A hybrid approach for GPS/INS integration using Kalman filter and IDNN

In the field of navigation, two system depends for the navigation is Inertial Navigation System (INS) and Global Positioning System (GPS). INS is an autonomous, standalone system which is available in the aircraft which gives the latitude, longitude and altitude position of the aircraft by the accelerometer and gyroscope measurements. GPS is used to provide accurate position information for the navigation. Both the system having its own drawbacks like bias error and drift error of accelerometer and gyroscope and satellite clock errors and multipath reflection errors of GPS. In order to overcome its drawback, both the systems are integrated to provide reliable navigation solution. Typically Kalman Filter (KF) is used to integrate the system. In recent years, AI (Artificial Intelligence) based systems are used for the same. In this paper, the hybrid approach of using Input delayed Dynamic Neural Network (IDNN) and KF is introduced for GPS/INS integration.