Continuous Trajectory Planning Based On Robot Dynamics

Based on the determined path in the path planning stage,industrial robot trajectory planning is a method of interpolating the path into a curve expressed by a time function.Kinematics trajectory planning with the robot kinematic constraints is generally feasible under low speed and light load conditions.But there may be large tracking errors when the robot runs in the high speed and heavy load situations.Compared with the kinematic trajectory planning,the dynamic constraints was considered by the dynamic trajectory planning,and the planning results are safe,stable and efficient.But the dynamic trajectory planning calculation efficiency is low,and it cannot meet the real-time calculation requirements of industrial production.Therefore,this thesis proposes a dynamic-based,constraint self-adjusting look-ahead trajectory planning method.It can meet the dynamic constraint requirements and has a good real-time performance.The dynamic model of the robot is the basis of dynamic control and planning.Since the dynamic parameter values cannot be obtained directly,and an identification experiment is designed,the parameters are identified by an optimization identification method with physical feasibility constraints.The least-square method is used to identify the inertial parameters of the robot system's load and convert the parameters to the robot body.Then a complete robot dynamics model is obtained.Although the robot trajectory planning algorithm based on reachability analysis solves the numerical integration method's low robustness,this algorithm's real-time calculation efficiency is not high enough.This thesis improves it to be a continuous segmented trajectory planning.By selecting a suitable path length for segments,this method can meet approximate real-time planning needs.Since the above method cannot strictly meet the online calculation requirements,a robot look-ahead trajectory planning algorithm that the dynamic equation is utilized to modify constraints is proposed.Kinematic trajectory planning is firstly used to calculate and verify whether the robot joint torque on each path point exceeds the limits;Map the corrected calculated joint torque to the acceleration constraint,and recalculate the kinematic trajectory planning to obtain the final result.Finally,the dynamics parameters of the robot are identified in simulation experiments and actual experiments,then the constraint self-adjusting look-ahead trajectory planning algorithm is verified through experiments.This algorithm satisfies the dynamic constraint conditions and realizes the goal of online dynamic trajectory planning.