A Cable-Driven Hyper-Redundant Robot with Angular Sensing

Hyper-redundant robots are usually used in confined space for the inspection and operations because of their dexterity and conformability. However, these robots usually use the kinematics to calculate their shape for control, and errors could be introduced inevitably due to the assembly, cable elongation, etc. This paper develops a cable-driven multi-section robot with angular sensing units for confined space. Kinematics based on D-H convention is built, and an optimization algorithm based on two-layer iteration is developed to achieve the follow-the-leader control. A feedback control using angular sensing is implemented to improve the overall accuracy. Also, force sensors are also employed to measure the tension on each actuation cable at the proximal end. Different target curves are generated for testing the feasibility, and motion simulations are conducted to successfully demonstrate the robot's insertion. Finally, six experiments on a three-link robot prototype are carried out to validate the proposed kinematics and control algorithm. Results show that the average positioning error for the tip is 5.27 mm.