Adaptive Control for a Constrained Soft Manipulator with Prescribed Performance

Soft manipulator is more safe than rigid manipulator when working with people, but it’s also more difficult to design and control. Unlike some existing works establishing kinematic model with static analysis, the kinematic model is established by using piecewise constant curvature methods and is divided into two subsystems. For the purpose of promoting the shape control for soft manipulators, an adaptive control scheme is proposed for the soft manipulator in this paper. In the framework of dynamic state feedback, the unknown function can be approximated by neural network (NN), and therefore an adaptive controller is devised. Furthermore, it is proven that the output tracking errors satisfy the stipulated performance by applying the time-varying barrier Lyapunov function. Finally, the simulation illustrates the rationality of the proposed control scheme.