Fuzzy Logic System for Position Control and Current Stabilization of a Robot Manipulator

This paper proposes the usage of the fuzzy logic for both position control and stabilization of current pulsations. The aim is to ensure precise control of the joint motion position with very quickly joints movements. The block diagram of a control system involves controller, actuator, robot and appropriate sensors. Each of these components is described. The dynamic model of a manipulator is based on the Newton-Euler formulation that provides a description of the relationship between the joint actuator torques and the motion of the structure. The fuzzy logic controller Mamdani type contains two control inputs: position error and armature current. Input variable fuzzification, inference based on fuzzy rules and output set defuzzification are described. The same fuzzy control algorithm is used in all joint servo loops, requiring only multiplexing and proper scaling of the fuzzy controller inputs and outputs. The electric dc (direct-drive) servomotors were used for actuating the joints of a manipulator. The influence of nonlinear loads, gravitation-dependent load, viscous friction torque, Coulomb's friction and torque due by reducer on position tracking performance was also considered