Trajectory Planning for Delta Robot Pick and Place Tasks Based on the Improved Grey Wolf Optimizer: An Experimental Validation

Abstract

Delta parallel robots are widely used in industry to run operations like sorting, mixing, shaping, labeling, assembling, packaging and pick and place. In order to reduce the trajectory length while performing repeated pick and place tasks, this work suggests an optimization approach that deals with a point to point trajectory planning for Delta parallel robots. At first, the trajectory is defined by five key points to be fitted during the task. Then, a non uniform b-spline curve is adopted to interpolate the initial key points where its weights and the knot vector are used as the decision variables for the optimization problem. Besides, the improved grey wolf optimizer technique is proposed to solve this problem. To find the optimal solution, the improved grey wolf optimizer population based metaheuristic uses two different search strategies, a group search and an individual search strategy. Simulation results show that the optimization of the non uniform b-spline weights and knot, using the improved grey wolf optimizer, can provide a length-optimal smooth trajectory. In comparison with b-spline, with fixed weights and knot, the obtained results also present an improvement of 2.65% of the trajectory length at each sequence. Experimental validation demonstrates that the proposed approach can be very useful in generating smooth trajectories while reducing the burden in repeated pick and place tasks.