Research On Full Coverage Route Planning Algorithm For Plant Protection UAV

Plant protection UAVs have the advantages of fast operation speed,low operation cost,little environmental pollution,and good prevention and control effect.With the development of aviation technology,control technology,artificial intelligence,and other technologies,the degree of automation of plant protection UAVs is increasing,however,there is still a certain distance from the fully autonomous operation.To further improve the ability of automatic operation of plant protection UAVs,this paper firstly studies the automatic identification method of plant protection UAVs,and then aims at the problem of whether there are obstacles in the boundaries of the plots,and plans the full coverage route that can adapt to different boundary types.method to conduct research.The main research contents and conclusions of this paper are as follows:（1）The automatic identification method of plant protection UAV operating plots was studied.The traditional image segmentation method and the image segmentation method based on deep learning are studied respectively.The image segmentation method based on deep learning is more robust and has better generalization ability compared to the other method.The research results show that,compared with the traditional image segmentation method,the image segmentation method based on deep learning can realize the rapid identification of land parcels,and can obtain competitive performance and model effect with a small amount of data sets.（2）A full-coverage route planning method for plant protection UAVs adapted to different boundary types is proposed.Aiming at the problem that there are obstacles on the boundary or not of the operation plot,a full-coverage route planning method for plant protection UAVs adapted to different boundary types is proposed.For the case where there are no obstacles on the border,the waypoints are expanded outwards to ensure maximum coverage.In the case of external obstacles,in order to ensure the safety of plant protection UAVs,the waypoints are retracted to avoid the border.（3）A program for a full-coverage route planning of plant protection UAVs with automatic identification of a land plots is developed.Based on Python and HTML/Java Script languages,a full-coverage route planning program for plant protection UAVs with land identification function was designed and implemented.Finally,the simulation verification is carried out through a numerical example,the data results of the route planning are analyzed,and the flight defense suggestions for different boundary types are given.