Research On Path Planning And Force Tracking Control For Autonomous Assembly Robot

The intelligent assembly and collaborative assembly of robots is the specific requirement of"Made in China 2025".The existing robot technology has realized the force servo control.In the face of diversified market demand,the existing assembly robots are less intelligent and have lower assembly efficiency.Therefore,it is urgent to use artificial intelligence technology to achieve rapid and efficient intelligent assembly of robots.Based on the six-degree-of-freedom robot with visual/force-sensing fusion,this paper studies the gravity compensation and the gripper calibration of assembly robot sensors,workspace path planning,and force tracking control.1.Assembly robot six-dimensional force sensor gravity compensation and the gripper calibration.The sensor space force model is established.The sensor data of N(N?3)groups under different loads is obtained.The least squares method is used to analyze and conduct experiments to obtain the sensor zero point value.The load center of gravity coordinates and the load gravity value accurately compensate the external environmental forces of the robot.Based on the visual guided grab axis model,the tool coordinate system position calibration is solved by the least squares method,and the attitude calibration is calculated by Halcon hand-eye visual calibration.Finally,the experimental calibration is carried out to obtain the transformation matrix of the tool coordinate system to the end coordinate system,which is prepared for the accurate grasp of axis in path planning.2.Path planning of assembly robot.An obstacle avoidance path planning algorithm based on improved artificial potential field method(local)and APF-RRT(global)is proposed.Firstly,according to the six-degree-of-freedom assembly robot model,the floating control point and fixed control point are introduced,the attraction potential field function is modified,the relative distance between the repulsion potential field and the target point is introduced,and the adj usting factor is added to realize the potential field regulation,so that the robot can escape towards the target point when it encounters obstacles;secondly,aiming at the problem that the potential field method is easy to fall into the minimum,the target gravity probability of APF is calculated.In order to make the robot move towards the target point and jump out of the local minimum,the RRT search tree expansion stage is introduced.Finally,the proposed algorithm is simulated in multi-environment with MATLAB,which proves that the algorithm can realize the motion planning of the six-degree-of-freedom assembly robot from the grasp to the assembly position,and can adapt to the changes of the environment.3.Research on the force tracking control algorithm of assembly robot.The force analysis of the six-dimensional force sensor is carried out when the shaft hole is in contact,and the mathematical model of the impedance control of the shaft hole assembly contact is established,and the force of the sensor is converted into the actual contact force of the shaft hole,and the impedance control based on the end force feedback is performed;The steady-state force error of the process impedance model is analyzed.Simulink is used to build the robot impedance simulation system.The control variable method is used to identify the desired impedance parameters to obtain the appropriate impedance parameters.For the uncertain environment,the model reference adaptive control algorithm is added.The assembly system performs online force level adjustment.Simulation experiments show that the impedance control added to the adaptive system is in the environment facing the following environment:plane position and sudden change of stiffness,slope,plane to slope transition,surface,plane to surface transition,etc.The tangential position tracking response is reduced from 2s to 0.2s,the steady-state contact force response is reduced from 4s to 0.4s,and the normal contact force tracking error is reduced from 4N to 0.3N.The adaptive impedance has better environmental adaptability than impedance control,accurate force position tracking simulation of flexible assembly is realized.4.Path planning experiment of assembly robot.The experimental scheme is designed,and the improved potential field method and APF-RRT fusion algorithm are used to carry out path planning in barrier-free environment,which verifies the effectiveness of the application of the algorithm.On the premise that there are obstacles in the spherical envelope cuboid box in the robot workspace,the proposed algorithm is used to carry out path planning,and the ability of obstacle avoidance path planning of the algorithm is verified.