Research On Local Obstacle Avoidance And Global Path Planning Of Famous Tea Plucking Manipulator

As a high value-added cash crop,famous tea has broad prospects for its development.With the development of urbanization and rising labor prices,the lack of labor has seriously affected the sustainable and healthy development of tea industry.Picking labor accounts for more than 50% of the labor in the tea garden.At present,the existing tea picking machines in the market are mainly for bulk tea.The requirements for picking famous tea are high.At present,they are still picked manually,with low degree of mechanization and strong dependence on labor,which has become the bottleneck restricting their development.With the continuous development of robot technology,robots have attracted more and more attention from all walks of life.Intelligent picking of agricultural products by robots has become a hot research field.Compared with apples,strawberries,kiwifruit and other crops,tea picking pays more attention to picking efficiency.The path planning of picking robot will directly affect the efficiency of picking operation.It is one of the key technologies of famous tea picking robot.The path planning of famous tea picking robot can be divided into two parts: global path planning and local obstacle avoidance path planning.Global path planning is used to determine the picking order of buds and leaves when picking famous tea,so as to ensure the shortest total distance in the picking process.The local obstacle avoidance path planning plans the collision free path of the manipulator moving from one picking point to the next picking point in real time,which ensures that the planning can meet the real-time performance of tea picking and the path is relatively excellent.The main research work is as follows:1)The improved max min ant algorithm is used to realize the global path planning algorithm for famous tea picking.The maximum minimum ant colony algorithm(MMAS)is determined as the basic algorithm of tea picking global path planning;According to the number,scale and distribution characteristics of buds and leaves in global path planning,the pheromone update rules,value range and path cost calculation method are determined.Combined with clustering algorithm(K-means),an improved MMAS is proposed to shorten the solution time of the algorithm by appropriately sacrificing the optimality of the solution;Tsp cases similar to tea distribution are selected as the experimental object,and the optimal parameters are selected through experiments to verify that the improved MMAS algorithm can meet the rapid planning and obtain a reasonable tea picking order at the same time.2)The improved fast random search tree algorithm(RRT)is used to realize local obstacle avoidance path planning,which meets the real-time requirements of famous tea picking.Combining visual graph method and informed sampling strategy to improve the sampling efficiency of the algorithm in the complex multi obstacle environment such as famous tea picking;At the same time,in order to ensure that the algorithm can meet the real-time requirements of famous tea picking path planning,a two-way heuristic expansion strategy is introduced;Finally,experiments show that the improved RRT*algorithm proposed in this paper can plan a better collision free path faster than other RRT algorithms in tea environment.3)The coordinate position set of bud and leaf picking points and obstacle information in tea environment are obtained by l515 depth camera,which is used for global path planning and local obstacle avoidance path planning of famous and highquality tea picking.The vision sensor used in picking path planning of famous tea was introduced;According to the transformation relationship in different coordinate systems in the process of bud and leaf imaging,the coordinate set of target bud and leaf picking points required for global path planning is obtained;The filtered tea original point cloud is transformed into a more efficient octree map,and the tea obstacle environment model required for the picking path planning of famous tea is established.4)The path planning experiment of famous tea picking manipulator is carried out.Combined with the real tea environment,the performance of the global path planning algorithm and local obstacle avoidance path planning algorithm proposed in this paper is further verified,whether the reasonable global path can be planned in real time,and then the collision free path of the manipulator moving in the tea environment is planned.This paper expounds the system framework of path planning of famous tea picking manipulator;Build the experimental platform and complete the hand eye calibration of the manipulator;Finally,the experimental results show that the local obstacle avoidance and global path planning of the famous tea picking manipulator proposed in this paper can meet the real-time and efficient picking of famous tea.