物流网络中多地区配送需求预测方法

摘要

物流网络较为复杂,当前预测方法只采用一种方法对物流网络中多地区配送需求进行预测,无法保证预测精度.提出一种新的需求预测方法, 通过两个或两个以上不一样的预测方法,对物流网络中多地区配送需求进行预测,构建物流网络中多地区配送需求组合预测模型,对不同方法进行加权组合,通过熵值法对组合预测模型的组合权重进行计算,充分利用不同预测方法的优点,获取较优的预测结果.通过残差检验和后验差检验,对组合预测模型的配送需求预测能力进行评价.实验结果表明,所提方法具有很高的预测精度.%Logistics network is relatively complex, the current forecasting methods use only one way to predict the demand for logistics network in many regions distribution, the forecast accuracy there is no guarantee.A new demand forecasting method was put forwarded, by two or more than two different prediction methods, to forecast the logistics network in the region distribution more demand, logistics network combination forecast model was built, many regional distribution requirements in weighted combination, the different methods of combination forecast model by entropy value method of combination weights to calculate, full use of the merits of different prediction methods were made, to obtain a better prediction results.Through the residual test and posterior difference test, the combination forecast model to evaluate the distribution of demand forecasting ability.The experimental results show that the proposed method has high prediction accuracy.