Relative Navigation Algorithm Based on Robust Filter for UAV Formation Flight

In order to solve the problem of determination of the relative position and the relative velocity of formation flight unmanned aerial vehicles, the relative navigation algorithm based on H-infinity filter with non-Gaussian noise input and the system model parameter uncertainty is developed. The system investigated here consists of two Inertial Navigation systems, ranging Radar and Vision sensor. The necessary and sufficient existence condition of desired estimator and its design method are solved by convex optimization based on linear matrix inequality. Numerical simulation results show that the filtering algorithm has a valid estimate of the relative status between the two machines when Kalman filter does not work as good in the case of non-Gaussian noise and system model uncertainty.