Research On Collaborative Method Of UAV For Flight Inspection

With the gradual promotion and application of drones in the military and civilian fields,in order to improve the efficiency of drone operations in complex air environment situations and ensure the safety of drone airspace flight,the cooperative operation of multiple drones has become particularly important.UAV collaboration technology not only guarantees the flight safety of the drone,but also is the core technology to realize the autonomous and intelligent of the drone.Collaboration of drones,including the use of sensors and data link data to effectively observe,evaluate,and determine threats to the operating environment.Based on this,multiple UAVs have multiple features and large clusters,covering a wide range of features,UAVs can share perceived content,realize collaborative work,and improve the efficiency of completing tasks.The subject is aimed at the application of drones in the field of civil aviation.It uses multi-drone cooperative operations instead of routine crews to inspect the civil aviation aircraft before and after flight.It focuses on the perception and planning of multi-drone collaborative sensing and evasion.The following two core topics have been studied:Firstly,A set of design schemes for multiple unmanned aerial vehicle collaborative surveys is proposed.With safety,accuracy,and efficiency as the core indicators,the joint optimization modeling from flight path planning and fleet size;for the appearance model of civil aviation aircraft,with reference to the inspection criteria of civil aviation aircraft before and after flight,construct a standard-compliant Drone bypass process.It provides an important basis for the experiment of the collaborative winding system.Secondly,A multi-UAV track planning solution is proposed,which is divided into track node assignment and global track planning.Among them,the track node assignment is to set the tracks of each drone according to the mission requirements,and plan multiple flight tracks offline;the global track planning uses the differential evolution particle swarm algorithm(DEPSO)to optimize the fleet size and introduce Differential evolution updates the particle swarm,adjusts the inertia weight of the particles through an adaptive method,calculates the cost function under the prescribed constraints,and selects the optimal fleet size based on the cost function index.Thirdly,a collaborative method suitable for multi-UAV roll-out is proposed.Aiming at the operation requirements of multiple drones in complex environments,the message publish / subscribe mechanism in MAVROS is adopted to enable the drones to communicate with ground stations and other drones in real time,enabling data sharing and sharing between multiple drones to achieve Collaboration of inspection tasks to improve the efficiency of inspection work.Finally,The hardware and software platform of the test system is set up for multi-UAV coordinated inspection.Conducted multi-UAV wrap-around experiments,and verified the rationality of the trajectory planning and the rapidity of multi-UAV wrap-around through the actual experiments of single-UAV roll-over and the simulation experiments of multi-UAV roll-over.The statistical verification was carried out through multiple experiments,and the stability of the multi-UAV coordinated winding system was examined.