Forward Kinematics Analysis of a 3-PRR Planer Parallel Robot Using a Combined Method Based on the Neural Network

Abstract

Forward kinematics problem has always been a controversial topic in the case of parallel robots because of the close loop structure and the complicated nonlinear system of equations. In this paper, the forward kinematics problem of a planar parallel robot in its workspace is considered and a neural network based solution is proposed. In this approach, the robot's workspace is segmented into several subspaces for improving the accuracy of the solutions. Then, the pose of the parallel robot is precisely estimated using a neural network for each subspace. In order to determine the validity of this approach, the circular motion of a 3-PRR planar parallel robot is simulated by applying the proposed method. The simulation results indicate the advantages of this approach compared to other methods.