Cooperative Obstacle Avoidance for Underwater Dual-arm Manipulators based on Traversing Adjacent Space Method

Xinyang Xiong,Xianbo Xiang,Yuhao Liu,Shaolong Yang

DOI: https://doi.org/10.23919/oceans44145.2021.9705888

2021-01-01

Abstract:Obstacle avoidance is a difficult subject in motion planning of marine robotic systems, especially for control of underwater dual-arm manipulators for underwater interventions, for instance, manipulation for marine biology and underwater archaeology, etc. Complicated ocean environment makes it tough to prevent the manipulator from collision with obstacles in dynamic environment, and collision avoidance between two arms is also complicated in such environment. In this paper, a novel obstacle avoidance control algorithm for underwater dual-arm manipulators is proposed, by using the Traversing Adjacent Space (TAS) method. Combined with the collision detection function, manipulators scan the adjacent work space, identifying the location of end-effector nearest to the target point, and reach the target gradually without collision. Moreover, a master-slave motion control mechanism is studied in this paper. The slave-manipulator moves after the master-manipulator, regarding the master-arm as obstacle, which successfully solves the difficulty in collision avoidance between two arms. The TAS method is computationally efficient, since obtaining a large number of random points is not required in TAS, and the collision-free trajectory to the target is normally the shortest. To avoid manipulator’s trapping in local optimum, generating virtual target is introduced to the paper. Simulations and experiment results show the good effectiveness and efficiency of the TAS obstacle avoidance control algorithm, which performs well both in obstacle avoidance and collision avoidance between two arms. In addition, the prospect of hydrodynamic modeling is presented after the simulation, which is important for future research on the robust control of underwater manipulator.