A Miniature Integrated Navigation System for Rotary-Wing Unmanned Aerial Vehicles

Abstract

This paper presents the development of a low cost miniature navigation system for autonomous flying rotary-wing unmanned aerial vehicles (UAVs). The system incorporates measurements from a low cost single point GPS and a triaxial solid state inertial/magnetic sensor unit. The navigation algorithm is composed of three modules running on a microcontroller: the sensor calibration module, the attitude estimator, and the velocity and position estimator. The sensor calibration module relies on a recursive least square based ellipsoid hypothesis calibration algorithm to estimate biases and scale factors of accelerometers and magnetometers without any additional calibration equipment. The attitude estimator is a low computational linear attitude fusion algorithm that effectively incorporates high frequency components of gyros and low frequency components of accelerometers and magnetometers to guarantee both accuracy and bandwidth of attitude estimation. The velocity and position estimator uses two cascaded complementary filters which fuse translational acceleration, GPS velocity, and position to improve the bandwidth of velocity and position. The designed navigation system is feasible for miniature UAVs due to its low cost, simplicity, miniaturization, and guaranteed estimation errors. Both ground tests and autonomous flight tests of miniature unmanned helicopter and quadrotor have shown the effectiveness of the proposed system, demonstrating its promise in UAV systems.