Research Of Positioning Algorithm In Coal Mine Based On Location Fingerprint

Safety is the priority of the coal mine production. However, the special environment of the coal mine brings tremendous security challenges to the production. In order to improve the level of coal mine monitoring, personnel location is used to real time locate the position of the mines. When disaster occurred in the coal mine, the personnel localization can provide the location information of the miners, which increases the rescue efficiency. This paper studied the localization algorithm based on position fingerprint and provided the improved algorithm.Firstly, this paper analyzed some kinds of localization algorithms and the feasibility of their application in the coal mine. The localization algorithm based on position fingerprint is more suitable for the application in the coal mine. Location fingerprint positioning algorithm is based on RSSI, Therefore, this paper studied the wireless signal propagation characteristics in the coal mine from the theoretical and practical, then simulation the influence of multipath effects on positioning. Then analyzed the key factors of Fingerprint Location,which became to the foundation for research of the following algorithm.Secondly, combining with the long and narrow characteristics of tunnel of the coal mine, the idea of regional division was introduced in this paper. Regional threshold was determined by the data that were collected in the coal mine. The localization algorithm divided the region into more sub-regions according to the regional threshold. Then, the sub-region of terminal device can be got by the received signal strength.The localization algorithm based on position fingerprint was used to accurate locate the terminal device in the sub-area.At last, comparing the K neighboring and bayes matching algorithm, this paper introduced weighted K-Bayes localization matching algorithm. This algorithm calculated k reference position points by using weighted Euclidean distance and ruled out the points whose error is bigger.The localization results was got by weighted processing the rest points probabilities that were got by the bayes theorem. Experimental results show the feasibility of the algorithm.