Study Of Three-dimensional Localization System And Method Based On WSN

Node localization is the key technology and application foundation of wireless sensor networks, of which the location technology has gradually transformed from two-dimensional space to three-dimensional space. In practical application, the two-dimensional plane positioning system cannot meet the requirements and has to be extended to the three-dimensional space. One of the key research directions is indoor positioning in three-dimensional space. There are some problems in the process of deploying indoor three-dimensional positioning system, such as the positioning problem of beacon nodes. The algorithm to solve the problem also has strong practicability. Three-dimensional space has its unique features, the algorithm of which is more complex than two-dimensional plane positioning but has a huge development space, making it worth researching. Therefore, this thesis mainly studies the characteristics of the relative position of the nodes in Three-dimensional positioning system, and puts forward some improved algorithm based on its characteristics.This thesis first expounds the basic concept, structure and characteristics of wireless sensor network (WSN) and analyzes the wireless sensor network positioning technology. And then the localization algorithm for 3D environment is introduced. A brief introduction on principle of Cricket 3D positioning system and the hardware and software structure are given. According to the characteristics of the Cricket node, the arrangement of beacon nodes is studied and an three-dimensional positioning system system is built up.After analyzing the application environment and characteristics of linear self-localization algorithm based on beacon node location estimate, an improved linear self-localization algorithm based on weighted least squares is proposed, which has inherited the characteristic of not relying on the position of the node coordinates. All it needs is an increase in the times that measured to obtain the characteristics of the relative position between the beacon node coordinates. According to the influence of different geometric distribution on propagation of error, the estimation precision is further improved and simulation experiments are taken on the proposed algorithm to analyze its performance.Then the optimization problem of beacon node position coordinates in three-dimensional positioning system is studied based on wireless sensor network. The mobile robot and positioning system are modeled, and then a system state prediction equation is presented to improve the precision of robot localization, making it possible to optimize the calibration of the position of the beacon nodes. Then the performance of EKF, HF and STF filtering algorithms are compared. HF has a better ability to adapt for uncertainty of system model; STF has a better tracking performance of the real state under the condition of great disturbance. Then the simulation data is validated.Finally, this thesis analyzes the characteristics of three dimensional localization algorithm based on Cricket positioning system and puts forward the three dimensional positioning problem for three beacon node. Two solution algorithms are as below: Gauss-Newton iterative method and localization algorithm based on Cayley-Menger determinant. Characteristics:Gauss-Newton’s iteration method has high precision, but low efficiency; positioning algorithm based on Cayley-Menger determinant has low accuracy, but high efficiency. This paper introduces GDOP in three-dimensional positioning, and further puts forward an easy-GDOP which is more suitable for our system, and then analyzes the application in beacon nodes selection. At last, the accuracy and characteristics of the weighted data fusion are researched, coming to the conclusion that the weighted data fusion with easy-GDOP value has a higher computational efficiency.