Deeply coupled GPS/INS integration using adaptive prediction filter

Abstract

Most applications using Global Positioning System (GPS) require constant, highly accurate navigation data with available satellite signals. GPS error sources can lead to reduction in accuracy of navigational information relevant to position, velocity, and attitude. For this reason, the integration of GPS and Inertial Navigation System (INS) produces a high-precision navigation system. In spite of considerable progress in recent years, it is still possible to improve the performance of this integration system. This thesis addressed Deeply Coupled GPS/INS Integration method based on using Adaptive Prediction Filter (APF) to increase accuracy and reliability of navigation data to mitigate effects of data collection errors. The main problem is outage or weakness of the GPS signal. There are several reasons for the outage of GPS signals, such as tunnels, high-rise buildings, urban canyons, heavy foliage, and high mountains. Reasons for weakness in a GPS signal include multipath signals, tropospheric effects, satellite orbit changes, etc. Represented in this thesis are the simulation and analysis of the INS system and its errors with detail components of X-axis, Y-axis, and Z-axis acceleration and velocity components, and INS performance in Euler angles (pitch, roll, and yaw) to find the attitude of a rigid body. Simulation and analysis of GPS with errors in latitude, longitude, and height and also represented here. Simulation trajectory for a vehicle on a banked figure-eight track has been proposed in this research.