基于改进的GM(1,1)模型的公路里程预测

Highway Mileage Forecasting Based on Improved GM(1,1) Model

分析了传统GM(1,1)模型在公路网里程预测中存在的问题,利用残差改进了GM(1,1)模型,运用马尔可夫状态转移矩阵判断残差预测值在k≥n时的符号,提高了预测精度。对广西省公路网里程进行了预测分析,结果表明该法具有较高的实用性。

Problems in highway network mileage forecasting using the traditional were analyzed. An improved model GM （1,1） using residual was provided. By using Markov State Transferring Matrix to determine the forecasting residual value, forecast accuracy was raised. Guangxi Province road network was taken as an example, the mileage was forecasted. The results showed that the method had a high practicality.