Confined spaces path following for cable-driven snake robots with prediction lookup and interpolation algorithms

While cable-driven snake robots are promising in exploring confined spaces, their hyper-redundancy makes the collision-free motion planning difficult. In this paper, by combining the prediction lookup and interpolation algorithms, we present a new path following method for cable-driven snake robots to high-efficiently slither into complex terrains along a desired path. In our method, we first discretize the desired path into points, and develop the prediction lookup algorithm to efficiently find the points matched with joints of the robot. According to geometric relations between the prediction lookup results and link length of the robot, we develop the interpolation algorithm to reduce the tracking errors caused by the discretization. Finally, simulations and experiments of inspections in two confined spaces including the obstacle array and pipe tank system are performed on our custom-built 25 degree of freedoms (DOFs) cable-driven snake robot. The results demonstrate that the presented method can successfully navigate our snake robot into confined spaces with high computational efficiency and good accuracy, which well verifies effectiveness of our development.