Modeling & Intelligent Control of a Multi-DoF Robotic Arm

Loading...
Thumbnail Image

Date

Journal Title

Journal ISSN

Volume Title

Publisher

University of Amar Telidji - Laghouat Faculty of Technology Department of Electrical Engineering

Abstract

This dissertation addresses the dynamic modeling and high-speed trajectory tracking control of a 6-DOF PUMA 560 robotic manipulator under severe dynamic stress and parametric uncertainty. Classical PID controllers, being structurally unaware of nonlinear multi-body dynamics, exhibit serious tracking distortion (16.8 mm error) and actuator saturation under an unmodeled 5 kg payload. To overcome these limitations, two opposing strategies are investigated: Model-free Mamdani Fuzzy Logic Control (FLC) offers robustness but lacks deterministic precision, while model-based Computed Torque Control (CTC) delivers perfect nominal precision (0.05 mm error) but fails catastrophically under parametric variations. The main contribution is a hybrid CTC-FLC architecture that synergistically integrates exact nonlinear cancellation from the dynamic model with intelligent uncertainty rejection from the fuzzy compensator. Under sustained high-frequency helical maneuvers and an unmodeled 5 kg payload, the hybrid controller entirely eliminates steady-state spatial shift and actuator chattering, achieving a remarkable mean Cartesian error of only 0.1 mm. This work provides empirical evidence that absolute spatial accuracy in unpredictable industrial environments requires the synergistic fusion of deterministic physical modeling and artificial intelligence.

Description

Citation

Endorsement

Review

Supplemented By

Referenced By