Neural Adaptive Kalman Filter's Application in MIMU/GPS/ MMC Integrated Navigation Systems

摘要

This paper presents a novel method for sensors data fusion based on Neural Adaptive kalman filter. The method is applied in fusing data from MIMU, GPS and Micro magnetism compass integrated navigation system for Micro autonomous Uninhibited Aerial Vehicles (UAV).The noise covariance of kalman filter is modified "online" by the Neural Adaptive Controller in order to modulate kalman filter to be optimal and to improve the positioning, velocity and attitude angle accuracy of the integrated navigation system. To demonstrate the effectiveness and accuracy of this method, an example is outlined. By simulation in the MIMU/GPS/MMC integrated navigation system, it is proved that the Neural Adaptive Kalman Filter has better accuracy than the regular Extended Kalman Filter.