Research On Key Technology Of Unmanned Aerial Vehicle All Source Navigation Based On Multi-Source Information Fusion

With the development of the national defense in our country,the requirement of unmanned aerial vehicle is higher.In order to improve the navigation and positioning accuracy of unmanned aerial vehicle,research institutions are doing more discussion and research on it.The satellite navigation is easy to be influenced by environment,and the navigation and positioning performance also can be disturbed.Because of that,“All Source Positioning and Navigation” appeared.The article regards the unmanned aerial vehicle as the background,and analyzes and researches the All Source Positioning and Navigation based on the multi-source information fusion.The article is divided into three parts,including the ASPN scheme,sensor selection mechanism,multi-source information fusion algorithm and fault detection.First of all,for ASPN scheme of the unmanned aerial vehicle,the thesis researches on the design thinking of the ASPN scheme,and analyzes the key technologies,and then designs the ASPN scheme of unmanned aerial vehicle.At the same time,based on the ranking rules of the navigation sensors,the option scheme of the subset of sensor is designed and the selection mechanism of the sensor subset is studied.Taking the high altitude unmanned drones as the example,the thesis constructs the flight environment and scene,and analyzes the sensor subset selection mechanism,and lays a foundation of the research on multi-source information fusion algorithm and fault detection.In order to deal with the issue of out-of-sync data and information,and adapt to the condition that each of the navigation sensor is unavailable,and meet the requirements of the changing environment and tasks demands,the thesis studied the multi-source information fusion algorithm based on factor graph.The thesis represents the status updates and measurement updates process of the navigation sensor through the open architecture of factor graph.And the thesis compares the multi-source information fusion algorithm based on factor graph and federal filtering algorithm by simulation.Taking the high altitude uav flight environment and scene as example,the thesis analyzes the multi-source information fusion algorithm.At the same time,restructuring work mode after isolating fault to optimize system performance.In order to solve the problem that navigation and positioning accuracy is influenced by the fault information of navigation sensors,the thesis studies the fault detection algorithm to detect hard fault and soft fault in the navigation system.The residual chi-square detection algorithm is effective for hard fault,but it is not effective for soft fault.Therefore,the soft fault in the navigation system is detected by improved fault detection algorithm.Finally,the paper sets up the high altitude uav navigation system simulation platform based on MATLAB software.The software platform realizes and simulates the sensor subset selection,multi-source information fusion algorithm based on factor graph and improved fault detection algorithm.