The Development of Subsidence Hazard Index for an Abandoned Mine Using Artificial Neural Network and Interaction Matrix

Yangkyun Kim; Min Son; Hyun-Koo Moon; Seung-Ah Lee; In-Jae Yang

doi:10.12972/ksmer.2017.54.5.480

All Issue

2017 Vol.54, Issue 5 Preview Page Next Page

Research Paper

The Development of Subsidence Hazard Index for an Abandoned Mine Using Artificial Neural Network and Interaction Matrix 인공신경망과 상관도 행렬을 이용한 폐광지역 지반침하 위험도지수 개발

31 October 2017. pp. 480-490

PDF XML

Abstract

A MSH(Mine Subsidence Hazard) index has been developed to estimate the subsidence possibility of an abandoned mine by means of an artificial neural network (ANN) and an interaction matrix. For this, 19 influence factors of mine subsidence were determined through a literature study, and the influence factors for 287 subsidence points at 38 abandoned mines were selected first, of which 108 points (34 abandoned mines) having high data quality were used for learning of ANN. Each influence factor of which range is widely distributed was classified into seven ranges to use it as a learning data of ANN. Interaction matrixes were constructed by an ANN analysis, and a MSH index for each subsidence point was calculated from influence factors as well as influence weights that are from the interaction matrix. The result shows that MSH index ranges between 37 and 66. The average of MSH index for Non-coal mines is 54, which is 13% higher than the one for coal mines of which average MSH index is 47.

Keywords

Abandoned mine

Subsidence hazard index

Artificial neural network

Interaction matrix

인공신경망과 상관도행렬을 활용하여 폐광지역 지반침하 가능성을 예측할 수 있는 침하위험도 지수(MSH index)를 개발하였다. 인공신경망 분석을 위해 먼저, 문헌연구를 통해 폐광지역 침하 영향인자 19가지를 결정하고, 한국광해관리공단에서 발간한 지반안정성 보고서를 분석하여 국내 38개 폐광지역에서 총 287개 개별침하지에 대한 영향인자값들을 1차 선별한 후 데이터의 완성도가 높은 개별침하지 108개(34개 폐광지역)를 최종 선정하였다. 선정된 데이터는 통계분석을 통해 일정한 구간별로 점수화한 후 인공신경망 학습에 활용하였다. 학습된 인공신경망 모델로 각 영향인자들의 상호간 민감도를 산출하여 상관도 행렬을 구성하고 이를 통해 학습에 사용된 개별 침하지의 침하위험도 지수를 산출하였다. 그 결과 침하위험도 지수는 전체적으로 37점부터 66점까지 분포하였으며, 평균점수는 51점으로 나타났다. 비석탄광의 MSH index 평균값은 54점으로 47점인 석탄광에 비해 평균 13% 높게 나타났다.

키워드

폐광

침하위험도 지수

인공신경망

상관도행렬

References

Cancelli, A. and Crosta, G., 1993. Hazard and risk assessment in rockfall prone areas: Chapter 15-Risk and Reliability in Ground Engineering. Proc. of the conference organised by the Institute of Civil Engineers, ICE, London, p.177-190.

Choi, S.O., Jeon, Y.S., Park, E.S., Jeong, Y.B., and Chun D.S., 2005. Analysis of subsidence mechanism and development of evaluation program. Tunn Undergr Space, 15(3), 195-212.

Hudson, J., 1992. Rock Engineering System : Theory and practice. Ellis Horwood Limited, UK, Chicester, p.185.

Jiao, Y. and Hudson, A., 1995. The fully-coupled model for rock engineering systems. 1nt. J. rock Mech. Min. Sci. & Geomech., Abstr. 32(5), 491-512.

10.1016/0148-9062(95)00038-I

Jung, Y.B., Song, W.K., and Kang, S.S., 2008. Development of subsidence hazard estimation method based on the depth of gangway. Tunnel & Underground Space(J. Korean Soc. Rock Mech.), 18(4), 272-279.

Kim, Y.K., Min, S., Moon, H.K., and Lee, S.A., 2017. A study on the development of an artificial neural network model for the prediction of ground subsidence over abandoned mines in Korea. Geosystem Engineering, 20(3), 163-171.

10.1080/12269328.2016.1254573

Kwon, H.H. and Nam, K.S., 2007. Mine Reclamation Engineering. Dongwha technique, Korea, p.92-203.

MIRECO, 2016. A Study on a Systematic Standard for the Evaluation of Subsidence Risk of an Abandoned mine area. 2016-059, Wonju, Korea, p.77-100.

Prechelt, L., 1998. Automatic early stopping using cross validation: quantifying the criteria. Neural Network, 11, 761-767.

10.1016/S0893-6080(98)00010-0

Shin, H.S., 2003. User manual for a neural network engine for training and prediction(neurosys v2003). Centre for Civil and Computation Engineering University of Wales Swansea, p.24.

Shin, H.S., Lee, S.S., Kim, K.Y., and Bae, G.J., 2013. Quan-titative evaluation of collapse hazard levels of tunnel faces by interlinked consideration of face mapping, design and construction data: focused on adaptive weights. J. of Korean Tunn Undergr Sp Assoc, 15(5), 505-522.

10.9711/KTAJ.2013.15.5.505

Suh, J.W., Choi, Y.S., Park, H.D., Kwon, H.H., Yoon, S.H., and Go, W.R., 2010. Application of Frequency Ratio and Analytic Hierarchy Process to Subsidence Hazard Assessment around Abandoned Coal Mine. J Korean Soc Miner Energy Resour Eng, 47(5), 690-704.

Toll, D., 1996. Artificial Intelligence Application in Geotechnical Engineering. Electr J Geotech Eng, 1, 1-3.

Yang, Y. and Zhang, Q., 1998. The application of neural networks to rock engineering systems (RES). Int. J. Rock Mech. Min. Sci & Geomech., Abstr. 35(6), 727-745.

10.1016/S0148-9062(97)00339-2

Information

Publisher :The Korean Society of Mineral and Energy Resources Engineers
Publisher(Ko) :한국자원공학회
Journal Title :Journal of the Korean Society of Mineral and Energy Resources Engineers
Journal Title(Ko) :한국자원공학회지
Volume : 54
No :5
Pages :480-490
Received Date : 2017-09-26
Revised Date : 2017-10-23
Accepted Date : 2017-10-24
DOI :https://doi.org/10.12972/ksmer.2017.54.5.480

Journal of the Korean Society of Mineral and Energy Resources Engineers ISSN:2288-0291(Print) 2288-2790(Online) 한국자원공학회지

All Issue