Robotic arm time–jerk optimal trajectory based on improved dingo optimization

Time and jerk are two pivotal factors in the optimal trajectory planning of a manipulator. A trajectory that minimizes time and jerk can ensure smooth and efficient operation of the manipulator. This paper introduces a novel, general time–jerk optimization method for manipulators, which integrates quintic non-uniform rational B spline (QNURBS) with an improved dingo optimization algorithm based on simulated annealing, namely SA-DOA. This method has been successfully applied to plan the time–jerk optimal trajectory of the Kinova Jaco2 robot arm. With maximum angular velocity, acceleration, and jerk as constraints, and running time and total jerk as optimization objectives, the method employs SA-DOA to determine the optimal time interval and then utilizes QNURBS to generate the angle curves of the six joints in joint space, based on the specified path point and optimal time interval. Experimental results demonstrate that SA-DOA can obtain the global optimal solution within a limited number of iterations. Compared with dingo optimization algorithm (DOA), particle swarm optimization algorithm (PSO), and simulated annealing (SA), the optimization results have improved by 26.57%, 21.87%, and 0.26%, respectively. The time–jerk optimal motion trajectory enhances the execution efficiency of the manipulator and produces less jerk during movement, thereby improving motion control performance.