Tightly coupled integrated inertial and real-time-kinematic positioning approach using nonlinear observer

Tightly coupled integration of inertial measurements and raw global navigation satellite measurements using a nonlinear observer is proposed for real-time-kinematic applications. The position, linear velocity and attitude estimates of a rover is aided by global pseudo-range and carrier-phase measurements from the rover and a base station. The tight integration is achieved using a modular observer design, where the attitude observer with gyro bias estimate is based on a nonlinear complementary filter. Two translational motion observers estimating position and linear velocity using single- or double-differenced range measurements, respectively, are presented. The range measurements include pseudo-range and carrier-phase measurements from the satellite constellation to rover and base station, where the integer ambiguity resolution is estimated as part of the nonlinear observer state vector. The feedback interconnection of the observer systems is shown to be exponentially stable. The proposed observers are tested with an unmanned aerial vehicle simulator.