Constant Rotation Ratio of Redundant Robot Manipulators Based on Neural Dynamics

In this paper, a constant rotation ratio (CRR) algorithm based on neural dynamics was proposed to solve the trajectory accuracy of motion planning for redundant robot manipulators. In the motion planning, the pseudo-inverse of the Jacobian matrix was solved by the neural dynamics network solver. The real-time and accurate joint velocity was acquired by the fourth-order Runge-Kutta methods to solve the average velocity ratio of each joint, and to obtain the optimal joint angle for the CRR. Differing from the traditional motion planning by weighted pseudo-inverse solution, the CRR algorithm was designed to solve at the joint angle level, being the key approach to improve the trajectory accuracy of redundant robot manipulators. In addition, computer-simulation models with three-link and six-link planar robot manipulators were established. Simulation results show the validity of the CRR based on neural dynamics algorithm.