Dynamic Modeling, Control, and Real-Time Implementation of a Planar Five-Bar Pantograph Robot
Description
This paper presents a complete dynamic modeling, and real-time implementation framework for a planar five-bar pantograph robot. Starting from a constrained Lagrangian formulation in redundant coordinates, the closed-chain dynamics are systematically reduced through the constraint null space to obtain a minimal two-coordinate model in the actuated joint variables. The resulting equations preserve the standard manipulator structure, with a symmetric positive-definite mass matrix, Coriolis and gravity terms, and an end-effector wrench mapping through the Jacobian transpose. The formulation covers both horizontal and vertical mounting configurations, allowing gravity effects to be either neglected or fully included depending on the application. Building on this model, several model-based control strategies are developed, including gravity compensation, computed-torque tracking, operational-space interaction control, hybrid force/position control, and Cartesian impedance control. A discrete-time implementation layer is then introduced through efficient forward- and inverse-dynamics algorithms suitable for sampled real-time control loops. Numerical studies verify the proposed framework through computed-torque trajectory tracking, energy-conserving forward simulation, inverse/forward-dynamics consistency, quasi-static loading comparisons, force regulation against a compliant environment, and impedance rendering under external disturbances. The results demonstrate that the proposed reduced model is physically consistent, computationally efficient, and suitable for real-time interaction-control applications such as robotic rehabilitation.