Flight Ad-hoc Network Ranging And Localization Algorithm For Non-cooperative Indoor Space

In non-cooperative scenarios such as urban search and rescue,the acquisition of target location is important.Collaborative positioning can only be realized through Ad-Hoc networks in urban scenarios.Restricted by time-frequency devices,the ranging accuracy cannot be achieved by one-way/two-way ranging mode.Serious non-line-of-sight observation problems make it impossible to accurately reflect the real electromagnetic signal propagation,and it is difficult to accurately model,predict and suppress NLOS signals,and the positioning accuracy decreases.It is necessary to adopt efficient and reliable positioning means to obtain high-precision position estimation to ensure the completion of the task of the drone group.In this paper,a range ranging architecture based on multiple polling responses and dynamic NLOS decision and a collaborative positioning method based on MDS assisted multilateral positioning are proposed to realize continuous and accurate relative positioning of multiple drones in non-cooperative environments.The main research contents of this paper include:First of all,the geometric ranging mode of multiple polling and response is designed,which not only makes the difference difference between different drone clocks,but also reduces the influence of frequency difference to millimeter level through simultaneous solution of multiple response information by using the objective condition that frequency is relatively stable in a very short observation period.Then,in view of the non-cooperative architectural space cannot be established by the NLOS identification threshold of problem,this paper proposes a NLOS identification algorithm based on dynamic threshold criterion,through geometrical observation distance between drone real-time collection and statistics,for unknown dynamic decision threshold of interior space,and in different scenarios by threshold adaptive adjustment,improve the effectiveness of the NLOS identification.Finally,in view of the failure of observation between two points due to the influence of NLOS,a multi-dimensional scaling method is proposed to assist multilateral positioning technology to achieve continuous relative positioning of the drone.Based on the artificial intelligence navigation test field of the State Key Laboratory of Satellite Navigation System and Equipment Technology,collaborative positioning tests were carried out,and the validity and reliability of ranging and positioning algorithms were verified.