Solving Forward Kinematics Problem of Stewart Robot Using Soft Computing

in this paper, we consider the problem of efficient computation of the forward kinematics of a 6 DOF robot manipulator built to use in rehabilitation purpose. Forward kinematics problem (FKP) of parallel robots is very difficult to solve in comparison to the serial manipulators. This problem is almost impossible to solve analytically. Numerical methods are one of the common solutions for this problem. But, accuracy, speed and convergence of these methods are fully dependent on the initial guess vector that is fed to the numerical algorithm. In this paper, soft computing approach like Artificial Neural Networks, fuzzy-neural network and nonlinear Auto Regressive eXogenous (NARX) identification method is used to solve the FKP of the Stewart robot. This problem is solved in the typical workspace of this robot. The results show the advantages of Nonlinear ARX identification method in providing very small modeling errors and provide excellent position and orientation angle estimation. Keywords— Stewart robot, Artificial Neural Networks, ANFIS, Nonlinear ARX, Identification.