Wifi-based Personnel Positioning System Of Coal Preparation Plant

China is a large coal producing country.Because of the particularity of coal mining environment,it is easy to have safety accidents in the process of coal mine operation.In order to ensure the safety of coal enterprises,ensure the personal safety of workers,reduce the occurrence of serious safety accidents,and standardize the operation process of workers,it is urgent to use the modern personnel positioning system for supervision.In the case of coal mine safety accidents,timely statistics of the number and location of the victims through reliable means has become the key to reduce the difficulty of rescue and improve the efficiency of rescue.In this paper,the key technologies of the personnel location system based on Wi-Fi are studied,and the main factors that affect the location performance by adopting the location fingerprint matching algorithm are analyzed.Fingerprint matching algorithm for position on both offline and positioning of from lower matching algorithm and the amount of calculation work,improve the quality of the RSS signal,the perspective of the real time and improve the stability of the system,improved the AP selection strategy and position of traditional fingerprint algorithm,to some extent,improved the performance of the Wi-Fi positioning technology based on the position fingerprint.First of all,on the offline phase fingerprint database creation,mainly to do the work of two aspects: first,according to the AP selection of fingerprint data redundancy problem,based on the influence factors of Wi-Fi signal,this paper analyses the commonly used AP selection strategy,and combined with the actual environment of coal preparation plant,was proposed based on the maximum average K-means clustering AP selection algorithm.Compared with the classical AP selection strategy,the algorithm proposed in this paper considers the effect of the whole AP set on the positioning accuracy,and avoids the problem of the system workload and large amount of computation caused by the redundancy of fingerprint data.Second,in view of the attenuation of RSS signal transmission in a complex environment,the Gaussian filter model and the Kalman filter model are selected for data preprocessing by analyzing and studying the commonly used filter models and aiming at the characteristics of the two key links of the location fingerprint algorithm.Secondly,from the system positioning accuracy,computation and the perspective of the workload,reduce system was proposed based on region partition-K-means clustering of the weighted K nearest neighbor algorithm,twice by the decrease of the spatial dimension,effectively reduce the need to match to the settlement of the locating position fingerprint number,reduce the workload and calculation of the system,improve the real-time performance of the algorithm,to improve the overall performance of the system have bigger.Finally,the personnel positioning system of coal preparation plant based on Android is designed.Starting from the actual needs,the functional modules and communication modes of the client and the server are designed.In addition,the system test program has been developed,and the system has been tried out in the coal preparation plant.The trial results show that the system runs stably and reliably and meets the actual demand.