The University of Alberta's Applied Nonlinear Controls Lab (ANCL) helicopter unmanned aerial vehicle's (UAV's) existing GPS-aided inertial navigation system (INS) does not provide observability of heading angle during hover. This motivates the integration of a triaxial magnetometer aiding sensor into the extended Kalman filter (EKF)-based navigation algorithm. A novel magnetometer calibration procedure is implemented and compared with conventional approaches. Experimental results in ground and flight testing confirm that the observability issue is resolved and demonstrate improvements in attitude estimation.
展开▼