A soft technique for measuring friction force using neural network

Abstract

There are two approaches to measure a friction force: force sensor, software estimation algorithm. This paper will focus on software approach to measure friction. The proposed approach uses a neural network (NN) to approximate the friction force in a mechanical system. Since the friction force considered is a speed-dependent function, a learning algorithm is adopted to update the NN weights so as to follow unknown friction behaviors. The advantage of the proposed friction estimation method is that it is based on the built NN model, and it does not require the force sensor measurement. Simulation test is given to verify the effectiveness of the proposed approach. © 2011 IFSA.