Control of the 2-DOF Underactuated Manipulators Based on Neural Network

The control problem of the underactuated manipulators system is one of the major topics in control theory and robotic engineering. In this paper, we apply the neural network method to analyze the 2-DOF joint manipulators and show the underactuated joint can be stabilized at the equilibrium point through the periodic disturbance of the actuated joint. And then, we design an open loop control strategy based on above results using neural network and use it to achieve the nonholonomic manipulator position control. We show the effectiveness of the neural network method and verify it by some simulation results