Coal Layer Thickness Estimation Using GPR Signal Processing Techniques

Abstract

One of the most important objectives in the present research technology of coal mining industries is the modernization of underground coal mining machine for personnel security. Conventionally, a thin remnant layer of coal is unmined while mining, in order to maintain the geographical stability of cutting face. This residual coal layer thickness primarily depends upon the existing geological conditions of the site and its range should be within a specified limit. Therefore, present research approach in the field of coal mining is to enrich the existing robust sensing methods for detecting the interface between coal and rock by enhancing resolution and sensing higher depth of penetration while estimating the thickness of coal layer.
This dissertation addresses the above mentioned challenges by adopting a step frequency continuous wave ground penetrating radar (SFCW GPR) as a sensor for coal-rock interface detection (CID). In order to process the SFCW GPR data, high resolution time delay estimation (TDE) algorithms are introduced, which not only efficiently estimates the thickness of thin remnant coal layer but, also capable in sensing weak echo signals from the target of higher depth. To improve the time resolution of GPR system, various high resolution algorithms are adopted for coal layer thickness estimation by TDE. The results are presented in terms of resolution power as well as relative root mean square error (RRMSE) in estimating the coal layer thickness.
In practice, the antenna of SFCW GPR yields frequency dependent gain, delay and other losses due to internal reflection. In order to incorporate all those effect in GPR signal modelling, three schemes, full wave model (FWM), plane wave model (PWM) and modified plane wave model (MPWM) are considered while generating synthetic data for experiment. The performance of the proposed signal processing methods are analysed using synthetic data generated by above modelling schemes using 1-3 GHz SFCW GPR system. The experimental results have revealed that the proposed methods outperform the conventional signal processing based methods while estimating the minimum and maximum amount of detectable thickness of coal layer, 3 cm and 441 cm respectively.