Taylor, Brian2014-10-142014-10-142014-10-10https://hdl.handle.net/11299/1668472: Guidance: wapoint_guidance_fast.c <- 150 meter x 150 meter square route initiated by autopilot on. Navigation: AHRS: AHRS_6state.c DR: DR_11state.c INSGPS: INSGPS_15state_quat.c Blender: Blender.c GPS Outage begins when Autopilot is on and the run_num is 2 (i.e. one waypoint route with GPS, one without). GPS outage ends when autopilot is flipped back off. Winds: light and variable out of the east Purpose: Test parallel filters and blending routine on standard UAV Lab Flightcode to ensure no issues. Result: Everything performed as expected. Good tracking with GPS (INSGPS weighted more heavily) and decent tracking without GPS (DR weighted more heavily). Position drift to West due to winds out of the East. Added Bonus: GPS initialized accurately but then diverged to 1000's of meters of error on all three position channels for a brief (<30 seconds) time. This behavior was seen in prior flight tests and would cause filter divergence. New innovation checking and blending routine maintained accuracy throughout bad GPS data and recovered well once GPS became accurate again.enTyrFlightUMoreNavigationWaypointGPSDeniedEKFFilterDataset