A New Fuzzy Adaptive Filtering for Relative Navigation System
Qingchang Ji · Electronics Optics & Control · 2012
In UAV formation flight,the relative position of the lead aircraft and follower needs to determined,which is to solve the navigation problem.Since the UAV has great maneuverability during formation flight,the inertial measurement components prone to having bias that may affect the motion state equations of the system.Sometimes statistical properties of the system noise and observation noise can not be obtained accurately.To solve the problems,an improved adaptive filtering method was proposed.According to the difference of the actual covariance of real-time measurement and the theoretical covariance,and the mean of the measurement values,an appropriate filtering method was selected based on the conditions.Then,the designed fuzzy inference systems was used to adjust the system noise matrix and the measurement noise matrix,or adjust the state error covariance matrix,thus the UAV formation flight could keep the original formation even under harsh environment.Simulation results show that the algorithm has fine adaptability.