Hybrid Position And Force Control For Rigid Manipulators With Bounded Torque Inputs

The actuator saturation characteristics of rigid manipulators are widespread.Currently,most controllers are designed based on that the manipulator link actuators are able to generate the arbitrarily necessary torque inputs.In practice,robotic actuators have output constraints.The unlimited available torque assumption could make the controlled system become degraded or even not able to work normally.Therefore,the research of the manipulators control with input constraints is of significance.In this thesis,controllers based on the actuator saturation characteristics are studied for a class of rigid manipulators with environmental constraints.The design procedure is based on a series of control theories,which are hybrid position and force control theory,adaptive control theory,neural network control theory,the Lyapunov stability analysis theory and so on.The main work in this thesis is organized as follows.Firstly,the normal hybrid position and force control method is studied for a class of rigid manipulators with environmental constraints.The dynamic equation of the constrained robot manipulator can be expressed in a reduced form,and then a controller that based on adaptive algorithm is used to estimate the uncertainty parameters of the model.The stability of the closed-loop systems is proved by the Lyapunov stability analysis theory.The simulation study indicates that the position and the force of the constrained manipulator are effective controlled by the proposed controller.What's more,the controller is proved to have strong robustness to the uncertainty parameters of the model.Secondly,a hybrid position and force tracking controller based on adaptive algorithm is designed for a class of environmental constrained rigid manipulators with input saturation characteristics.The controller has bounded torque outputs and the bounds can be regulated.The stability of the closed-loop systems is proved by the Lyapunov stability analysis theory.The simulation study indicates the effectiveness of the proposed controller.Lastly,a hybrid position and force tracking controller that based on Radial Basis Function(RBF)neural network algorithm is designed for a class of constrained rigid manipulators with input saturation characteristics.A dynamic model is integrally approximated by RBF neural network.A robust algorithm is also considered to overcome the errors of the neural network model at the same time.The dynamic models of manipulators do not need to be linear parameterized and less rely on the model,which is better than the adaptive control.The controller has bounded torque outputs and the bounds can be regulated at the same time.The stability of the closed-loop systems is proved by the Lyapunov stability analysis theory.The simulation research indicates the effectiveness of the proposed controller.