A Fractional Order Grey Prediction Algorithm

Abstract

In order to improve the prediction accuracy to uncertainty signal, the fractional order grey prediction algorithm is proposed in this paper, and the measurement method of the grey information of signals is proposed for the first time. The paper attempts to deduce the fractional order grey prediction algorithm through fractional order Taylor expansion formula and provide the measurement method of the grey information of signals by analyzing the relations between fractional-order differential orders and the prediction accuracy. Fractional order grey prediction algorithm spares the complex operation of whitening grey information and makes full use of the grey information. Experiments have proven that the prediction accuracy of fractional order grey prediction algorithm has been greatly increased as compared with GM (1,1) algorithm.