Optimal Trajectory Planning for Manipulators with Efficiency and Smoothness Constraint

Abstract

Path planning to generate an appropriate time sequence of positions for a complex trajectory is an open challenge in robotics. This paper proposes an optimization method with the integration of an improved ant colony algorithm and a high-order spline interpolation technique. The optimization process can be modelled as the travelling salesman problem. The greatest features of this method include: (1) automatic generation for complex trajectory and a new idea of selecting the nearest start point instead of using the traditional way of human operation; (2) an optimized motion sequence of the manipulator with the shortest length of the free-load path improves efficiency by nearly 65% and (3) trajectories both in Cartesian space and joint space are interpolated with good smoothness to reduce shocks and vibrations. Simulations and experiments are conducted to demonstrate the good properties of this method.

Optimizing Robotic Task Sequencing and Trajectory Planning on the Basis of Deep Reinforcement Learning

“…Lee Y D et al [18] then attempted to use GA to solve the time-optimal trajectory planning for a two-link manipulator. With the development of intelligent algorithms, some algorithms with good performance have been introduced to address the robot trajectory planning problem, such as the DE algorithm [19], the PSO algorithm [20,21], the ant colony algorithm [22], the whale algorithm [23], and so on. However, focusing solely on minimizing time as the objective in robot trajectory planning frequently results in significant vibrations of the robot, leading to a loss of tracking and positioning accuracy and even exacerbating motor wear.…”