Research On Multi-UAV Collaborative Search Method Based On Random Targets

In recent years,with the advancement of technology,economic development and the reduction of raw material costs,UAV have been used in the military industry from the beginning to the civil and commercial fields.In the same time,UAV have been widely used.Target search is one of the classic application scenarios of UAV Due to the complexity of tasks,single UAV is often limited by their own portable resources and functions when performing these tasks.Therefore,the collaborative operation of UAVs has become a hot spot for researchers and an important trend in the development of future UAV This paper mainly studies the multi-UAV collaborative operation to complete the search of random targets,realizes the UAV task allocation by Voronoi diagram,completes the path planning of the single UAV based on Dubins curve,and adopts the centroid V-graph method to achieve multi-UAV collaborative search.The main research contents are as follows:1、This paper introduces the current research status of multi-UAV collaborative search and UAV flight path planning,and describes the commonly used UAV mission environment modeling method and path planning algorithm.2、For multi-UAV collaborative search,it is first necessary to complete the task allocation.This paper introduces a way to divide area based on the Voronoi diagram,which uses the location of the UAVs and the no-fly zone as the input site for Voronoi diagram.Because each Voronoi cell needs to be obtained vertex sequence,this paper designs an algorithm for constructing Voronoi diagrams in bounded regions,which solves the task allocation of multi-UAV collaborative search.Task allocation transforms multi-UAV collaborative search problem into a single UAV search for a given area which reduces the difficulty of path planning.3、According to the characteristics of the no-fly zone in the area that each UAV is searching for,this paper uses Dubins curve to solve the problem of target search path planning for single UAV When the UAV detects the target,UAV needs to fly to the target to complete the task of collecting target information.Because the target has no direction angle limitation for the UAV,the trajectory shape of UAV when flying to the target is arc-line.This paper use way that solve calculation of Dubins curve to complete the calculation of the key points of this trajectory to realize collecting the target information.This paper completes the path planning of a single UAV based on Dubin curve.4、According to the characteristics of random targets,the UAV needs to perform patrol detection targets in the area to be searched.This paper realizes the multi-UAV collaborative search path planning algorithm by adding the no-fly zone to improve the centroid V-graph method.Base on Qt program development framework realizes the multi-UAV collaborative search visualization platform.The experimental results and the spiral coverage algorithm are compared and analyzed to vertify the effectiveness and practicability of the method.