Research On Application Of UAV Farmland Patrol Route Planning

As a new intelligent device in the field of agriculture,UAV provides powerful power for agricultural production,and is a new hotspot in the research and application of precision agriculture.Using UAV instead of human to complete the field inspection task can timely obtain the growth and appearance of crops,weeds and crop diseases and insect pests,and timely guide the fertilization,irrigation,herbicide and pesticide application and other field management work,so as to solve the problems of labor shortage and labor-intensive manual inspection.One of the key problems of UAV field inspection is route planning.The rationality of route planning directly affects the quality of UAV field patrol mission.Therefore,taking Yunyuan base of Hunan Agricultural University as the experimental scene,combined with the characteristics of UAV,this paper studies the field patrol route planning of UAV(1)In view of the fact that there is no unified UAV field inspection operation specification,this paper proposes a reasonable and effective UAV field inspection operation process.First of all,in order to observe the overall situation of a single piece of farmland,a full coverage UAV field first patrol is proposed.On the basis of UAV’s first field inspection,the disease points with low emergence rate and vigorous growth of weeds in the field are screened out.The UAV’s second fixed-point inspection for rapid and accurate positioning of the disease points can grasp the follow-up development of the disease points,which has certain practical significance and reference value for the realization of all-round monitoring of the state of farmland and its crops.(2)Aiming at the problem of full coverage route planning for UAV’s first field inspection,Canny edge detection algorithm is used to extract the ideal boundary of farmland,and then according to the range of farmland boundary,the grid method is used to establish the environment model of UAV field initial inspection.A * search algorithm is used to realize the connectivity between grids,which ensures that UAV can avoid obstacles independently and complete the inspection task smoothly during the first field inspection.The route evaluation of cattle ploughing and inward spiral full coverage patrol operation was carried out by using three indexes of operation range,turning times and coverage repetition rate.The simulation results based on MATLAB software show that cattle plough is more suitable for UAV’s first patrol in the field.At the same time,field verification is carried out.By controlling the parameters of UAV aerial photography,the image quality of aerial stitching for the first inspection is ensured,and the full coverage route planning for the first inspection of farmland is realized.(3)This paper aims at the point-to-point route planning problem of UAV’s second fixed-point inspection in the field,The Gauss Kruger projection method is used to construct the environmental coordinate system of the disease points,and the more accurate two-dimensional plane coordinates of the disease points are obtained.The adjacency matrix representation is introduced to associate the location information of the disease points,and the distance relationship between the disease points is determined.The environmental model of UAV field secondary fixed-point inspection is established.Based on MATLAB simulation platform,genetic algorithm,simulated annealing algorithm and improved genetic simulated annealing algorithm are used to solve the environment model of UAV field secondary fixed-point inspection,and the optimal scheduling routes of the three algorithms are obtained.The results show that the improved genetic simulated annealing algorithm has the best scheduling results,reduces the distance of non operational routes,reduces the energy consumption of UAV,and improves the efficiency of inspection.