Real-time Obstacle Avoidance and Position Control for a Soft Robot Based on Its Redundant Freedom

The problem of obstacle avoidance for a soft robot is studied by using the motion redundancy of the soft robot when the tip of the soft robot is controlled to the desired position. The kinematic model of the soft robot based on piecewise constant curvature hypothesis is established. And a controller for both position control of the tip and real-time obstacle avoidance is designed based on this model. In the algorithm, a warning zone is artificially divided around the obstacle. With the position feedback of the marker points on the soft robot, the moving strategy of the tip when it's outside the warning zone, and the moving strategies of the middle points as well as the tip when they're inside the warning zone are given. The required control parameters are solved by the generalized inverse matrix of the Jacobian matrix. Then the stability of the inverse Jacobian control is verified using Lyapunov theory. Finally, experiments are implemented in 2-D space. The results show that the tip can reach the target point, and meanwhile the soft robotic arm successfully avoids the obstacle, which verifies the effectiveness of the obstacle avoidance algorithm and the stability of position control.