Model-Free Ground Velocity and Position Estimation for Manned and Unmanned Aircraft

Navigation is a critical task for the operations of both manned and unmanned aircraft systems. Current positioning systems rely primarily on satellite systems such as the Global Positioning System (GPS) or alternative sensor fusion algorithms, which typically require vision sensing and processing. Due to the possibility of temporary GPS outages and/or GPS jamming, it is critical for aircraft sensing systems to predict the position as well as the ground velocity of the aircraft in the absence of GPS signals. This work proposes two state estimation algorithms for predicting the position and ground velocity of aircraft. These methods do not require vision sensors or aircraft dynamic model information, thus providing a portable approach applicable to any aircraft. The proposed methods consider infrequent GPS position updates. Although not completely GPS free, these algorithms do not require GPS velocity measurements and can predict the aircraft position in between the position updates. The proposed methods use the information filter and unscented information filter; they are first validated using unmanned aircraft flight data and later applied to flight data from a high-speed manned military trainer jet. The results indicate the effectiveness of this approach for model-free position and ground velocity estimation.