Research On Underground Positioning Algorithm Based On WiFi Location Fingerprint

Our country is abundant in coal and deficient in oil and gas.Therefore,coal,which is the main energy resources,is the basic raw material to support China's economic development and industrial transformation for a long time.How to ensure the safety of coal mine production and improve production efficiency is a problem that the country and coal companies attach great importance to.In recent years,with the development of wireless communication technology,the underground positioning system based on wireless communication technology has made great contributions to production safety and disaster relief.However,low positioning accuracy,low positioning efficiency and poor anti-interference ability are the main problems existing in the current underground positioning system.Therefore,further research on the positioning algorithm to improve the accuracy,real-time and stability of the positioning system has important theoretical significance and application value in improving the production efficiency of coal enterprises and ensuring the life safety of miners.First of all,this thesis analyzes and compares the commonly used positioning technology and positioning algorithm underground,and determines the positioning algorithm based on WiFi location fingerprint as the main research object.Then,based on the introduction of the underground positioning environment,this thesis analyzes the key factors affecting the WiFi Positioning accuracy from both theoretical and experimental aspects.The layout scheme of access point(AP)in the underground tunnel is proposed.In the off-line stage,Gaussian filter and mean filter are used to process the sampled data,and a more reliable off-line fingerprint database is constructed.Secondly,In the online positioning stage,a dynamic fingerprint library based on online fingerprints is constructed,which can filter out APs with large fluctuations and APs that are not related to this positioning,reduce the algorithm traversal time,and improve positioning accuracy.Thirdly,the sensitivity of the Fuzzy C-means clustering algorithm(FCM)to singular values was used to analyze the positioning results of k-nearest-neighbor algorithm(KNN),and the FCM-KNN algorithm for underground personnel positioning was designed.It effectively solves the problem that the staffs travel route is uncertain,which easily causes the positioning error to increase,and improves the positioning accuracy.Finally,the correlation between online fingerprint and neighboring fingerprint is judged by using the theory of Spearman rank correlation,and the correlation weight coefficient is constructed.A weighted KNN algorithm based on Spearman rank correlation(SW-KNN)for underground vehicles is designed,and the positioning results are filtered by the improved Sage-Husa adaptive Kalman filter,which improves the positioning performance of the algorithm.The simulation results show that the dynamic fingerprint database can greatly improve the positioning accuracy.The underground personnel positioning algorithm based on FCM-KNN can not only ensure the positioning efficiency,but also further improve the positioning accuracy and its positioning credibility within 2m reaches 96%.The vehicle positioning algorithm based on SW-KNN underground can meet the positioning efficiency requirements of high-speed driving vehicles,and provide reliable positioning accuracy,which can meet the needs of underground vehicle positioning,navigation and scheduling.