Path Planning Of Greenhouse Mobile Robot Based On Improved RRT

It is of great significance to study a greenhouse autonomous operation basic platform with strong adaptability and capable of carrying various sensors and actuators to accelerate the construction of modern greenhouses.Therefore,this paper takes the greenhouse mobile robot and the global path planning of the greenhouse mobile robot based on the Rapidlyexploring Random Tree as the research direction,and builds a mobile robot platform suitable for the greenhouse,and focuses on the research on the global path planning algorithm of mobile robot.The main research contents are as follows:(1)Combined with the results of the field investigation in the greenhouse and the task requirements of the mobile robot,a mobile robot platform based on the Robot Operating System was built: First,the mechanical structure of the mobile robot was modeled through Solidworks,and then determining its master-slave control structure and writing the upper and lower layers of code,and finally deploying the individual modules on the mobile robot.(2)To meet the requirements of single-target point operations such as picking and harvesting on mobile robots in greenhouses,a goal-biasing RRT based on Max Turning Angle Constrained is proposed.Through the realization and analysis of the basic RRT and the goal-biasing RRT,the MAC-RRT integrates the maximum steering angle constraint of the differential mobile robot into the goal-biasing RRT on the basis of studying the kinematics model of the mobile robot.Restricting node expansion conditions enhances the smoothness of the path,making it more suitable for tight greenhouse environments.(3)In view of the slow planning speed of MAC-RRT in the greenhouse environment and the lack of security of the path,a goal-biasing RRT with efficient branch pruning and path smoothing under expansive map is proposed.BSP-RRT improves the security of the path in the greenhouse environment through the expansion preprocessing of the map.By improving the node expansion method that the target is biased towards RRT,it makes it more greedy to approach the target point,which improves the convergence speed of the algorithm.Finally,by cutting Post-processing methods such as branching,interpolation,and Bezier curve path smoothing further optimize the path,reduce the path cost,enhance the path smoothness,and greatly improve the applicability of the algorithm to the greenhouse environment.(4)A multi-objective path planning algorithm integrating BSP-RRT and ant colony optimization is proposed for the needs of multi-objective point operations such as handling and inspection in the greenhouse.The algorithm uses BSP-RRT to plan the collision-free path between each target point and enter the cost matrix,and then calculates the heuristic factor between each target point through the cost matrix,and finally optimizes the optimal running order through the ant colony algorithm iteratively.The optimization of the global path of multi-objective points in the greenhouse is completed.(5)The stability of the mobile robot platform control system is verified by the mobile robot debugging software,and the feasibility of the improved algorithm is verified as an experimental platform.