Abstract
As research into parallel mechanisms deepens, numerous specific configurations and modeling methods have emerged. However, many of these methods lack universality. This paper proposes a general modeling framework for both the kinematics and dynamics of "spoke inclined" parallel mechanisms. Using geometric methods and the Lagrange approach, the researchers established a mapping between platform pose and driving forces. A prototype with a 45° inclination angle was developed and tested using a dSPACE real-time simulation system. Experimental results validate the mathematical models, demonstrating that this class of mechanism offers low inertia and a large rotational range (up to 51.4°), making it highly suitable for industrial applications like automotive coating.
The Challenge of Specialized Parallel Robots
Parallel mechanisms, such as the Stewart platform, are widely used in high-precision robotics due to their stiffness and rapid response. Despite their advantages, most research focuses on modeling specific, fixed geometries. A significant gap exists in providing a "general" modeling method for mechanisms where the lead rails are inclined at an angle ($\theta$). Since the performance—such as workspace and payload capacity—varies significantly with this inclination, engineers require a universal mathematical toolset to predict behavior across a range of designs before physical prototyping.
A Dual-Layered Modeling Approach
The researchers developed a comprehensive framework to address both motion and force:
- Kinematic Modeling: By employing geometric methods and Roll-Pitch-Yaw (RPY) descriptions, the team established the relationship between the platform's position/orientation and motor displacements. They solved the inverse kinematics problem using quadratic equations, incorporating a selection criterion to identify the physically valid solution among multiple mathematical possibilities.
- Dynamic Modeling: To calculate the required motor forces during complex maneuvers, the authors used the Lagrange method. This energy-based approach is often more efficient than traditional Newtonian mechanics for complex robots. By treating connecting rods as particles, they simplified the dynamics of 18 different kinematic pairs into a solvable set of equations.
Experimental Validation and Results
To prove the model's efficacy, a physical prototype was constructed with a 45° inclination angle. The system was controlled via a dSPACE real-time simulation system, which utilized feedback from displacement and current sensors for precise closed-loop control.
The prototype was subjected to a six-dimensional trajectory task, following paths defined by various sine functions to test its full range of motion. Key findings include:
- High Fidelity: Actual sensor data closely matched the theoretical curves for both displacement and force, confirming the model's accuracy.
- Superior Range of Motion: The "spoke inclined" design achieved impressive rotational capabilities, reaching a maximum angle of 51.4° around the z-axis, significantly higher than many standard parallel configurations.
- Industrial Utility: The low inertia and high angular range make this mechanism ideal for the automotive industry, particularly for automated coating processes that require agile, complex movements.
Limitations and Future Directions
While the results were positive, small discrepancies between theory and experiment were noted. These were attributed to manufacturing tolerances in the prototype's components and minor sensor measurement errors. Future research will focus on refining the models to account for these physical imperfections and optimizing the inclination angle ($\theta$) for diverse industrial tasks beyond automotive manufacturing.
Summary updated