Development and Flight Test of a Robust Optical-Inertial Navigation System Using Low-Cost Sensors; Master's thesis

摘要

This report develops and tests a precision navigation algorithm fusing optical and inertial measurements. This algorithm provides an alternative to the Global Positioning System (GPS) as a precision navigation source, enabling navigation in GPS denied environments, using low-cost sensors and equipment. A rigorous study of the fundamental nature of optical/inertial navigation is accomplished through the examination of the observability grammian of the underlying measurement equations. Through this analysis, a set of principles of design are proposed guiding the development of a navigation algorithm. An algorithm titled the Simultaneous Location Aiding And Mapping Recursively (SLAAMR) is thus described incorporating these principles of design, an extended Kalman Filter (EKF) and a Simultaneous Location And Mapping (SLAM) process. The SLAAMR algorithm is designed to provide robust navigation performance in realistic, full scale environments at a low cost. The principles of design and the SLAAMR algorithm are tested and evaluated using data collected at the United States Air Force Test Pilot School (USAF TPS). A full- scale aircraft flying operationally representative parameters and profiles was used to collect the data, and was correlated with highly precise Time Space Position Information (TSPI) truth data for validation and evaluation purposes. The flight test data supports the principles of design and highlights the challenges faced by any opticallybased navigation system. The resultant performance of the SLAAMR algorithm provides a robust, practical navigation solution to Air Force aircraft.