Design and implementation of robust controller for robotic manipulator

نوع: Type: thesis

مقطع: Segment: masters

عنوان: Title: Design and implementation of robust controller for robotic manipulator

ارائه دهنده: Provider: Mohammad Yazdani

اساتید راهنما: Supervisors: Soheil Ganjefar (Ph. D)

اساتید مشاور: Advisory Professors:

اساتید ممتحن یا داور: Examining professors or referees: Seyed Manouchehr Hosseini Pilangorgi Majid Ghaniee Zarch

زمان و تاریخ ارائه: Time and date of presentation: March,15, 2021

مکان ارائه: Place of presentation:

چکیده: Abstract: In this research, two free chattering robust control schemes with fixed time convergence are proposed for first and second order nonlinear systems in the presence of model uncertainties and external disturbances. Control approaches are established by introduction of novel hybrid sliding surfaces. These hybrid systems are consisting of two loops and in both internal and external loops are used from new integral fixed-time nonsingular terminal sliding modes. Using of these hybrid sliding surfaces can guarantee the convergence of system states in bounded time independent of initial conditions and without singularity problem. In addition, this structure decreases the reaching interrupt noticeable and as a result, robustness and system transient response speed enhance. In proposed approaches, adaptive super-twisting algorithm is utilized as switching control law that obtains free chattering and continuous strategy, without depending to knowledge of perturbations boundary. In present research, at first challenging problem of force sensorless position/ force control of a 3 DOF articulated manipulator is considered in uncertain environment and in presence of model uncertainties and external disturbances, then proposed control scheme for second order systems is applied to it. In this approach, model uncertainties and interaction force between robot and environment are estimated by two separate intelligent observers based on wavelet neural networks and an error estimator compensates the neural networks error. Thus, control algorithm will not be dependent on system precise model and position/ force control of system will be possible without requiring to force sensor. Besides, in contrast to presented previous research in force sensorless position/ force control field, finite time convergence of system states obtains. As well, control algorithm is designed so that provides possibility of manipulator interaction with unknown environment and smooth contact control realize during transition from free space to constrained space. Then, presented scheme for first order systems is used for torque control of permanent magnet DC motors that can achieve to finite time convergence of system states. Alike to first problem, summation of load side torque and disturbances are estimated based on adaptive neural network and compensated in control loop. Finally, Stability and fixed-time convergence of control systems are proved by Lyapunov theory. performance and effectiveness of control schemes are evaluated by simulations, then for verification of simulations results and investigation of their feasibility, practical implementations of control schemes are presented. Key Words: Integral Fixed-Time Nonsingular Terminal Sliding Mode, Adaptive Super-Twisting Algorithm, Wavelet Neural Network Observer, Position/ Force Control, Robotic Manipulator.

فایل: ّFile: Download فایل