Approximation of Coefficients Influencing Robot Design Using FFNN with Bayesian Regularized LMBPA
Raja Mohamed S, P. Raviraj · Procedia Engineering · 2012
We propose a method to calculate the coefficients that will be influencing the design of a Robot based on a mathematical equation in a discrete time model using a (FFNN) feed forward neural network as it is a really challenging task to design them for any of the industrial applications especially in medical field. The Proposed mathematical model derived from Lagrange-Euler equations to calculate the coefficients of the Robot with help of neural networks and also guides us to generate control signals from it. Here PUMA 560 robot arm has been taken in to consideration.