数据驱动的公路货运合同物流投标报价模型

Study on Data Driven Logistics Bidding Model for Highway Freight Transport Contracts

提出了一种基于数据驱动的公路货运合同物流的投标报价模型,该模型综合了国内生产总值、燃油价格、市场需求、货物重量、运输距离、承运商、货车长度七个因素,从而提高了第三方物流企业的投标预测水平。针对运输成本的非线性特征,利用BP神经网络模型,通过案例分析获得的实证数据进行训练,对短期报价进行预测。另外,利用遗传算法提高搜索性能。通过与多元线性回归算法和BP算法进行比较,提出的混合GA-BP算法具有较佳的准确度和性能。该研究为第三方物流公司带来的启示在于,历史数据在公路货运合同物流的招标报价决策中是有价值和有效的。

This paper presents a data-driven logistics service bidding model for highway freight transport contracts, which comprehensively considers the seven factors, namely GDP, fuel price, market demand, cargo weight, shipping distance, carrier and truck length, for the purpose of enhancing the bidding forecasting ability of the third-party logistics companies. Then according to the non-linear characteristics of the transportation cost, the BP neural network model is used and trained based on empirical data obtained through a case analysis to forecast the short-term quotation of the logistics services. In addition, the genetic algorithm is used to improve search performance of the model. At the end, by comparing the multivariate linear regression algorithm and the BP algorithm, we conclude that the hybrid GA-BP algorithm proposed in this paper has better accuracy and performance.