Trajectory Planning And Grasping Of Manipulator Based On ROS

With the rapid advancement of technology,more and more industries tend to be automated and intelligent,and robotic arms have become a part of role in the modern industry field,active in the assembly lines of various factory workshops,undertaking increasingly complex work tasks.Therefore,this paper will focus on the series manipulator,combined with the related work of manipulator trajectory planning and grasping.First of all,based on the laboratory self-developed five-axis lightweight robotic arm,this paper takes the robotic arm as the research object,establishes the DH parameter table of the model according to the structural size of each joint.Based on this,the forward and inverse kinematic model of the manipulator are derived.And the kinematic simulation experiments based on MATLAB are conducted to verify the correctness of the kinematic derivation.Then the maximum torque required for each joint of the robot arm is determined according to the working range of the robot arm,use this as a constraint to complete the selection of the main hardware facilities.Then hardware platform of the robot arm is assembled and built,and finally the programming of the robot arm control system is completed by combining the hardware modules and relevant communication protocols.Secondly,according to the classification of trajectory planning in different spaces,trajectory planning of cubic polynomial and quintic polynomial in joint space is introduced,and the motion parameters of both are compared through simulation experiments,which shows that the manipulator has better smoothness of motion when higher-order polynomial algorithm is used;furthermore,the principle and process of interpolation about spatial straight line and circular arc in Cartesian space are introduced,and the experiments of spatial straight line and circular trajectory planning in MATLAB combined with Robotic Toolbox are conducted respectively.Then,for the problem of trajectory optimization,the single-objective trajectory optimization based on jerk minimization is studied,focusing on the segmented polynomial expression of a complex trajectory in the optimization process,as well as establishing the objective function and related constraint equations for the optimization objective.The effectiveness of the optimized trajectory is analyzed by comparing the original trajectory with the optimized trajectory through simulation experiments and combining with the motion parameter curve.It is confirmed that the optimization results are helpful to solve the problem of excessive impact and vibration during operations.In the follow-up,we further study the multi-objective optimization problem based on the shortest time and the least jerk.Sometimes,there are contradictions and conflicts between multi-objectives.Therefore,NSGA-II algorithm is used to solve the multi-objective optimization problem.The Pareto front of the optimization solution set can be calculated,and then the best optimal solution relative to other positions can be selected according to the weight coefficient.In the simulation experiments of the multiobjective optimization problem,it is shown that the trajectory optimization does shorten the running time and mitigation of jerk by comparing the trajectory curves and the graphs of running time and motion parameters.Finally,based on the robot operating system(ROS),complete the system construction of the entire robotic arm platform,including the camera hand-eye calibration in the vision system,then the detection and localization of the target object is completed by using the SIFT algorithm.Based on the above work,a real environment robotic arm grasping experiment was conducted.The depth camera is used to detect and calculate the pose information of the target object.The motion trajectory curve is calculated using trajectory planning,and combine the forward and inverse kinematics model to solve the rotation angle of each joint.Then the hardware system drives joints of the manipulator to control the end effector of robotic arm to reach the target position to complete the pick-and-place action,and the feasibility of the research work is verified by actual experimental process.