طراحي كنترل كننده فازي بر مبناي الگوريتم ژنتيك براي ربات موازي شبيه سازهاي حركتي با محرك هاي دوراني

چكيده انگليسي :

Fuzzy Controller Design Based on Genetic Algorithm for a Parallel Manipulator of Motion Simulators with Rotary Actuators Hamid Azizan hamidazizan@gmail com 5th April 2011 Department of Mechanical Engineering Isfahan University of Technology Isfahan 84156 83111 IranDegree MSc Language FarsiSupervisor Mehdi Keshmiri mehdik@cc iut ac irABSTRACTDesign of a fuzzy controller based on genetic algorithm for a 6 DOF parallel manipulator is studied in thisthesis In the analytical part the study deals with kinematic and dynamic analysis of the manipulator as wellas its control synthesis via Parallel Distributed Compensation PDC method In the experimental part implementation of the controller on a prototype of this kind has been carried out This manipulator isbasically made up of two platforms The base platform is fixed to the ground and is linked to the movingplatform through six legs Each leg comprises two links with a universal joint connection in between Thelower link is connected to the base platform through a revolute joint and upper link is connected to themoving platform through a spherical joint Kinematic constraint equations are extracted in both algebraic and differential forms As a result theforward and inverse kinematics of the robot are solved The full nonlinear dynamic equations of themanipulator are derived using Lagrange s method for constrained systems Using orthogonal complement ofthe constraint Jacobian matrix and eliminating the Lagrange multipliers dynamic equations are reduced to aset of six independent differential equations As an assumption flexibility and looseness of the joints as wellas joint frictions are ignored in the modeling Based on the kinematic and dynamic analysis of the manipulator a Takagi Sugeno fuzzy model of thesystem is presented through a combination of linear systems The concept of Parallel DistributedCompensation PDC is used to design the fuzzy controller for the system To linearize the nonlinear system some points in the workspace of the manipulator are chosen and dynamical model is linearized at these points Stability of the designed fuzzy control system is guaranteed via Lyapunov approach The sufficientconditions for the existence of an appropriate controller are presented in terms of Linear Matrix Inequalities LMIs These LMIs are used to determine the common positive definite matrix and the feedback gains This manipulator has many singular points large degrees of freedom and a challenging dynamics tocontrol Therefore choosing the points in the workspace of the manipulator to design the fuzzy controller isvery important and optimization of the controller is necessary Therefore in this thesis to design the optimalPDC fuzzy controller genetic algorithm is used and the linearizing points are chosen optimally To reach thisend a fitness function is introduced with 6 points in the workspace of the manipulator as the inputs andintegral of trajectory tracking errors of the closed loop system as the output It is shown that the optimizedPDC controller with 6 points has a better performance compared with even a PDC controller with 10 points Due to deficiencies in the mechanical part of the system as well as model based nature of the controller practically implementation of the controller faced several problems However two sliding mode and PIDcontrollers are implemented in the experimental prototype Therefore implementation results of controllersare presented at the end Keywords Parallel robot Motion simulator Rotary actuator PDC Fuzzy control Genetic algorithm

عزيزان، حميد

طراحي كنترل كننده فازي بر مبناي الگوريتم ژنتيك براي ربات موازي شبيه سازهاي حركتي با محرك هاي دوراني

روش لاگرانژ , خطي سازي معادلات ديناميك