Sie sind auf Seite 1von 52

RESEARCH PAPER

GROUP 4
RESHMINDER KAUR A132843
NOR IZHAM NOHANZI A133921
MOHD IMRAN MOHD JUNAIDI A133239
MUHF NURHILMI SAHARIN A134058
AMIEROUL IEFWAT AKASHAH A133697

Spatial Prediction of Ground Subsidence Susceptibility
Using an Artificial Neural Network
TERMINOLOGIES
Spatial Prediction of Ground Subsidence Susceptibility
Using an Artificial Neural Network
SPATIAL PREDICTION OF GROUND SUBSIDENCE SUSCEPTIBILITY
USING AN ARTIFICIAL NEURAL NETWORK
Subsidence
is the motion of a surface (usually, the Earth's surface) as it shifts
downward relative to a datum such as sea-level.
is the sinking or settling of the ground surface
Ground subsidence
is the settlement of native low density soils, or the caving in of
natural or man-made underground voids.
Spatial data
Data that define a location. These are in the form of graphic
primitives that are usually either points, lines, polygons or pixels.
Vector data
A representation of the world using points, lines, and polygons.
Vector models are useful for storing data that has discrete
boundaries, such as country borders, land parcels, and streets.
Susceptibility
The state or fact of being likely or liable to be influenced or harmed
by a particular thing
SPATIAL PREDICTION OF GROUND SUBSIDENCE SUSCEPTIBILITY
USING AN ARTIFICIAL NEURAL NETWORK
Artificial Neural Network (ANN)
are computational models inspired by an animal's central nervous systems (in
particular the brain) which is capable of machine learning as well as pattern
recognition.
Geographic Information System (GIS)
is a computer system designed to capture, store, manipulate, analyze, manage,
and present all types of geographical
Digital Elevation Model (DEM)
The representation of continuous elevation values over a topographic surface by a
regular array of z-values, referenced to a common datum. DEMs are typically
used to represent terrain relief.






Rock Mass Rating (RMR)
Is a geomechanical classification system for rocks developed by Z.T. Bieniawski
between 1972-1973


INTRODUCTION
Spatial Prediction of Ground Subsidence Susceptibility
Using an Artificial Neural Network
SPATIAL PREDICTION OF GROUND SUBSIDENCE SUSCEPTIBILITY
USING AN ARTIFICIAL NEURAL NETWORK
In South Korea, the coal industry played an important role in 1960s
1970s.

Began to decline in 1980s along with the decrease in international oil
prices.

In 2005, only seven of 345 coal mines operating nationwide.
There is NO MEASURES taken to protect against environmental
damage after the mine closed.

Various heavy metal flow from mine waste heaps in leachate,
causing serious pollution in rivers and soil.

Underground subsidence can cause damage to surface structure as
well as human injury.

Ground subsidence is treated by simple reinforcements after the
subsidence has occurred.
SPATIAL PREDICTION OF GROUND SUBSIDENCE SUSCEPTIBILITY
USING AN ARTIFICIAL NEURAL NETWORK
The objective of this study is:

To assessed and predicted discontinuous
residual subsidence to produce a ground
subsidence susceptibility (GSS) map of an
area near abandoned underground coal
mines using an artificial neural network
(ANN) in a geographic information system
(GIS) environment.
SPATIAL PREDICTION OF GROUND SUBSIDENCE SUSCEPTIBILITY
USING AN ARTIFICIAL NEURAL NETWORK
Why use GIS?
Provide a way to introduce information and knowledge
from other data sources into the decision making
process and they aid in the handling and manipulation
of classified remote-sensing data.
Enables quantitative assessment in environmental
systems over a broad range of spatial and temporal
scales.

Recent Studies
GSS have been studied and examined based on
Geological and geotechnical investigation
Probability, statistical, and fuzzy algebra
ANN model tandem with GIS applications
SPATIAL PREDICTION OF GROUND SUBSIDENCE SUSCEPTIBILITY
USING AN ARTIFICIAL NEURAL NETWORK
The present study assessed and predicted GSS using raster
databases, in an ArcGIS grid format, of topographic, geologic,
and geotechnical data and the locations of subsidence areas
already discovered in the study area.

ArcGIS 9.3 software (ESRI, CA) was used for database
construction, coordinate conversion, grid production, overlay
analysis, and spatial analysis.

Using the major factors, GSS maps were drawn by application
of ANN models and then validated by area-under-the-curve
analysis and a field survey.

By this approach, the major influences on ground subsidence
were determined in a limited 1-km
2
region, and a method for
predicting GSS efficiently was established.

Jeong-am (3712037130N, 1285310
1285410E; see Fig. 2), was an important coal
mining area
lies between Mt. Baek-Wu to the southwest and
Mt. Ham-Beak to the southeast
still has many cavities remaining from mining.
Thus, areas of likely ground subsidence exist in
Jeong-am.
all coal in South Korea is anthracite, 85% - upper
Paleozoic era & lower Mesozoic era in the
Jangseong Formation of the Pyeongan
Supergroup
STUDY AREA
Sadong Group
Bamchi formations
Jangseong formations
Gobangsan Group
Hambaeksan formations
Tosagok formations
Kohan formations
PYEONGAN
SUPERGROUP

Jangseo
ng
Formatio
n
includes several coal beds, -
workable quality and thickness.
Coal mining - occurred
1967 until 1989
Coal seams - have steep slopes
(>6070).
Average seam thickness was
~1.32.5 m
Jangseong
Formation
composed
mainly of
alternating
sandstone and
shale
deposits, with
the shale
having
intercalations
of two to three
coal bed
seams
Hambaegsan
Formation
an upper
stratum of the
major coal
drifts, is
composed of
coarsely
grained meta-
sandstone and
gray shale and
is relatively
resistant to
weathering

However,
the
developmen
t of
cleavages
from severe
folding has
led to
weaknesses
in the rock
masses of
this
formation
A local road (no. 38) shows
features of typical sinkhole
collapse and deformations and
cracks in the road surface
Ground subsidence can occur in
areas of past underground mining
activity.
In underground mines, ground
subsidence develops from the
mine roof to the ground surface.
Mine collapse with time is
attributable to decreased shear
strength, groundwater injection,
and increased seepage force after
coal mining

depth and height
of mine cavities
excavation method
degree of
inclination of the
excavation
scope of mining,
structural geology
flow of
groundwater
the mechanical
characteristics
represented by the
rock-mass rating
(RMR)
In this study,
locations of ground
subsidence and
factors governing the
occurrence of ground
subsidence were
collected in a vector-
type spatial database
and then represented
on a grid using the
ArcGIS software
package.
The spatial database
is listed in Table 1
INPUT FACTOR
> national organizations, such as the Coal Industry Promotion
Board (for ground subsidence),
> the National Geographic Information Institute (for topography
and land use),
> the Mine Reclamation Corporation (for mine tunnels and
boreholes), and
> the Korea Institute of Geoscience and Mineral Resources (for
geology)



+





OK



Areas of buildings, mountains, railways, fields, rivers,
complex area, roads, and multi-purpose area use were
extracted from the land-use map
The slope angle was obtained from the digital elevation map.
a triangulated irregular network was made using the
elevations.
Contours (5-m intervals) and survey base points of elevation
were from the topographic map,
METHODOLOGY
Spatial Prediction of Ground Subsidence Susceptibility
Using an Artificial Neural Network
An ANN is a computational mechanism able to acquire,
represent, and compute a mapping from one multivariate
space of information to another, given a set of data
representing that mapping

Purpose of an ANN to build a model of the data-
generated weighting process so that the network can
generalize and predict output from inputs that is has not
previously seen

Hidden- and output-layer nodes process their input by
multiplying them by a corresponding weight, summing the
products, and processing the sum using a nonlinear
transfer function
An ANN learns by adjusting the weights between the
nodes in response to errors between the actual and
target output values

At the end of phase, neural network provides a model
that should be able to predict a target value from a given
input value

Two stages are involved in using neural network for
multisource classification
1) Training stage
2) Classifying stage
PROCEDURE
Training sites selected based on scientific and objective
criteria, location considered likely and unlikely to have
GSS were selected as training sites

Areas where ground subsidence has not occurred were
classified as areas not prone to ground subsidence and
50% of areas where ground subsidence was known to
have occurred were assigned to the areas prone to
ground subsidence.
The back-propagation algorithm was then applied to
calculate the weights between the input and hidden
layers and between the hidden and output layers

An 8x16x1 structure was selected for the network, and
input data was normalized in the range of 0.1-0.9

Learning rate was set as 0.01, initial weights were
randomly set as values between 0.1-0.3

The weights calculated from ten test cases were
compared with determine whether the variation in the
final weight depended on the selection of the initial
weight
Weight between layers acquired by neural network
training were calculated in reverse, and the contribution
or importance of each factor were determined

Weight that represent the contribution or importance of
each factor were determined

MATLAB software used for weight calculation and
interpretation of the weight

The model was trained for 5000 epochs, and the root
mean-square error(RMSE) value used for the stopping
criterion was set at 0.01.

If this RMSE value was not achieved, then the maximum
number of iterations was terminated at 5000 epochs.

An epochs means the entire training set to the neural network

The maximum RMSE value when the latter case occurred was
0.214. The final weight between layers acquired during training
of the neural network, and the contribution or importance of
each of the eight factors, were used to predict GSS

Finally, the weights were applied to the entire study area, and
GSS maps were created for each training case
RESULTS
Spatial Prediction of Ground Subsidence Susceptibility
Using an Artificial Neural Network
WEIGHT DETERMINATION AND GSS MAPPING
Final weights between layers acquired during training of the neural network and the
contribution or importance of each of the eight factors used to predict GSS.

The calculations were repeated ten times to allow the results to achieve similar values.
The SD of the results ranged from 0.014 to 0.033.
The average values were calculated, and divided by the average of the weights of the
factor having the minimum value.
Among the weights, distance from lineament had the highest value (1.5491) and
RMR had the lowest (1.000).
The GSS values were classified by equal areas and grouped into five classes (% of area) of
GSS rank for easy visual interpretation:
- very high (5%), high (5%), medium (10%), low (20%), and low (60%).
The minimum and maximum values were 0.008 and 0.952.
The mean and SD were 0.205 and 0.221
VALIDATION
Predictions made using the ANN method were compared with expected
results based on knowledge of the factors.
Rate curves - the calculated GSS values of all grids in the study area were
sorted in descending order which divided into 100 classes in accumulated
1% intervals.
AUC was calculated to compare the results. A total area of 1 denotes perfect
prediction accuracy for all cases.
The AUC method can be used to assess prediction accuracy qualitatively.
The percentages of validated results appear as a line in Fig. 7.

For case no. 1, the highest accuracy, 10% of the study area having a greater GSS
could explain 90% of all ground subsidence.
20% of the study area where the GSS value had a greater rank could explain 99% of
ground subsidence (Fig. 7).
AUC values for the ANN produced GSS maps were between 0.9484 and 0.9598

DISCUSSION
Spatial Prediction of Ground Subsidence Susceptibility
Using an Artificial Neural Network
HIGHLIGHTS
1. The need to have an overall and systematic analysis method
for understanding the effects of each factor and interactions
among factors

2. Relative environmental factors played important roles in
producing the final map products

3. Primary value of results proved to be a robust and usefull
tool for estimating and mapping subsidence even with some
incomplete data

1. THE NEED TO HAVE AN OVERALL AND
SYSTEMATIC ANALYSIS METHOD
May take a very long time from the
beginning of ground subsidence in
underground cavities till visible damages
occur at the surface
When an underground mine is
abonded,ground subsidence develops from
the mine activity roof to the ground surface
Various factors can generate ground
subsidence and there is complex relation
among those factors
Underground
Facilities
Construction
Abandoned
Coal Mines
Soft Soil In
landfill Area
Corrosion of
Limestone
Ground
Subsidence


In this Study,

1) GIS techniques used to study the prediction and management
of ground subsidence in abandoned mines
2) An ANN model applied to assess and predict GSH in Jeong-
am,South Korea,a region where ground subsidence is
expected to continue in the future
3) Influences of factors that are expected to affect were
quantitavely analysed
4) Maps of GSS were made using ANN and repeated 10x
5) Training Sites extracted from ground-subsidence areas
6) Validation Showed 94.84 and 95.98 % prediction accuracy
(Ave:95.41%)-Similar & Satisfactory

2. RELATIVE ENVIRONMENTAL FACTORS PLAYED IMPORTANT
ROLES IN PRODUCING THE FINAL MAP PRODUCTS

RMR had little influence on final model results considering
paucity of data and narrow value variation range for those few
data that were extrapolated

DOES NOT MEAN that RMR is Unimportant Factor in
defining subsidence hazard but available data is insuffiecient to
support any other conclusion

Spatial Distribution of Lineament Data extended across the
entire area,these data are expected to have had greater
influence on model results






3. PRIMARY VALUE OF RESULTS PROVED TO BE A ROBUST AND
USEFULL TOOL FOR ESTIMATING AND MAPPING SUBSIDENCE EVEN
WITH SOME INCOMPLETE DATA


In this study,
Using the ANN,the relative importance and
weights of factor were calculated
Factor Average Weight
Slope 0.106 1.055
Depth of Drift 0.023 1.303
Distance from Drift 0.027 1.184
Depth of Groundwater 0.021 1.180
RMR 0.014 1.000
Distance from lineament 0.033 1.549
Geology 0.016 1.376
Land Use 0.019 1.360
Weights of Factor
Slope
10%
Depth of
Drift
13%
Distance
from
Drift
12%
Depth of
Groundw
ater
12%
RMR
10%
Distance
from
lineament
15%
Geology
14%
Land Use
14%
Weights of
Hydrogeological factors
in GSH Analysis
The determined weights
indicates geological factors such
as geology and lineament are
important for ground subsidence
compared to others

Slope was not important

RMR data have limited
availability and low accuracy
,showed low weight
CONCLUSION
Spatial Prediction of Ground Subsidence Susceptibility
Using an Artificial Neural Network
Study have shown factors involved in ground subsidence and
the method and findings can be applied to GSS mapping in
other regions

GSS Map produced can be used to mitigate hazards to people
and facilities

Locating monitoring and facility sites for establishing plans
to prevent ground hazards

Das könnte Ihnen auch gefallen