Please use this identifier to cite or link to this item:
|Title||Trajectory Tracking Control of A 2-DOF Robot Arm Using Neural Networks|
This thesis investigated several control strategies to handle the trajectory tracking problem for a two degree-of-freedom (2-DOF) robotic arm using artificial neural networks (ANNs). Feed-forward two layer neural networks were designed and utilized in both model-based and non-model based control structures to conduct online learning and identification of the inverse dynamics of the robotic manipulator and to compensate for both structured and unstructured uncertainties. The simulation results obtained proved the superiority of the proposed neural network controllers to dramatically reduce the error between the desired and actual position trajectories even in the presence of uncertainties unlike other conventional methods such as the PD-computed torque method. The neural network-based controllers proposed in this thesis provide solutions to the trajectory tracking problem of robotic manipulators with or without a mathematical model which would make them effective controllers for both planned and unplanned trajectory tracking problems for any degree of freedom robotic manipulator. The development of the mathematical models for the 2-DOF robotic arm and its joints driving motors as well as their simulation experiments were carried out under the Dynamic Modeling Laboratory (Dymola) environment which uses the Modelica object-oriented multi-domain system modeling language. The simulation results obtained in the thesis were accompanied by three dimensional (3D) figures in order to visualize the results and to help establish a deeper analysis and understanding of these results.
|Publisher||الجامعة الإسلامية - غزة|
|Files in this item|