
2021,
Volume and Issue:
12(5), P. 101177 - 101177
Published: Feb. 23, 2021
This
paper
introduces
an
intelligent
framework
for
predicting
the
advancing
speed
during
earth
pressure
balance
(EPB)
shield
tunnelling.
Five
artificial
intelligence
(AI)
models
based
on
machine
and
deep
learning
techniques—back-propagation
neural
network
(BPNN),
extreme
(ELM),
support
vector
(SVM),
long-short
term
memory
(LSTM),
gated
recurrent
unit
(GRU)—are
used.
geological
nine
operational
parameters
that
influence
are
considered.
A
field
case
of
tunnelling
in
Shenzhen
City,
China
is
analyzed
using
developed
models.
total
1000
datasets
adopted
to
establish
The
prediction
performance
five
ranked
as
GRU
>
LSTM
SVM
ELM
BPNN.
Moreover,
Pearson
correlation
coefficient
(PCC)
sensitivity
analysis.
results
reveal
main
thrust
(MT),
penetration
(P),
foam
volume
(FV),
grouting
(GV)
have
strong
correlations
with
(AS).
An
empirical
formula
constructed
high-correlation
influential
factors
their
corresponding
datasets.
Finally,
performances
method
compared.
all
perform
better
than
method.
Journal of Rock Mechanics and Geotechnical Engineering,
Journal Year:
2021,
Volume and Issue:
13(6), P. 1340 - 1357
Published: Oct. 22, 2021
Tunnel
boring
machine
(TBM)
vibration
induced
by
cutting
complex
ground
contains
essential
information
that
can
help
engineers
evaluate
the
interaction
between
a
cutterhead
and
itself.
In
this
study,
deep
recurrent
neural
networks
(RNNs)
convolutional
(CNNs)
were
used
for
vibration-based
working
face
identification.
First,
field
monitoring
was
conducted
to
obtain
TBM
data
when
tunneling
in
changing
geological
conditions,
including
mixed-face,
homogeneous,
transmission
ground.
Next,
RNNs
CNNs
utilized
develop
prediction
models,
which
then
validated
using
testing
dataset.
The
accuracy
of
long
short-term
memory
(LSTM)
bidirectional
LSTM
(Bi-LSTM)
models
approximately
70%
with
raw
data;
however,
instantaneous
frequency
transmission,
increased
80%.
Two
types
CNNs,
GoogLeNet
ResNet,
trained
tested
time-frequency
scalar
diagrams
from
continuous
wavelet
transformation.
CNN
an
greater
than
96%,
performed
significantly
better
RNN
models.
ResNet-18,
98.28%,
best.
When
sample
length
set
as
rotation
period,
achieved
highest
while
proposed
model
simultaneously
high
feedback
efficiency.
could
promptly
identify
conditions
at
without
stopping
normal
process,
parameters
be
adjusted
optimized
timely
manner
based
on
predicted
results.