本文選題：物流配送 + 操作成本　； 參考：《物流技術(shù)》2014年01期

【摘要】：為了有效地利用物流成本估計中線性和非線性數(shù)學模型的優(yōu)點,把線性預測性能優(yōu)異的ARIMA數(shù)學模型和RBF神經(jīng)網(wǎng)絡相結(jié)合,使模型非線性數(shù)學變化上形成估計優(yōu)化,可以捕捉物流成本價格的線性和非線性規(guī)律,有效地減少傳統(tǒng)預測數(shù)學模型中一些非線性因素的影響。以某物流公司1991～2012年物流操作成本為數(shù)據(jù),將所提出的數(shù)學模型與網(wǎng)格搜索SVR模型、PSO-SVR模型、Levenberg-Marquardt BP神經(jīng)網(wǎng)絡模型及背景值優(yōu)化GM(1,1)模型進行對比實驗。結(jié)果表明所提出的優(yōu)化數(shù)學模型能夠解決上述問題且具有更高的預測精度。
[Abstract]:In order to effectively utilize the advantages of linear and nonlinear mathematical models in logistics cost estimation, Arima mathematical model with excellent linear predictive performance is combined with RBF neural network to optimize the estimation of nonlinear mathematical variation of the model. It can capture the linear and nonlinear laws of logistics cost and price, and effectively reduce the influence of some nonlinear factors in the traditional predictive mathematical model. Taking the logistics operation cost of a logistics company from 1991 to 2012 as the data, this paper compares the proposed mathematical model with the grid search SVR model and the Levenberg-Marquardt BP neural network model and the background value optimization GM (1t1) model. The results show that the proposed optimal mathematical model can solve the above problems and has higher prediction accuracy.
【作者單位】： 邢臺職業(yè)技術(shù)學院;
【基金】：河北省社會科學基金項目“基于數(shù)學模型的物流配送成本優(yōu)化體系構(gòu)建”(HB12YJ006)
【分類號】：F253.7;F224

【參考文獻】

相關(guān)期刊論文 前2條

1 郝洪;王s，

本文編號：2083343