Study on Mechanical Data Logging for Small Diameter Drilling using Neural Network Techniques
| Project/Area Number |
06651089
|
|---|
| Research Category |
Grant-in-Aid for General Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
資源開発工学
|
|---|
| Research Institution | Muroran Institute of Technology |
|---|
Principal Investigator |
ITAKURA Ken-ichi Muroran Institute of Technology, Department of Computer Science and Systems Engineering, Associate Professor, 工学部, 助教授 (20168298)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
SATO Kazuhiko Muroran Institute of Technology, Department of Computer Science and Systems Engi, 工学部, 教授 (30002009)
|
|---|
| Project Period (FY) |
1994 – 1995
|
| Project Status |
Completed (Fiscal Year 1995)
|
| Budget Amount *help |
¥2,000,000 (Direct Cost: ¥2,000,000)
Fiscal Year 1995: ¥400,000 (Direct Cost: ¥400,000)
Fiscal Year 1994: ¥1,600,000 (Direct Cost: ¥1,600,000)
|
|---|
| Keywords | Neural Network / Adaptive Resonance Theory / ART2 / Drilling / Logging / Rockbolt / MWD / Mechanical Data |
|---|
| Research Abstract |
In underground mining it is important to obtain information about the rock structure, including distribution of discontinuities, before excavation and constructing supports. Mechanical data from a drilling machine at the working face such as torque, thrust, revolution and stroke can be useful for understanding the underground structure. Therefore, both laboratory and field experiments were carried out in order to develop a mechanical data logging system. A data analysis system using neural network techniques was also developed for objective interpretation of the logging data and to display the log of strata and discontinuities.
1.From the laboratory experiments it was found that the level of torque or thrust of a drilling machine reflected the difference in rock types, and the separation of strata and cracks appeared as a specific pattern in the log of torque/thrust ratio or torque/drilling speed per one revolution.
2.As an algorithm of neural network techniques, adaptive resonance theory (ART2) and back propagation network (BPN) were employed to extract the pattern corresponding to a rock layr boundary and cracks from mechanical data of a drilling machine. Using both algorithms sequentially the location of discontinuities were successfully estimated.
3.For the field experiments, a data acquisition system was attached to a portable drilling machine for the collection of mechanical data. A feasibility study was conducted using the drilling machine at an underground coal mine. It was confirmed that the discontinuity log estimated by the analysis conformed to the core sample reconstructed from borehole TV data.
4.Combining several sets of the logging data measured at a local area in the field study, a two-dimensional structure of the strata including discontinuities was reconstructed. From this structure it was possible to estimate the orientation of discontinuities, which was difficult to evaluate from the logging data obtained from only one drill hole.
|
Report
(3 results)
Research Products
(16 results)
