Browsing by Author "Yilmaz, BM"

Now showing 1 - 4 of 4
  • No Thumbnail Available
    Item
    Velocity observer design for a class of uncertain nonlinear mechanical systems: a self-adaptive fuzzy logic-based approach
    Yilmaz, BM; Tatlicioglu, E; Selim, E
    This study focused on designing a smooth velocity observer (VO) for mechanical systems whose mathematical model is uncertain. The uncertainties that appear in the observer dynamics are, via utilising their universal approximation property, modelled with fuzzy logics. A novel self-adaptive fuzzy logic (SAFL)-based term in which control representative value matrix (CRVM), centres and widths of membership function are all dynamically updated is used as part of the observer design. Through the application of Lyapunov-type stability analysis techniques, the practical stability of the observed velocity error was guaranteed. The outcomes derived from experimentation on a planar robotic manipulator are showcased to illustrate the performance of the devised VO design.
  • No Thumbnail Available
    Item
    Adaptive Neural Network-Based Backstepping Control of BLDC-Driven Robot Manipulators: An Operational Space Approach with Experimental Validation
    Unver, S; Yilmaz, BM; Tatlicioglu, E; Saka, I; Selim, E; Zergeroglu, E
    This study concentrates on end effector tracking control of robotic manipulators actuated by brushless direct current (BLDC) motors, having parametric uncertainties in their kinematic, dynamical and electrical sub-systems. Specifically, an operational space controller formulation is proposed that does not rely on inverse kinematics calculations at position level and still ensures practical end effector tracking despite the presence of uncertainties related to the mechanical and electrical dynamics, and the kinematics of the robotic manipulator. Compensation for the uncertainties throughout the entire system is achieved via the use of neural network-based dynamical adaptations, and the overall stability of the closed-loop system is guaranteed via Lyapunov-based arguments. We would like to note that the work addresses the following problems: (i) incorporation of actuator dynamics into the error system in order to achieve increased efficiency, (ii) elimination of the need for position level inverse kinematics calculations for the controller formulation to remove the computational burden and (iii) compensation of the uncertainties throughout the entire subsystem. Experiment studies were carried out on a two degree of freedom planar robot manipulator equipped with BLDC motors to evaluate the effectiveness of the proposed formulation.
  • No Thumbnail Available
    Item
    An adaptive self-adjusting fuzzy logic-based robust controller formulation for a class of uncertain MIMO nonlinear systems
    Yilmaz, BM; Tatlicioglu, E; Zergeroglu, E
    This study presents a novel continuous controller, in conjunction with a fuzzy logic-based estimator, designed to address the compensation of parametric uncertainty in a category of high-order, multiple-input-multiple-output nonlinear systems. The proposed controller-estimator methodology tackles parametric uncertainties with self-adjusting adaptive fuzzy logic-based robust integral of sign of error algorithm. In the employed adaptive fuzzy logic (AFL) framework, the means and variances of the membership functions are updated online in each iteration, enabling a more accurate estimation of uncertainties. The boundedness of the closed-loop system and asymptotic stability of the error signals are verified via Lyapunov-based arguments. Numerical simulations are additionally presented to evaluate the efficacy of the proposed methodology.
  • No Thumbnail Available
    Item
    Fuzzy Logic based adaptive control of robot manipulators driven by BLDC Motors
    Yilmaz, BM; Unver, S; Selim, E; Saka, I; Tatlicioglu, E
    This research focuses on designing a joint space controller for robot manipulators whose joints are driven by brushless DC motors. By appending the actuator dynamics to control synthesis and stability analysis, the sensitivity in the sense of positioning and reliability of the robot manipulators is aimed to increase. Uncertainties in both robot manipulator and actuator dynamics are taken into account to obtain improved tracking performance. To overcome these uncertainties, self-adjusting fuzzy logic networks, that is having adaptively updated centers and widths, are used. The developed adaptive controller structure does not rely on acceleration measurements and incorporates only full state feedback. Via utilizing Lyapunov type stability analysis methods, semi-global uniform ultimate bounded tracking result is ensured. To further verify the applicability of presented controller, comparative tests are performed on a two degree of freedom planar robot manipulator equipped with brushless DC motors and satisfactory tracking performance is demonstrated.