Research On Grasping Control Of 6-dof Manipulator

With the development and progress of science and technology,the integration of robots and society continues to deepen,and gradually penetrates into every corner of life.The manipulator is an important actuator of the robot,replacing or assisting manual tasks to complete repetitive,cumbersome and high-intensity tasks and improve production efficiency.For humans,the grasping task is simple,move to the grasp point to grab the object,move to the drop point to place the object.Since the intelligence,autonomy and flexibility of the manipulator cannot be compared with that of the human,it is of great significance to study the grasping and control of the manipulator so that the manipulator can complete the grasping task.The grasping point and the placing point are also the target points of the manipulator movement,and the end load of the manipulator will change during the grasping task.In order to control the manipulator to complete the grasping task,the following four problems need to be solved:judging whether the target point is in the working space of the manipulator;planning the path for the manipulator to reach the target point;analyzing and ensuring that the joint driving force of the manipulator meets the load changes before and after grasping;control the manipulator to track the planned path and move to the target point to complete the grab and place tasks.This paper takes a six-degree-of-freedom manipulator as an object to analyze and solve four problems in grasping tasks.The main content completed and the main results obtained are as follows:(1)Determine whether the target point is in the working space of the manipulator.Firstly,the structure of the manipulator is analyzed.Based on the improved D-H parameter and coordinate transformation theory,the forward kinematics equation of the manipulator is deduced,and the relationship between the joint variables and the position and attitude of the end gripper is determined.Then,the correctness of the forward kinematics model is verified by simulation using Robotic Toolbox in Matlab.Finally,the working space of the manipulator is calculated using the Monte Carlo method and the forward kinematics equation,and a dimension-by-dimension judgment method is proposed to determine whether the target point is in the working space.(2)Plan the path for the manipulator to reach the target point.The target point is also the expected position of the center of the gripper at the end of the manipulator.Therefore,the inverse kinematics model of the manipulator is first established by geometric method,and the method of judging the existence of the inverse kinematics solution is given.The joint variables of the manipulator are solved separately.The validity of the geometric method is verified by the forward kinematics simulation model.Then,aiming at the time optimization of the trajectory,the quintic polynomial interpolation method and the particle swarm optimization algorithm are used to plan and optimize the trajectory between the target points,and a smooth joint trajectory is obtained,and the speed and acceleration performance of the joint are fully utilized.The manipulator can reach the target point faster and improve the work efficiency of the manipulator.(3)Analyze and ensure that the joint moment of the manipulator meets the requirements of load changes.Firstly,the dynamics of the manipulator is modeled by the Lagrange method,and the influence of the end load changes after grasping on the dynamic and static performance of the manipulator was analyzed.In order to ensure that the driving force provided by the joint meets the performance requirements of the manipulator,the influence of the joint transmission system on the dynamic model is studied,and the joint motor is selected based on the dynamic model after grasping.Finally,the Simulink and Simscape simulation models of the manipulator system are built in Matlab,which lays the foundation for the design of the control algorithm.(4)Control the manipulator to to follow the planned trajectory and move to the target point to complete the grab and place tasks.First,a single-joint controller is designed with the selected motor as the object.In the position control of a single joint,torque compensation and velocity feedforward are introduced,which reduces the tracking error of the joint,and enables the manipulator to reach the target point accurately and quickly under both loading and unloading conditions.By analyzing the grasping process,the impedance control model of the grasping claw is established,and the grasping force of the grasping claw is controlled.In order to reduce the influence of nonlinear and strong coupling of the manipulator and improve the ability to track the trajectory,a model-based feedforward PD controller is designed and optimized.Simulation results show that the effect of the feedforward PD control is better than that of the PD control,and the position error between the end of the manipulator and the target point will increase after loading.Increasing the proportional gain of the feedforward PD control can reduce the effect of loading,but if the proportional gain is too large,it will increase the load.Causes jitter of the joint moment and reduces the stability of the system.By using the neural network to fit the characteristics of the joint shaking torque,the smooth joint torque can be obtained and input into the system as a new feedforward torque,so that the robotic arm can reach the target point more accurately and improve the mechanical performance.The stability and accuracy of the arm control enables the optimization of the feedforward PD control.