Model-Less Feedback Control for Soft Manipulators with Jacobian Adaptation

In this paper we propose a novel closed-loop model-less feedback control scheme for soft manipulators with Jacobian matrix adaptation. As the major merit, this approach does not rely on a specific kinematic model and thus has wide applicability and portability for different kinds of soft manipulators. We make use of closed-loop control, convex optimization and adaptation strategy to estimate the Jacobian matrix for trajectory tracking tasks. In the simulation validation, we adapt the algorithm to a two-section soft manipulator of piece-wise constant curvature to track a desired trajectory. Furthermore, we compare our algorithm with two other model-less algorithms and a model-based algorithm with noise on their tracking performances. Validation results prove the feasibility of this simple yet effective model-less feedback control scheme.