The Enhanced Performance of an Integrated Navigation System in a Highly Dynamic Environment.

Abstract : For the U.S. Air Force to maintain an accurate and reliable Navigation Reference System (NRS) with Carrier-Phase Global Positioning System (CPGPS) measurements, it must develop an accurate and robust NRS in the face of cycle slips caused by highly dynamic maneuvers. This research investigates the implementation of a double differencing between receivers/satellites scheme to improve the accuracy of current NRS models. The removal of the 'perfect Doppler velocity aiding measurements' (a very poor assumption of past research) was completed with stable and accurate results. The double differencing implemented showed improvement in the accuracy of the NRS. An investigation of two Failure Detection, Isolation, and Recovery (FDIR) algorithms for large cycle slip failures is conducted. The two FDIR techniques are the Chi-Square test and a Multiple Model Adaptive Estimator (MMAE). The FDIR results show that a Chi-Square test as a stand-alone algorithm can work accurately for detection and isolation of failures with an accurate and reliable recovery algorithm. The MMAE algorithm as conjectured seems to be the best FDIR technique to handle single and multiple cycle slips accurately and reliably.