Research On Three-dimensional Positioning Technology Based On RFID

With the rapid development of Chinese economy, an increasing number of enterprises haveestablished their own factories, inside which the goods are stored in disorder. The level of on-sitescheduling and management directly affects the production efficiency of enterprises, how toeffectively carry out real-time tracking of the goods on the shelves inside factories is the mostfundamental premise to improving the ability of back-stage scheduling and management.This thesis adopts RFID (Radio Frequency Identification), an emerging automatic identificationtechnology, for real-time positioning of goods inside factories, which has various advantages, suchas non-contact, long distance, low cost of labels, non-visibility, multi-target fast recognition, etc.Such features of RFID perfectly make up the defect of the positioning technology of GPS in indoorpositioning.This thesis firstly introduces the research significance and objective of this project, andcompares the current situation of research on this technology at home and abroad. Then, severaltraditional positioning methods are described, and combined with the scenario set in this thesis, thepositioning system based on RSSI is selected for positioning.In the stage of positioning system design, the positioning algorithm for the two-dimensionalplane is described. Four readers are placed in the two-dimensional plane, three of which with thelargest power value of received signals are chosen to determine the estimated range of objectpositioning by using the circular model, then the boundary points of intersecting regions areweighted, and finally the final position is obtained. This thesis proposes a new3D spatialpositioning algorithm. Nine readers are placed in the space, four of which with the largest powervalue of received signals are chosen to determine the estimated range by using the sphere model,and then the centers of gravity for the intersecting bodies of each two spheres are weighted. In orderto remove the influence of the vector field in the space, four times of weighting are conductedrespectively taking four readers nearest to the object to be positioned as the benchmark. Finally, theaverage of the four results is taken as the final positioned coordinates.The final stage is experimental simulation and analysis. Firstly, keeping signal to noise ratiofixed, simulation is conducted for the algorithm proposed in this thesis, so as to plot theperformance curve of positioning accuracy by statistical analysis. Then, changing signal to noiseratio, simulation is conducted with different signal-to-noise ratios, the influence of different signal-to-noise ratios on the positioning error is obtained by statistical analysis. Finally, the spatialpositioning algorithm proposed in this thesis and traditional algorithms are compared, and we getthe conclusion that the accuracy of the algorithm proposed in this thesis is reduced compared withothers, however, taking comprehensive consideration of the improvement space, cost control andother aspects, the algorithm proposed in this thesis has certain engineering applications significance.