Research On Motion Planning And Human-Robot Collaboration Technology Of Redundant Manipulator For Fruit-Oriented Operation

The labor in the process of picking and packing economic forest fruit products,often in great demand,has problems of low efficiency and low accuracy and is now facing shortages due to the COVID-19 epidemic,which has inflicted the development of the whole industry.This has also brought out other problems such as excessive customization of automation equipment and large industrial robots for economic fruit-oriented operations,followed by high R&D costs and maintenance as well as low versatility.At present,the lightweight collaborative manipulator has been used to replace these devices,but it only has six degrees of freedom and unable to meet the task for the more complex space.Meanwhile,the problem of fruit posture calculation combined with environmental constraints in the fruit picking and packing is not well solved.It is rarely considered that there are obstacles or the path obstacle avoidance problem crossing the workspace and the trajectory synchronization problem for multi-point continuous motion.It is not taken into account in the current research how to update task as well as how to integrate the key human-robot collaboration technology perfectly into the actual operation scene under the condition the manipulator grabs the wrong fruits.Focusing on the shortages of the motion planning and human-robot collaboration of fruit products in economic forest fruit picking and packing,this paper calculates the posture that satisfies the environmental constraints and generate the joint configuration calculated by inverse kinematics with redundant optimization;Aiming at the fruit picking and packing,the path to avoid obstacles and synchronous trajectory planning algorithm for achieving motion planning is designed;This paper studies human-robot collaboration technology to fix the deficiency of planning for fruit picking and packing motion.Main research contents are as follows:(1)Study the fruit poses generation technology that satisfies environmental constraints.In the scene of deep-container gripping fruits,the manipulator needs to generate obstacle avoidance grasping posture,to solve this problem,fruit pose calculation methods are designed based on deep-container region segmentation and remote center of motion.A hybrid inverse kinematics algorithm based on the combination of heuristics and numerical iteration is proposed,combined with redundant optimization,joint configuration of the manipulator for grasping fruits can be achieved under multiple constraints.The results of experimental show that calculating the grasping posture combined with redundant optimization,can achieve the optimal joint state solution that meets the environmental constraints,and the error at the singular point is significantly reduced compared with existing research methods(2)Study the trajectory planning synchronization technology.In order to achieve the task of smooth motion to the target joint state meeting the obstacle avoidance constraints and fruit picking,a multi-dimensional space synchronization trajectory planning technology is proposed based on full speed quadrant motion and speed braking technology.The algorithm is applied to the joint space of the manipulator to complete the task of fruit picking and packing,achieving good movement.The results show that the algorithm can achieve synchronous motion,avoiding collisions in picking tasks and squeezing fruits caused by asynchronous.(3)Research on the path planning technology of the manipulator in the fruit operation.If depending on manual teaching and fixed path setting methods in practical applications,reprogramming is needed after changing picking scene or deep container.In order to solve the above problems,the DRRTC algorithm is proposed,based on the discretized map,to obtain the initialized priori map path,which is used as the sliding constraint of the configuration space sampling,and the search space is compressed to improve the efficiency.Through simulation and experimental validation,the results show that the proposed DRRTC algorithm can meet obstacle avoidance and human task constraints to generate a smooth path completing the fruit task,which is less time-consuming,more motion stability.(4)Study the force interaction control strategy in human-robot collaboration.Due to the diversity of products,quality control cannot be described qualitatively in fruit operations.In order to solve the problem that the manipulator cannot be intervened in time and modify the motion control decision after grasping error,the research on the dynamics and force control is carried out.The three-loop dynamics identification method and dynamic threshold collision detection algorithm are further proposed to improve the accuracy of dynamics estimation and collision detection.A new motion scheme can be formulated according to the duration of the external force,and the task of fruits accurately packing can be completed collaboratively.The results of experiments show that the manipulator can accurately produce decisions for human intervention,and then coordinate to complete the accurate fruits packing,which fix the deficiency that the changes of motion decision-making in fruits packing cannot be realized by relying only on motion planning,thereby improving operation efficiency and accuracy.In summary,this paper studies motion planning and human-robot collaboration technology of redundant manipulator for fruit-oriented operation in a comprehensive way.From the results of many simulations and real-world robot experiments,the proposed method can effectively improve the production efficiency.Its ideas and methods not only satisfy the requirements of fruit-oriented operation,but also are available for forestry agile manufacturing.