Measurement of feeder performance during coal discharge from an underroof seam using machine vision

Abstract

The technology for extracting and discharging coal from an underroof seam uses the so-called gravitational extraction method in which coal is extracted and discharged from under the roof by gravity. Here, coal can be discharged onto the main conveyor (face conveyor, located in the supported area), central conveyor (rear conveyor in Western literature), and tail conveyor (discharge conveyor, located in the unsupported area). The most common facilities used currently are longwall sets of equipment providing discharge onto tail conveyors. The purpose of this study is to measure the performance of a motorised plate feeder supplying coal from the outlet port of a roof support to a conveyor during the extraction of thick seams with discharge onto the face conveyor. To achieve the goal, it is proposed to measure the coal volume using machine vision. Methods for calculating a unit volume in a measuring section using a three-dimensional model were investigated. Laboratory studies were carried out to estimate the relative errors of the methods. The research allowed properly defining: a method for collecting data to calculate the unit volume of coal; a method for calculating the unit volume in the measuring section; a method for calculating the feeder performance using machine vision, and approaches for physically simplifying the video scene examined by machine vision. A relative error of less than 10 % with the existing measurement accuracy for constructing a coal layer surface height map indicates the sufficiency of the proposed calculation method for engineering use. The developed mathematical apparatus for calculating the unit volume of coal at the measuring section and measuring the feeder performance allows creating algorithmic software using the elementary mathematical functions of addition, subtraction, multiplication, and division. This aspect is important because it lower sights for the software development environment, and therefore expands the range of hardware suitable for calculating the feeder performance.

Keywords

• mining
• coal mining
• coalface
• performance
• coal discharge
• face conveyor
• rock mass volume
• machine vision
• pattern recognition
• video image recognition
• height map