GPS/INS Trajectory Refinement Employing Automatic Supervision and Tuning of Kalman Filters

The accuracy of positioning data derived from mobile platforms for navigation, surveillance, reconnaissance, and target designation are only as accurate as the position and orientation measurements of the platform itself.  While Kalman filter based integration of Global Positioning System (GPS) and Inertial Navigation System (INS) technology have revolutionized navigation, further refinements are necessary to allow improved tracking and designation. Presently, tuning and refinement of such Kalman Filters and associated Rauch, Tung and Striebel (RTS) smoothers within packages such as the Multisensor Optimal Smoothing Estimation Software (MOSES) require manual intervention from trained analysts to optimize trajectory and orientation estimates. The need for a skilled is one of the primary bottlenecks in trajectory and orientation refinement. In response to the need for fully automated tuning and refinement, AS&T proposes the GPS Inertial Fuzzy Trajectory Expert Determination (GIFTED) system. The GIFTED system will monitor the state of the estimation filter employed via a fuzzy logic module which will generate discrete input states for a rules-based expert system which will in turn, select appropriate tuning and refinement actions. In Phase I the refinement of simple Kalman and RTS estimation filters will be demonstrated, while in Phase II it will be integrated with MOSES. BENEFITS: The use of Carrier Differential GPS coupled with INS has revolutionized commercial aerial photography and photogrammetry. In the same manner high-precision GPS/INS systems have proven indispensable for mapping and survey operations. The use of the GIFTED system to further refine the accuracy of such application will greatly increase the value of the associated services