Baseline analysis of predicted tracking system performance

Communications and tracking systems are required to be installed and maintained in all underground coal mines in the United States. Systems offer a wide range of technologies to provide the miner's location. The means to evaluate and predict the performance of the tracking system regardless of the tracking technology in use is presented. Performance measures employed in this work are descriptive of where the tracking system works, how accurate it is and the variation with time. An example of a typical Central Appalachian Coal mine is presented that has installed a wireless mesh communication and tracking system. This example system provides continuous tracking, but the same techniques can be applied to discontinuous tracking systems, such as those based on Radio Frequency Identification. The analysis flow begins with the mine map and tracking system design provided by the manufacturer. Future mine projections will work with this technique as well. The mine geometry is extracted from the mine map and simulation locations emplaced. Infrastructure locations and configurations, being input parameters from which the lowest attenuation signal paths from each antenna to each simulation point are found. These values are used in a statistical calculator to make pseudo-random values for the range and variation of loss along each path. The output from these predictions is used to calculate prediction measures for the tracking system. Performance measures were extensively tested using an apparatus developed in this project. This experiment results in predicting the performance of a tracking system that is in line with the observed performance. More importantly it demonstrates that the predicted measures and observed values using standardized performance metrics do adequately describe the tracking system because the general values and trends are correct.