Artificial Neural Network-based error compensation procedure for low-cost encoders - Computer Science > Neural and Evolutionary ComputingReport as inadecuate




Artificial Neural Network-based error compensation procedure for low-cost encoders - Computer Science > Neural and Evolutionary Computing - Download this document for free, or read online. Document in PDF available to download.

Abstract: An Artificial Neural Network-based error compensation method is proposed forimproving the accuracy of resolver-based 16-bit encoders by compensating fortheir respective systematic error profiles. The error compensation procedure,for a particular encoder, involves obtaining its error profile by calibratingit on a precision rotary table, training the neural network by using a part ofthis data and then determining the corrected encoder angle by subtracting theANN-predicted error from the measured value of the encoder angle. Since it isnot guaranteed that all the resolvers will have exactly similar error profilesbecause of the inherent differences in their construction on a micro scale, theANN has been trained on one error profile at a time and the correspondingweight file is then used only for compensating the systematic error of thisparticular encoder. The systematic nature of the error profile for each of theencoders has also been validated by repeated calibration of the encoders over aperiod of time and it was found that the error profiles of a particular encoderrecorded at different epochs show near reproducible behavior. The ANN-basederror compensation procedure has been implemented for 4 encoders by trainingthe ANN with their respective error profiles and the results indicate that theaccuracy of encoders can be improved by nearly an order of magnitude fromquoted values of ~6 arc-min to ~0.65 arc-min when their correspondingANN-generated weight files are used for determining the corrected encoderangle.



Author: V.K.Dhar, A.K.Tickoo, S.K.Kaul, R.Koul, B.P.Dubey

Source: https://arxiv.org/







Related documents