Enhanced Positioning Research For Moving Target In Mine Underground

Accurate target positioning is the important guarantee for mine safety and efficient production. However, the wireless signal transmission has the phenomenon of Non Line of Sight(NLOS) and multipath fading considering the long and narrow space under coal mine, which lead to low positioning accuracy and poor adaptation for the positioning system. From the view of deployment, even if high accuracy positioning system is developed, the large scale replacement for the existing localization system will be a tremendous waste of pre-investment. The development of Mine Internet of Things provides new opportunities and ideas for accurate localization of coal mine. In order to real-time sense, monitor and warning the activities of the staff and equipment in the mine, enough sensors need to be installed which can not only sense the information nearby, but also be helpful for accurate localization of coal mine based on Internet of Things(IOT) characteristic of other sensor or the existing positioning system.Therefore, this paper developed two methods of improving original positioning accuracy of sensors and one extended Kalman filter method based on restraining the influence of NLOS error.(1) Improve the positioning accuracy based on existing sensor nodes and geographic information in coal mine. Firstly, this method acquires coordinate of the targets by the existing positioning system. Secondly, auxiliary positioning needs calling sensors which can communicate and be near to the target, and combine coordinates of system positioning and sensors positioning. At last, the outside coordinates can be corrected to the roadway by the geographic information of underground, and the precise positioning for moving targets is implemented. The simulation results show that the positioning accuracy is improved on the basis of the original positioning system.(2) An Enhanced Localization Method based on Witness Nodes(WitEnLoc) is built by the existing sensors in the roadway. The method calls information from the sensor nodes in the area of moving target communication from IOT control platform and selects some of the sensor nodes as witness. The search radius of target driven by witness node is shrinked by changing the witness searching radius. Wit EnLoc is primarily comprised of three stages: initial value calculation, target node searching and search result correction. The simulation results indicate that WitEn Loc can improve the accuracy of existing localization systems regardless of the original localization methods.(3) Influence of NLOS error is eliminated when positioning by extended Kalman filter. Taking an enhanced localization method of witness node based on Kalman filter for example, this method uses filter twice for positioning coordinates by extended Kalman filter to further improve the positioning accuracy. The simulation results indicate that this method can restrain the influence of NLOS error on the positioning accuracy.