Research On The Obstacle Avoidance Path Planning Optimization And Prediction Of Plant Protect UAV

Plant protection unmanned aerial vehicle(UAV)has the advantages of high efficiency,good working quality,wide adaptability and high flexibility.It is attracting more and more attention,and its applications are becoming more and more common.With the development of aeronautical technology,control technology and artificial intelligence,the degree of UAV automation has been improved rapidly.How to realize autonomous operation with low energy consumption and high efficiency has become the main research focus of the plant protection UAV.To improve the efficiency and quality of spray operation,path planning optimization has become one of the most important basic research issues in its autonomous operation system.In this paper,the path planning of plant protection UAV sprayed in non-obstacle environment and obstacle environment is studied.Firstly,in the non-obstacle and obstacles environment,a path planning optimization algorithm based on genetic algorithm is proposed separately.Secondly,according to the improved Dubins path,a turning path optimization algorithm is designed in accordance with the maneuverability of UAV.Thirdly,a track points extraction algorithm based on curvature feature is proposed considering the influence of different curvatures on the difficulty of track tracking.Finally,a tracking prediction algorithm based on particle filter algorithm is achieved for uncertain disturbance state of UAV.The main research work and results are as follows:(1)This paper analyzes and studies the existing path planning algorithms for plant protection UAV.To solve the problems existing in the current path planning algorithm,an algorithm based on genetic algorithm is proposed to optimize the flight path of UAV by encoding the UAV's heading angle.The algorithm can reduce the number of turns during plant protection while taking into account the length of track.This study can effectively reduce the energy consumption of UAV plant protection operations.(2)A path planning optimization algorithm is proposed for the path planning with multiple obstacle constraints.First,after modeling the common obstacles in the spraying area,an obstacle avoidance path planning optimization algorithm based on improved Dubins path was proposed.Second,when there are many obstacles in the spraying area,the genetic algorithm(GA)is used to search optimal obstacle avoidance path.Third,for turning in the path planning,a turning path optimization algorithm is presented considering the maneuverability of UAV.Fourth,the obstacleavoidance path planning is simulated in MATLAB.The results show that the proposed algorithm can decrease the area of overlap and skip to 205.1%,while the path length is increased only by1.6% in comparison with the traditional Dubins obstacle avoidance algorithm under the same conditions.Finally,the obstacle avoidance path planning is verified by experiments and the experimental results verify the feasibility of this algorithm.(3)Aiming at the obstacle avoidance path generated by the optimization algorithm based on genetic algorithm in this paper,a track points extraction algorithm based on curvature feature is proposed considering the influence of different curvatures on the difficulty of track tracking.When the curvature is 0,the constant C is set to accommodate the track point extraction.In order to extract the representative track points better in the path,the points are also extracted as track points if they are 2 changing points whose curvatures in the obstacles avoidance route are positive,negative.(4)Aiming at the track prediction to avoid obstacles of the plant protection UAV,a tracking prediction method of UAV was achieved based on the algorithm of particle filter under the situation of uncertain disturbances.The simulated result by MATLAB shows that proposed algorithm can achieved the tracking prediction better with less smaller error and better robustness when there are uncertain Gaussian noise,wind speed below 2m/s as well as error and random noise of the sensor existing.The simulation results provide a reference for the UAV spraying operation environment,and it has certain guiding significance for the plant protection UAV operation environment.