The Application Of Kalman Filtering In INS/GPS Integrated Navigation

The inertial navigation system(INS)can provide the information continuously,has superior precision in short time, but position error accumulates with time.The GPS has long-term stability, but is disturbed easily and the data renewalfrequency is low. As a result of MEMS component shortcoming, the inertialnavigation system with MEMS component needs assistance from GPS. INS/GPSintegrated navigation system achieves good navigation accuracy, obtainswidespread application in actual project.This thesis introduces basic principle and composition of GPS and INS,infers the system error equations, gets the error mode, studies application ofkalman filtering in the INS/GPS integrated navigation system mainly. UsingMATLAB simulation language, the general filter, the federated filter and theadaptive filter are researched and simulated.MEMS component bias is not constant, it undulates along with time andenvironment. If it is not considered, introduces the constant bias to thestrap-down system, affects the system precision. This paper studies improvedkalman filter--the kalman filter with unknown and variant bias, analyzes itsapplication background, infers its algorithm. Bias is introduced in theequations, estimated with states vector respectively. Then the bias estimaterevises state estimation, updatas the state estimation. According to thesimulation of the experiment data, the improved kalman filter achieves superioreffect, is valuable in practical application.