Research On Underground Personnel Localization Technology For Wireless Sensor Networks

Localization algorithm is a basic work for personnel localization system for coal miners,If localization algorithm is the more robustness and the more reliability,it is important to provide accurate information for policymakers.According to national standards structure for underground mining. first of all, the RSSI ranging error is influenced not only by distance but also by complex underground environment. In APIT localization algorithm,ranging error is excessively accumulated,which leads to a larger positioning estimation error in per iteration.Secondly, with a larger target area to be monitored in mine localization system, particulars are searching solutions in a wide range of space, and maybe some individual particulars repeat to search solutions in the area which has been searched in, which made time delay throughout the WSN. Finally, the positions of unknown nodes in the target area are changing at any time, if the anchor nodes are installed fixedly along the roadway, it can not cover all unknown nodes effectively.Major tasks of the thesis are as follows: (1) We do a deep research about the current localization system of Shanxi's underground miners, and find some problems about localization algorithm. Secondly, we analyzed mines'structure, daily activity rhythm of miners and their environment.(2) We make a comparison of some staple covering algorithms at present to sum up their advantage and disadvantage.Give a performance evaluation index of coverage algorithm. We analyzed the coverage efficiency, in which anchor node cover mobile unknown nodes. The thesis proposed that based on virtual grids area density coverage algorithm, in which we firstly identify target area which may be hot area, rare area, or normal area; then gravity, repulsion and boundary coefficient on the virtual force are updated according to area density. Anchor node position is readjusted by virtual forces to achieve to cover the unknown nodes in the boundary or triangle area, dispelling effectively the blind spots of boundary or triangle.(3) From the two angles of monitor system network administrator and ground monitoring personnel, we analysis the demands of the WSN localization technology in underground miners'localization system. Network administrators ask to grasp WSN coverage efficiency and localization precision at any time; the ground monitoring personnel requests to keep track of the miners'activity.(4) We give a priority to the study of PSO localization algorithm according to the demands of underground personnel localization system, and proposed PAH-PSO(Probability Area Hopsize Particular Swarm Optimization)localization algorithm, which makes use of the least square method to predict the node location in order to narrowing the solution space range,and particulars can search subtly. It can improve the node localization accuracy. The algorithm avoided accumulated ranging error in APIT at per iteration. In unknown nodes'communication range, there are no three anchor nodes which can constitute of triangular area.(5) Finally, we verify the coverage efficiency of based on virtual grids area density coverage algorithm and the localization accuracy of PAH-PSO algorithm. On the Matlab7.0 simulation platform, we make a great of comparison between improved algorithm and other similar algorithm, PAH-PSO node localization errors reduced by 36% compared with the PSO node localization algorithm.