GM(1,N) Model Multi-variable Forecast

摘要

In light of mutual connections and restriction relations between independent variables, the primitive series of gray system was transformed into the AGO series. The optimal one-step transfer matrix P was solved from k-1 time to k time using Markov chain and lingo software, meeting recurrence formula x~(1)(k)=P x~(1)(k-1) by the least-squares method. Then the prediction was carried on with this transfer matrix. Finally, the AGO series were returned to the IAGO series. Through the example analysis, it is proved that the new model has a high simulation precision and forecast precision through the example analysis.