Optimal Trajectory Planning of a Cable-Driven Parallel Robot by Direct Collocation Approaches

Abstract

Trajectory planning in cable-driven robots is more challenging than rigid-link ones. To maintain the robot control, the cable tensions must be positive during motion. This paper presents a direct collocation approach to solve the optimal trajectory planning based on the minimization of a robot's tension and tension-rate objective functions. Besides, during robot motion, the cables must be tensile. The configuration of a cable parallel robot composed of a 3-cable and a prismatic actuator neutralizes the moving platform’s weight while improving tensionability. To generate smooth trajectories, the proposed method is compared with two standard approaches: GPOPS-II software package which uses Legendre-Gauss-Radu quadrature orthogonal collocation polynomials and direct collocation by using B-spline interpolation curves. Despite the efficiency of using B-spline functions in trajectory planning, numerical simulations demonstrate that the Hermite-Simpson direct collocation approach has a substantial benefit in the computation cost and accuracy for trajectory planning of a cable-driven parallel robot. Also, by choosing appropriate constraints and cost functions, the cable forces in the parallel robot can be well managed.

Keywords

• b-spline interpolation curves
• cable-driven parallel robot
• direct collocation
• gpops-ii
• trajectory planning