Error compensation of optical encoder based on RBF network

Abstract

A new method to correct and compensate the dynamic error of a optical encoder was presented by using the neural network and the digital signal process technology.A modeling method based on the Radial Basis Function(RBF) was set up,in which the output was the test value of high precision instrument and the input was the angle values of sample points.According to the inhibiting condition between the test value and the output of network,the power factor formula,the center and width of the RBF were adjusted to make the model have a good learning ability and a generalization ability.The relationship between sampled angles and errors was determined by training the neural network.After the sampled angles were measured,the unknown error of the encoder could be calculated via a trained neural network even if the error was nonlinear.The tested results show that the precision of system error by this method is 1~3 times that by tradiotional method.And not only the error of each network point but also the interpolated network point can be corrected automatically.The practical application proves that the precision of measuring system is improved greatly by using the RBF model as error compensation,and the effect of nonlinear errors on the system is reduced also.