| Under the condition of known surrounding environment,the intelligent obstacle avoidance motion planning and control method of seven degree of freedom tomato picking manipulator is studied,and the obstacle avoidance control system is structured to realize the obstacle avoidance path planning of the manipulator.The main research contents are as follows:(1)A new intelligent obstacle avoidance path planning method is proposed,which integrates artificial potential field method,genetic algorithm(GA)and reinforcement learning(RL).Based on the principle of bionics and the traditional artificial potential field method,a new and improved artificial potential field method snake signal method is proposed.Based on the slope potential field and referring to the hunting principle of snake,the snake signal method guides the trajectory of manipulator through the height difference of potential energy in the potential field.At the same time,genetic algorithm is used to strengthen the path search ability of artificial potential field method,so that it can complete the path planning under the distribution of various complex obstacles with generally shorter path search results.Finally,reinforcement learning algorithm is used to optimize the path search results,simplify the path nodes,reduce the path length,reduce the amount of calculation of robot inverse kinematics analysis,improve the working efficiency of manipulator and reduce energy consumption.(2)The D-H parameter method is used to simulate the kinematics of the seven degree of freedom picking manipulator,and the forward kinematics and inverse kinematics of the manipulator are analyzed.Two inverse kinematics analysis methods are proposed.One is the angle sequence iteration method,which classifies the joints according to the influence of joint motion on the grasping center coordinate position of the manipulator,arranges the calculation sequence of each joint,and obtains the variation of each joint through cyclic iteration.Another method is geometric method,which analyzes and calculates the variation of each joint motion according to the workspace and structural characteristics of the picking manipulator used in this paper.(3)The motion control system of remote manipulator based on MCU and WiFi module is designed.STM32 single chip microcomputer is selected as the controller,stepper motor as the joint motor,and WiFi module is used to realize the communication between mobile phone and STM32 single chip microcomputer.Then STM32 single chip microcomputer converts the received information into pulse signal to control the rotation of each joint motor of picking manipulator.(4)In this paper,the intelligent obstacle avoidance control system and experimental platform of picking manipulator are built for picking test.The picking manipulator moves according to the planned path,and records and analyzes the time required for the manipulator to complete the path and obstacle avoidance.The planned path is divided into three types: the path planned by artificial potential field method,the path planned by artificial potential field method optimized by genetic algorithm and the path simplified by reinforcement learning on the basis of the former.The results show that through the optimization of genetic algorithm and reinforcement learning,the path length is reduced by 14.042%,the number of path nodes is reduced by 73.333%,the inverse kinematics analysis time is reduced by 48.813%,and the working time of manipulator is reduced by 47.755%.During the experiment,the manipulator avoided obstacles well and there was no mutual interference between joints.The experimental results show that the intelligent obstacle avoidance path planning and control method proposed in this paper can improve the working efficiency of the manipulator. |
PDF Full Text Request