Design of a Gough-Stewart Platform Based on Visual Servoing Controller

Area of Science:

• Robotics
• Control Systems Engineering
• Mechanical Design

Background:

• Achieving high accuracy in robotic systems is a key challenge.
• Gough-Stewart platforms require precise design for optimal controller performance.
• Existing design methods often overlook the interplay between robot parameters and controller accuracy.

Purpose of the Study:

• To present a novel control-based design methodology for Gough-Stewart platforms.
• To optimize geometric parameters and camera placement for enhanced robot accuracy.
• To evaluate the effectiveness of the methodology with different visual servoing controllers.

Main Methods:

• Application of three visual servoing controllers: leg-direction-based, line-based, and image moment visual servoing.
• Analysis of positioning error models, incorporating camera observation errors and controller singularities.
• Formulation of optimization problems to determine optimal robot geometric parameters and camera placement.
• Co-simulation of optimized platforms to assess positioning accuracy and robustness against manufacturing errors.

Main Results:

• The control-based design methodology successfully optimizes geometric parameters and camera placement.
• Co-simulations demonstrate improved positioning accuracy and robustness for the designed Gough-Stewart platforms.
• The methodology effectively integrates robot design with dedicated controller performance.

Conclusions:

• The proposed control-based design methodology offers a superior approach to designing accurate Gough-Stewart platforms.
• This method ensures optimal robot design parameters and enhanced controller performance.
• The findings contribute to the development of more precise and reliable robotic systems.