Trajectory Planning and Nonlinear Control of a Pneumatic Flexible Manipulator With a Gripper Using Neural Networks

Pneumatic flexible manipulators are often used in grasping and manipulation tasks, there exist nonlinear characteristics and singular configurations in the motion processes. To enhance dynamic control performance, this article proposes a trajectory planning and nonlinear control strategy based on adaptive radial basis function (RBF) neural networks. The trajectory planning is developed for the pneumatic flexible manipulator using a particle swarm optimization (PSO) algorithm combined with a cubic spline interpolation (CSI) method. To achieve the grasping and handling tasks of the pneumatic flexible manipulator, a backstepping integral sliding mode controller is designed based on the adaptive RBF neural networks. The adaptive RBF neural networks are employed to provide online compensation for load disturbances from the gripper. Experimental results of a grasping and handling task have shown effectiveness and superiority of the trajectory planning and nonlinear control strategy in this article.