Research On Fault-tolerant Integrated Navigation Of UAV Based On The Multi-source Information Fusion

The UAV suffer from severe electro-magnetic interference and human errors during its missions.Besides,the platform motion characteristics and flight environment make a strong demand of the navigation system's autonomy,reliability and accuracy that a single combination of navigation system can not satisfy.Therefore considering the flight characteristics of UAV,this paper deals with the fault-tolerant integrated navigation technology based on multi-source information fusion,with a view to a comprehensive set of high-precision,high-reliability integrated navigation solution algorithms that adapted to drones.Firstly,concerned with the application background and work characteristics of UAV,the Federal filter SINS/GPS/CNS/SAR multi-navigation program is constructed.In order to solve the shortage that the traditional scalar information sharing coefficients can not correctly reflect the differences between local state variables,this paper proposes an adaptive federated filtering algorithm based on information factor optimization,which effectively improves the navigation accuracy.Secondly,considering the failure of UAV,this paper studies the fault detection algorithm of integrated navigation and conducts a thorough research on the residual chi-square detection algorithm.Regarding the disadvantages of the algorithm,which has low sensitivity to small information fault,this paper proposes an improved residual chi-square detection algorithm based on eigenvalue and dual-threshold.To further consider the asymptotic changing and components of small information fault,a kind of adaptive fault-tolerant algorithm based on sequential probability mapping is proposed.The quality of local estimation states under small information fault is calculated by sequential probability ratio test and then used in adaptive tuning of global fusion.This algorithm is proved to improve the fault-tolerant performance under small information fault.Finally,base on the fusion of multi-source information integrated navigation algorithm and two fault detection algorithms,this paper has established the fault-tolerant integrated navigation simulation,and a digital simulation system is designed based on multi-source information fusion and navigation,which provides the simulation environment for UAV auto-navigation technology.