The Application of Combination Forecasting Method in Total Power of Agriculture Machinery Based on RS

Abstract

Forecast of agriculture machinery total power is a complicated non-linear system，combination forecasting can take full advantage of known information, to improve prediction accuracy .The relative data model between forecast objective and forecast model, and knowledge system and decision table was established respectively by means of converting continuous attribute values into discrete attribute values. Then, the weight of combination forecast model was calculated according to estimating dependence and significance of attributes in rough set theory, from this ,then constructed the combination forecasting model and conducted combination forecasting agriculture machinery total power in Heilongjiang province. The results showed that the forecast average error of combination forecast model is 3.13,which is lower than 3.71、5.25 and 3.63 of quadratic curve model, GM(1,1) and cubic exponent smooth model , and is also lower than 3.34 and 3.26 of the combinatorial forecast model based on the divergence coefficient method and Shapley value method.