An optimization strategy for anchor point placement for underground coal mine area positioning

Abstract

Coal mine underground positioning technology plays an important role in safe mining, personnel monitoring and dispatching, and post-disaster rescue. Recently, there has been an increasing interest in indoor localization due to the demand for location-based services. Diverse techniques have been described in the literature to improve indoor localization services, but their accuracy is significantly affected by the number and location of the anchors, which act as a reference point for localizing tags in a given space. This work describes APOTSA, a novel approach for discovering a high-quality placement of anchors to improve the accuracy of area-based indoor localization systems while requiring a shorter execution time than existing approaches. The proposed algorithm is based on Tabu search and optimizes the localization accuracy by minimizing the expected residence area, which indicates a geographical area containing the tag. APOTSA's effectiveness with respect to solution quality and time of execution is shown by evaluating different scenarios involving up to five anchors. Moreover, a simple yet effective discretization approach is introduced to speed up the calculations, and the trade-off between the running time and solution quality of our algorithm can be controlled by a parameter. Our results indicate that the expected residence area and the time of execution can be reduced by up to 9.5 % and 99 %compared to the state-of-the-art local search anchors placement (LSAP) algorithm, respectively.