Dual-arm robotic manipulation represents a promising solution for the automation of complex tasks that require coordination, precision, and controlled interaction with the environment. One of the most relevant applications is automated welding, where it is necessary to follow constrained trajectories while constantly maintaining controlled contact with the workpiece. In this thesis, a dual-arm system is developed and simulated to mimic automated welding operations through precise trajectory tracking. The system is modeled in MATLAB, integrating CasADi for symbolic and numerical optimization and the Robotics System Toolbox for robot modeling and simulation. Reference trajectories are generated during the path planning phase using the OrientationConstrained Rapidly exploring Random Tree (OC-RRT) algorithm [1], which enables the creation of kinematically feasible and optimized paths. Trajectory execution is handled by a 12- degree-of-freedom Nonlinear Model Predictive Controller (NMPC), designed to ensure accuracy and stability even under non-ideal conditions. The system is tested in simulation under dynamic disturbances, sensor noise, and in addition with a distance constraint between the end-effectors of the two manipulators. A comprehensive description of the system architecture, planning strategies, and control techniques is provided, along with a performance analysis in both nominal and perturbed scenarios. The results highlight the robustness and effectiveness of the proposed approach, paving the way for potential future implementations on real robotic manipulators in advanced industrial environments.

Incoming