UWB Radar Imaging Method for the Longwall Retreat and Creep Displacements Measurement in Underground Coal Mine

The longwall retreat and creep displacements are two important parameters for the safety production in underground coal-mine. At present, the problem restricting the displacements measurement of longwall retreat and creep is mainly the influence of coal dust and water vapor at the end of longwall face. Therefore, this paper introduces UWB radar imaging method with strong imaging capability all day and all weather to measure the longwall retreat and creep displacements. That is, the radar moves along the slide-way mounted on the head of armored face conveyor (AFC) to scan the bolt plates fixed in the roadway. Two images obtained before and after the longwall face movement are matched to acquire displacements. The data processing has two steps: radar imaging and image registration. First, according to the signal model of synthetic aperture radar (SAR), the modified range-doppler (RD) algorithm with the fourth-order range model is derived to obtain near-field high-resolution image. Second, on the basis of SAR-scale invariant feature transform (SARSIFT) algorithm, this paper replaces Gaussian filter (GF) with bilateral filter (BF) to remove the speck noise, and designs an principal component analysis-improved feature matching (PCAIFM) algorithm to eliminate the wrong matching points. Therefore, the proposed image registration algorithm is called BF-SARSIFT-PCAIFM. According to the correct-match pairs, the similarity transformation model is calculated to acquire the information of retreat and creep displacements. The simulation experiment in building corridor and field test in underground coal mine are carried out, and the results show that the measurement errors of these two displacements are less than 10 mm. Note to Practitioners —The sliding up and down of mining equipment along longwall face, called the longwall creep phenomenon, often occurs during the mining processing, which may cause the roof falling accident, affect the overall advancement of longwall face, and so on. In addition, if the measured retreat displacement is larger or smaller than the normal value, the safety risk may arise and the mining efficiency would be reduced. Therefore, the monitoring of longwall retreat and creep displacements is an urgent problem to be solved. This paper proposes the UWB radar imaging method to measure the retreat and creep displacements. Due to its imaging capacity all-day and all-weather, UWB radar imaging method has the potential for application in underground coal-mine. The experiment results performed in this paper verify the effectiveness and practicability of the proposed method. However, because the two-dimensional (2D) radar image has only range and azimuth resolutions and not height, the radar movement route must be in the same horizontal plane as bolt plates to be scanned, which increases the difficulty of practical application. In addition, the scanning mode of radar linear motion is unsuitable for underground limited space. Therefore, UWB radar three-dimensional (3D) imaging method under new scanning mode is the focus of our next research.