铁路货车再制造领域零件换新率动态预测方法  

Dynamic Prediction Method for Part Replacement Rate in the Freight Train Remanufacturing Field

作　　者：孙超凡 乔运华 何山 赵鸿浩 SUN Chao-fan;QIAO Yun-hua;HE Shan;ZHAO Hong-hao(Beijing Research Institute of Automation for Machinery Industry Co.,Ltd.,Beijing 100120)

机构地区：[1]北京机械工业自动化研究所有限公司,北京100120

出　　处：《制造业自动化》2025年第3期94-99,共6页Manufacturing Automation

摘　　要：铁路货车零部件的换新率是货车再制造企业制定预投预买计划的重要数据来源,预测出较为准确的换新率可以帮助企业缩短交付时间和降低库存。目前企业在估算换新率时未能充分考虑到各维度因素对换新率的影响,这使得企业不能有效制定准确的预投预买计划,无法确保缩短交期和降低库存成本的要求。分析了实际业务场景,针对这一业务特点和需求建立了动态模型,并提出了一种铁路货车再制造领域零件换新率动态预测方法。此方法结合影响铁路货车换新率的多维度因素,运用机器学习算法中的支持向量机回归模型构建了基于动态模型的零件换新率预测模型。以实际铁路货车转向架零部件历史换新数据为基础进行验证,实验结果显示,该模型在测试集上的绝对误差低于15%,验证了模型较高的预测准确性,表明其可有效应用于实际场景。此方法为企业在再制造过程中进行预投预买决策与库存控制提供了数据支持,有助于企业降低成本及提高竞争力。The replacement rate of freight train parts is an important data source for freight train remanufacturing enterprises to formulate pre-investment and pre-purchase plans,while the prediction of a more accurate replacement can help enterprises shorten delivery time and reduce inventory.At present,enterprises that fail to fully consider the impact of various factors on the replacement rate during estimation find it difficult to effectively formulate accurate pre-investment and pre-purchase plans,thereby shortening the delivery time and reducing the inventory costs.This paper analyzes the actual business scenarios,and proposes a dynamic model-based method for predicting the replacement rate of freight train parts.This method combines the multi-dimensional factors affecting the replacement rate of freight trains,and uses the support vector machine regression model to construct a dynamic model of the parts replacement rate.By validating the model against historical replacement data of freight train bogie components,the experimental results demonstrate that the model achieves a mean absolute error of less than 15%on the testing set,confirming its effectiveness and applicability in real-world scenarios.This method provides data support for enterprises'pre-investment and pre-purchase decision-making as well as inventory control in the remanufacturing process,which helps enterprises to reduce costs and improve competitiveness.

关 键 词：动态模型 铁路货车零件换新率 支持向量机回归 ERP 

分 类 号：U279[机械工程—车辆工程]