Novel terrain integrated navigation system using neural network aided Kalman filter

Abstract

A terrain integrated navigation system is proposed to adapt the characteristics of the underwater environment and high accuracy requirements of AUV navigation, which is composed of the strapdown inertial navigation system (SINS), Terrain-aided navigation system (TAN),the Doppler velocity log (DVL) and the magnetic compass(MCP). An improved federated Kalman filter based on the back-propagation neural network(BPNN) for adjusting the information sharing factors is designed and implemented in the AUV integrated navigation system. Linear filter equations for the Kalman filter and measurement equations of navigation sensors are addressed. Simulation experiments are carried out according to the mathematic model. The comparable results indicate that the AUV navigation precision and adaptive capacity are improved substantially with the proposed sensors and the intelligent Kalman filter.