Algorithm Of Information Fusion For Integrated Train Positioning System

Low cost GPS/INS integrated positioning system plays a vital role in train positioning and becomes one of hot topics both home and abroad. Kalman filter is generally applied in train positioning, and information fusion algorithm based on Kalman filter in GPS/INS integrated positioning system can improve positioning continuity and accuracy. Research on information fusion algorithm applied in train positioning has practical significance.GPS/INS integrated positioning system is selected as the research object, and Kalman filter is selected as the foundation of positioning algorithm. Fusion results affected by three different Kalman filter algorithms in different parameters and models are compared and analyzed in detail. Speed threshold is determined according to statistical angles measured by GPS. Algorithms are designed and compared among three-dimension filtering, four-dimension filtering and mixed-filtering. The information fusion algorithm with dynamic delay Kalman filter and four-dimension mixed adaptive Kalman filter is proposed.Data collected from experimental car or Qinghai-Tibet railway is tested and analyzed for GPS and INS in detail. Algorithms are compared and evaluated in both qualitative and quantitative ways. Fusion results demonstrate that the proposed algorithm has high accuracy, robust performance, and prospective application in train positioning. Software of GPS/INS integrated positioning platform is designed to process all the information needed by algorithms in integrated environment.