Solving Kinematics of a Parallel Manipulator Using Artificial Neural Networks

Abstract

Artificial Neural Networks (ANNs) are known for their ability to map nonlinear relations between inputs and outputs. This paper presents ANN-based kinematic modeling of a recently developed parallel manipulator. The manipulator has 3 limbs of prismatic-universal-universal (3-PUU) structure. To avoid the computational complexity of solving the kinematics problem in real-time application, two artificial neural networks are trained to estimate the forward and inverse kinematics solutions. Simulation results show that the developed ANNs have great prediction capabilities, providing accurate kinematic solution and can then be applied in the control design of the proposed manipulator.