A nine-degree-of-freedom modular redundant robotic manipulator: Development and experimentation

Abstract

Performing interactive tasks in complex obstacle environments is a formidable challenge for manipulators. Redundant manipulators have the advantage of being highly flexible and can be used to perform interactive tasks in complex environments. This paper presents a nine-degree-of-freedom (DOF) modular redundant manipulator with nine identical modular joints. To improve the obstacle avoidance capability in narrow environments, the bidirectional rapidly-exploring random trees (bi-RRT) algorithm and the oriented bounding box (OBB) are combined for trajectory planning of redundant robotic manipulators. To enhance the interaction capability in uncalibrated dynamic environments, an improved virtual semi-active damping control (VSADC) algorithm based on particle swarm optimization (PSO) is proposed, and it can effectively estimate impedance parameters and positions of the dynamic environment. Finally, combined with the proposed control schemes, simulation and experimental results demonstrate that the designed redundant manipulator can traverse complex obstacle environments and interact with the unknown dynamic environment.

Modular robotic manipulator and ground assembly system for on-orbit assembly of space telescopes

“…Moreover, the redundant design of the robotic manipulator's degrees of freedom improves the flexibility and reliability of the robotic manipulator. 29,30 The ground assembly is completed by the developed modular redundant robotic manipulator, which verifies the autonomous assembly technology and paves the way for the ground verification of large space telescope on-orbit assembly technology.…”

“…Next, the assembly process of SMT4 is used as an example to illustrate the control strategy of different stages. In the first stage, the bi-RRT algorithm in our previous work 29 was used to find the path points during the assembly process of SMT 4, while the OBB algorithm was used to determine whether there was a collision between the manipulator joint and the obstacle, and the time-optimal asymmetric S-curve trajectory planning algorithm 33 that we developed for the redundant robotic manipulator was used to quickly plan a suitable motion trajectory for the end-effector of the robotic manipulator moving to the gripping point near the passive end of the adapter of SMT4. Then, based on an improved uncalibrated visual servo strategy for hyper-redundant manipulators in on-orbit automatic assembly proposed in our previous work, 34 fine adjustment of the end-effector was completed by the robotic manipulator’s end camera, which ensured that the active end adapter of the robotic manipulator was aligned with the passive end adapter of SMT4.…”

Adaptive neural controller with predefined boundary constraints for 3D printed cable-driven soft actuators

“…This makes it difficult to achieve trajectory tracking and output force control using traditional model-based control algorithms. Adaptive neural controllers have been developed to deal with such complex scenarios [34,35]. To guarantee that the tracking error ultimately converges to within the predetermined boundaries, Bechlioulis and Rovithakis [36,37] proposed a significant error transformation method, called 'prescribed performance', for the single-input/single-output strict-feedback nonlinear systems and the multiple-input/multiple-output affine nonlinear systems.…”