Forward kinematics analysis of a six‐DOF Stewart platform using PCA and NM algorithm

The purpose of this paper is to present an adaptive numerical algorithm for forward kinematics analysis of general Stewart platform.

Unlike the convention of developing a set of kinematic equations and then solving them, an alternative numerical algorithm is proposed in which the principal components of link lengths are used as a bridge to analyze the forward kinematics of a Stewart platform. The values of link lengths are firstly transformed to the values of principal components through principal component analysis. Then, the computation of the values of positional variables is transformed to a two‐dimensional nonlinear minimization problem by using the relationships between principal components and positional variables. A hybrid Nelder Mead‐particle swarm optimizer (NM‐PSO) algorithm and a modified NM algorithm are used to solve the two‐dimensional nonlinear minimization problem.

Simulation experiments have been conducted to validate the numerical algorithm and experimental results show that the numerical algorithm is valid and can achieve good accuracy and high efficiency.

This paper proposes an adaptive numerical algorithm for forward kinematics analysis of general Stewart platform.