考虑样本特征的第三方物流安全库存预测模型

A Forecast Method of Safety Stock of Third Party Logistics Considering Sample Characteristics

安全库存的准确预测和设置是第三方物流提高客户服务水平、控制自身成本、优化供应链整体性能的重要问题。针对汽车制造业中第三方物流安全库存预测中存在的数据样本有限、线性相关度低、维数高等问题,运用支持向量机算法建立第三方物流安全库存预测模型,并将预测过程中得出的所有支持向量进行分析得到安全库存预测模型中输入变量对输出结果的影响的特征变量权重计算公式。最后运用某品牌汽车第三方物流中心各参量历史数据进行实例应用与验证,证实了该预测模型的良好准确性和鲁棒性。

The accurate forecasting and setting of the safety stock is the important problem of the third party logistics to improve the customer service level,self-cost control and optimize the overall performance of the supply chain.Aiming at the characteristics of less data,low linear correlation and high latitude in the third party logistics inventory forecasting,using support vector machine method to establish safety stock of third party logistics prediction model.Through the support vector analysis,the calculation formula of the characteristic variable weight in the third party logistics safety inventory prediction is obtained.The accuracy and robustness of the model are verified by the historical dataof athird party automobile logistics center.