Open AccessA Coal Mine Underground Localization Algorithm Based on the Feature Vector
Author(s) -
Yinjing Guo,
Xianqi Song,
Ling Yang,
Wenhong Lv
Publication year - 2019
Publication title -
journal of engineering and technological sciences
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.202
H-Index - 14
eISSN - 2338-5502
pISSN - 2337-5779
DOI - 10.5614/j.eng.technol.sci.2019.51.2.3
Subject(s) - position (finance) , coal mining , algorithm , feature vector , process (computing) , matching (statistics) , feature (linguistics) , signal (programming language) , vector map , computer science , pattern recognition (psychology) , direction vector , support vector machine , artificial intelligence , data mining , mathematics , engineering , coal , statistics , linguistics , philosophy , waste management , programming language , finance , economics , operating system
To enhance the position estimation accuracy of an underground localization system for coal mine roadways, an algorithm based on the feature vector of received signals is presented in this paper. The algorithm includes three steps: the construction process of a feature vector database and a distance database, the vector matching process and the localization process. When a signal vector is received, it only needs to calculate the distance from the received vector to the center vector of each subset and then compare it with the data in the distance database. After multiple filtering and comparing the source of the strongest signal, the coordinates closest to the received vector are found. The experiment showed that the maximum error of this algorithm was 4 m and the average error was 1.62 m. Furthermore, within a localization error of 1 m, the X-axis localization accuracy was 98% while the Y-axis localization accuracy was 86%. Also, the algorithm took much less time compared to the KNN and WKNN algorithms, so the algorithm meets the requirements of coal mine safety systems and underground personnel localization systems.