Survey paper on the Forward Kinematics Solution of a Stewart Platform

Abstract

Over the past few decades, researchers have become increasingly interested in the Stewart Platform, a parallel manipulator with six degrees of freedom (DOF) that offers remarkable functionality over serial manipulators. The main objection to the implementation of closed loop control in Stewart Platform is its difficulty in solving the forward kinematics. In the forward kinematics problem, the pose and orientation of the moving platform are determined from the leg lengths. This is a high degree nonlinear problem with multiple solutions. Even though many studies were conducted about the kinematics of parallel robots, forward kinematics continues to be a hindrance to its practical application. Consequently, the real-time solution of forward kinematics is gaining interest. To address the direct kinematics problem, several approaches were presented, including numerical techniques, analytical methods, sensor-based methods, methods using neural networks, and observer-based methods. In this paper, the different approaches and limitations of the forward kinematics problem on Stewart Platform are analyzed.