Particle Swarm Optimization Based Time Optimal Pick-place Trajectory Planning In A Dobot Manipulator

Trajectory planning of manipulator is one of the main research directions of manipulator motion control system.Its core task is to ensure the continuity and smoothness of joint position,velocity curve as well as the continuity of acceleration curve of each joint,and further obtain the end effector trajectory with smooth operation and no violent shaking,so as to reduce joint wear and improve work efficiency.In this paper,Dobot Magician,a four degree of freedom desktop manipulator,is taken as the actual research object.A practical application scene based on the robot arm is designed,and the planning method of how to get the time optimal trajectory is discussed.There are various methods for trajectory planning of industrial robot in joint space.Firstly,using B-spline function or low-order polynomial method for trajectory planning is easy to cause sudden changes in velocity and acceleration,which will increase the joint wear of the manipulator and reduce its life;secondly,a single high-order polynomial method can obtain continuous smooth position,velocity curve and continuous acceleration curve,it will increase the computation and is easy to cause Runge phenomenon.Therefore,the three-segment combined polynomial trajectory planning method is proposed in this paper,which divides the trajectory of the manipulator into three segments artificially to avoid using a polynomial function to fit all the path points and causing the Runge phenomenon,and greatly reduces the computational complexity.In addition,due to the combination of different orders and the introduction of acceleration constraints of related path points,the three-segment combined polynomial method can avoid the problems of uncontrollable acceleration and unsmooth trajectory of end effector while using single polynomial trajectory planning.On the other hand,using single polynomial or combined polynomial trajectory planning method for trajectory planning only takes the nature of the trajectory into account.And in actual production,under the condition of meeting the basic requirements,the efficiency and energy consumption problem should be further considered.Based on this,the improved particle swarm optimization(PSO)algorithm is used in this paper for the purpose of optimizing the trajectory obtained by combined polynomial method to reduce the time consumed in the whole operation process of the industrial robot,to achieve the ultimate goal of not only continuous and smooth joint position,velocity curve and continuous acceleration curve,but also smooth operation,optimal time of the end effector trajectory.Finally,a practical application scene based on the Dobot manipulator is designed in this paper,and the single polynomial and combined polynomial method are used for trajectory planning.Through comparison,it is found that the combined polynomial method not only obtains the continuous and smooth curve of joint position and joint velocity,but also its joint acceleration curve is continuously controllable.After getting the time-optimal trajectory by the improved PSO,a certain number of path points are selected according to the equal step size to control the manipulator to complete the corresponding motion,which verifies the practicability of the trajectory obtained by this method in the actual manipulator.