Sanal gerçeklikle hesaplamalı zeka temelli robot kolu kontrolü

This work addresses the problem of trajectory tracking control of the robotic manipulator. Different methods are proposed in kinematics and dynamic control of robotic manipulator. In inverse kinematics, two methods are presented based on closed loop strategy. First method is based on proportional Derivative (PD) like fuzzy controller while second method exploits the effectiveness of Sliding Mode Control by designing a robust method with solving the problem of singularity. Different desired Cartesian trajectories are used to illustrate the effectiveness of proposed methods with two links and 4 Degree of Freedom (DOF) SCARA robots. Obtained results reveal the performance of proposed methods. In dynamic control of robotic manipulator, efficient robust control methods are proposed for controlling robotic manipulator subjected to external disturbance and model uncertainties. Unlike most existing nonlinear robust control schemes, the proposed control methodologies do not require the exact dynamic model of robotic manipulator. The proposed controller's gains are selected by using Lyapunov stability theorem. Three robust control methods have been proposed, proportional–Sliding Mode Control (P-SMC), Hybrid Computed Torque Control (CTC)-SMC, and adaptive SMC. P-SMC requires upper bound of uncertainty while in Hybrid CTC-SMC method only nominal dynamic model of robot manipulator is required. These requirements have been avoided in the third proposed method by using adaptation technique. Linear matrix inequality technique is applied to select the gains of linear part of proposed controller in P-SMC and Hybrid CTC-SMC and Lyapunov stability theorem is used to derive the updating laws for the controller gains. The performances of proposed methods are compared with other different methods by simulating these methods applied on a two-link robotic manipulator for different desired trajectories. Moreover, performance index of integral absolute error is used to measure the performance of each method. All proposed theorems and methodologies are considered for the general robotic manipulator regardless of the degrees of freedom. Simulation results illustrate in a comparative fashion the performance of proposed methods in terms of cumulative error, robustness against disturbances and uncertainties, and trajectory tracking.