摘要
基于大数据环境提出考虑误差修正的两阶段船舶中间产品装配工时预测模型。从船舶设计软件中提取中间产品装配工艺信息,建立反向传播神经网络(Back Propagation Neural Network,BPNN)模型,实现装配工时的初步预测。采集对装配工时预测可能造成影响的外界因素大数据,建立基于极端梯度提升(Extreme Gradient Boosting,XGBoost)算法的装配工时预测误差修正模型。两阶段预测结果相加得到装配工时预测值。实例验证该预测模型的有效性,可为船舶企业完善装配工时管理提供切实可行的解决思路。
Based on the big data environment,a two-stage prediction model with error correction for assembly working hours of ship intermediate products is proposed.The assembly process information of intermediate products is extracted from the ship design software,and the Back Propagation Neural Network(BPNN)model is established to realize the preliminary prediction of assembly working hours.The big data of external factors that may affect the prediction of assembly working hours are collected,and an error correction model for the prediction of assembly working hours based on Extreme Gradient Boosting(XGBoost)algorithm is established.The predicted value of assembly working hours is obtained by adding the prediction results of the two stages.The validity of the prediction model is verified by an example,which can provide feasible solution for shipbuilding enterprises to improve the management of assembly working hours.
作者
苏翔
徐瑞林
杨玉雪
史恭波
SU Xiang;XU Ruilin;YANG Yuxue;SHI Gongbo(School of Economics and Management,Jiangsu University of Science and Technology,Zhenjiang 212100,Jiangsu,China;Zhenjiang Jinzhou Software Co.,Ltd.,Zhenjiang 212003,Jiangsu,China)
基金
江苏省社会科学基金项目“江苏区域绿色经济效率评价、影响因素及提升策略研究”(编号:22EYD004)。
关键词
船舶
中间产品
装配工时
预测模型
误差修正
BPNN
XGBoost
ship
intermediate product
assembly working hour
prediction model
error correction
BPNN(Back Propagation Neural Network)
XGBoost(Extreme Gradient Boosting)
作者简介
苏翔(1965-),男,教授,主要从事管理信息系统、成本控制、智能制造等研究。
