| With the rapid development of chip technology,control technology and industrial technology,harvesting,plant protection and other equipments in the agricultural field are also developing towards automation,informatization and intelligence.Because the robot has the advantages of large load,sustainable operation and reducing the harm of liquid medicine to human body,it is of great significance to use the robot instead of manual for greenhouse plant protection.Path planning is one of the core technologies of greenhouse spraying robot,which has a great influence on improving the working efficiency of the robot.In order to realize the safe and efficient operation of greenhouse spraying robots,this thesis studies the plant protection requirements of greenhouse orchards,proposes an improved A* and Dynamic Window Approach(DWA),which improves the path-finding strategy of A* algorithm,and improves the ability of DWA algorithm and A* algorithm to achieve uniform application of crops on both sides under the premise of obstacle avoidance in greenhouse.The main work of this thesis is as follows :(1)The operation requirements of greenhouse spray robot are introduced,and the software and hardware platforms of the system are analyzed,including the software and hardware of the robot based on ROS system.Meanwhile,based on the analysis of the environment map model and the mapping algorithm,the map model and the mapping algorithm suitable for the greenhouse robot are selected.(2)The global path planning algorithm is studied,and an improved A* algorithm which can plan the middle line path between crop rows is proposed.The path planned by the traditional A*algorithm may lead to uneven spraying of crops on both sides of the greenhouse spraying robot during operation.In order to solve this problem,the heuristic function of A* algorithm is improved to optimize its path finding strategy.By calculating the reciprocal of the diagonal distance of the obstacle closest to the feasible point and using it as a new heuristic function term,the A* algorithm pathfinding evaluation standard is improved,so as to plan a path suitable for greenhouse spraying robot plant protection.Then the robot can be guaranteed to spray along the center line between crop rows during the operation in order to ensure the utilization of the liquid and the safety of the operation.Finally,the effectiveness of the improved A* algorithm in greenhouse environment is verified by simulation experiments.(3)Aiming at the problem that DWA algorithm is easy to fall into local optimum,the evaluation function of DWA algorithm is optimized to improve the fusion effect of global path planning algorithm and local path planning algorithm.Firstly,the relationship between the weight of the evaluation function and the time cost is determined by simulation experiments,and the better weight of the evaluation function is selected to improve the working speed of the robot.Considering that the robot needs to cooperate with global and local path planning,the sub-target point distance evaluation function is constructed in the original DWA evaluation function term to ensure that the DWA algorithm can efficiently track the path planned by the improved A*algorithm under the premise of dynamic obstacle avoidance.(4)In order to make the greenhouse spraying robot perform dynamic obstacle avoidance on the basis of the global optimal path,and solve the problem of many turning points and large corners in the global path,a path planning method for greenhouse spraying robot based on improved A* and DWA algorithms is proposed.Firstly,based on the improved fusion algorithm,the guidance point optimization fusion algorithm is added to realize the global path planning of coverage operation and the real-time dynamic obstacle avoidance of the greenhouse spraying robot.Then,the experimental scene is designed through the MATLAB platform to verify the ability of the fusion algorithm to plan the paths and avoid obstacles.Subsequently,based on the Gazebo physical simulation platform,the simulated greenhouse environment is built,and the plug-in improved path planning algorithm is written into the ROS path planner to further verify the reliability of the algorithm.Finally,the improved algorithm is applied to the greenhouse spraying robot to verify the operation in the real environment.The experimental results show that the path planned by the improved algorithm is smooth and the path curvature is continuous.Meanwhile,the path can be well located in the center line between crop rows.The dynamic obstacle avoidance effect is also good,which has a certain reference for the research of greenhouse spraying robot in path planning. |
