Application Of A New Adaptive Kalman Filitering Algorithm In Integrated Navigation System

Inertial Navigation System (INS), which can calculate the position and orientation of amoving object in any circumstances worldwide, is a kind of self-aid navigation system thatworks without the need for external references or radiant information outside. However,astime passes, with errors of INS cumulated, its precision decreases. Contrary to INS, the moretime GPS works, the higher precision it has.Hence, integrated navigation system of INS/GPSis a relatively ideal combination pattern which can complement each other, and is widely usedin many fields-such as aviation, navigation and aerospace.This article focuses on the Error signal of INS/GPS integrated navigation system, whichis the main part of navigation data fusion. First of all, this article studies the theoretical basisfor kalman filtering, which is part of the optimal estimation theory, analyzes the causes thatmake it divergent and how to inhibit it. Secondly, introduces classic sage-husa adaptivekalman filtering, based on what this article put forward how to improve it.In the last place,introduces several ways designing filter for INS/GPS integrated navigation system. Accordingto the need of this research, this article adopted a loose combination of position and velocity.Based on it has been given the strapdown inertial navigation system’s error、 the GPS ’serror model and the equation separately, then infers INS/GPS inertial navigation system’sequation of state and the gauging equation.The selection has the representative guidanceparameter, establishes corresponding the simulation model,uses the classics kalman filteringalgorithm and the improvement Sage-Husa adaptive filtering separately in assigns under theguidance path to carry on the simulation.The simulation result indicated that,the relativeconvention Kalman filtering speaking of a system mode stricter request, the improvementSage-Husa adaptive filtering may be possible to suppress the filter radiation effectively in asystem mode more loose situation the question, and can maintain filter’s precision and thereliability.