[1]毛文涛,高 祥,罗铁军,等.基于张量表示的间歇性序列自适应区间预测[J].郑州大学学报(工学版),2024,45(04):79-86.[doi:10.13705/ j.issn.1671-6833.2024.01.007]
 MAO Wentao,GAO Xiang,LUO Tiejun,et al.Adaptive Interval Prediction of Intermittent Series Based on Tensor Representation[J].Journal of Zhengzhou University (Engineering Science),2024,45(04):79-86.[doi:10.13705/ j.issn.1671-6833.2024.01.007]





Adaptive Interval Prediction of Intermittent Series Based on Tensor Representation
毛文涛12 高 祥1 罗铁军3 张艳娜12 宋钊瑜1
1.河南师范大学 计算机与信息工程学院,河南 新乡 453007;2.河南师范大学 智慧商务与物联网技术河南省工程实验室,河南 新乡 453007;3.株洲中车时代电气股份有限公司,湖南 株洲 412001
MAO Wentao12 GAO Xiang1 LUO Tiejun3 ZHANG Yanna12 SONG Zhaoyu1
1.College of Computer and Information Engineering, Henan Normal University, Xinxiang 453007, China; 2.Engineering Lab of Intelligence Business and Internet of Things of Henan Province,Henan Normal University, Xinxiang 453007, China; 3.Zhuzhou CRRC Times Electronic Co., Ltd., Zhuzhou 412001, China
需求预测 间歇性时间序列 张量分解 配件管理 区间预测 时间序列聚类
demand forecast intermittent time series tensor decomposition parts management interval predic tion time series clustering
TP301 F272.1
10.13705/ j.issn.1671-6833.2024.01.007
在实际业务中,配件需求发生随机、需求量波动大,配件序列数据呈现明显的间歇性分布,同时由于人工报 单失误或特殊事件等因素的影响,实际配件需求易发生异常变化,导致传统的时间序列预测方法难以捕捉配件需 求量的演化规律,预测结果不确定性高、可靠性不足。为解决上述问题,提出了一种基于张量表示的间歇性序列自 适应区间预测方法。首先,利用层次聚类,基于间歇性序列的平均需求间隔和平方变异系数指标筛选相似序列形 成序列簇,用于提取簇内公共需求演化信息,增加可预测性;其次,通过张量分解重构原始需求序列,在最大限度保 留序列核心信息的前提下平滑序列中的异常值;最后,构建一种自适应预测区间算法,通过动态更新机制得到配件 需求量的预测值和预测区间,以确保结果的可靠性。利用某大型轨道交通制造企业实际售后数据进行验证,与现 有典型时间序列预测方法相比,所提方法可有效挖掘不同特点间歇性序列的演化趋势,提高小样本间歇性序列的 预测精度。实验结果表明:所提方法在间歇性特有指标均方根标准误差(RMSSE)上,相较于需求预测主流的深度 学习方法平均降低了0.32,且当预测结果出现失真时,可提供一个可靠的弹性预测区间,为实际应用中企业智能备 件计划决策提供了一种新的解决方案。
In the actual business, parts demand occured randomly and demand fluctuates, so the demand sequence for spare parts showed obvious intermittent distribution. At the same time, due to factors such as manual reporting errors or special events, the actual demand for spare parts was prone to abnormal changes, making it difficult for traditional time series prediction methods to capture the evolution of the demand for accessories, resulting in high uncertainty and insufficient reliability of prediction results. To solve this problem, an adaptive interval prediction method for intermittent series based on tensor representation was proposed. Firstly, hierarchical clustering was used to screen similar sequences based on the average demand interval and square of the coefficient of variation of acces sory sequences, forming sequence clusters to increase predictability. Secondly, the original demand sequence was reconstructed by tensor decomposition. The outliers in the sequence were then corrected while retaining the core in formation of the original sequence to maximum extent. Finally, an adaptive prediction interval algorithm was con structed, which could obtain the predicted value and prediction interval of the parts demand through the dynamic update mechanism to ensure the reliability of the results. The proposed method was validated on the aftersales data set from a large vehicle manufacturing enterprise. Compared with existing time series prediction methods, the pro posed method could effectively extract the evolutionary trend of various types of intermittent series and improve the prediction accuracy on the intermittent time series with small size as well. Experiments showed that the average root mean square scaled error (RMSSE) of this method was 0.32 lower than that of the mainstream in-depth learning method of demand prediction. More importantly, when the prediction results were distorted, the proposed method could provide a reliable and flexible prediction interval, which could be helpful to provide a feasible solution for in telligent parts management.


[1] BAO Y K, WANG W, ZOU H. SVR-based method forecasting intermittent demand for service parts invento ries[C]∥ International Workshop on Rough Sets,Fuzzy Sets,Data Mining,and Granular Computing. Cham: Springer,2005:604-613. 

[2] FU G Q, ZHENG Y, ZHOU L F, et al. Look-ahead prediction of spindle thermal errors with on-machine measurement and the cubic exponential smoothing unscented Kalman filtering-based temperature prediction model of the machine tools[J]. Measurement, 2023, 210: 112536. 
[3] 贾茹宾, 高金峰. 基于ARIMA模型的变压器油中溶 解气体含量时间序列预测方法[J]. 郑州大学学报 (工学版), 2020, 41(2): 67-72. 
JIA R B, GAO J F. Time series prediction method of dissolved gas content in transformer oil based on ARIMA model[J]. Journal of Zhengzhou University (Engineer ing Science), 2020, 41(2): 67-72. 
[4] KARMY J P, MALDONADO S. Hierarchical time series forecasting via support vector regression in the European travel retail industry[J]. Expert Systems with Applica tions, 2019, 137: 59-73. 
[5] VAN STEENBERGEN R M, MES M R K. Forecasting demand profiles of new products[J]. Decision Support Systems, 2020, 139: 113401. 
[6] SUENAGA D, TAKASE Y, ABE T, et al. Prediction accuracy of Random Forest, XGBoost, LightGBM, and artificial neural network for shear resistance of post installed anchors[J].Structures,2023,50:1252-1263. 
[7] CAO D D, CHAN M, NG S. Modeling and forecasting of nanoFeCu treated sewage quality using recurrent neu ral network (RNN)[J]. Computation, 2023, 11(2): 39. 
[8] ABBASIMEHR H, SHABANI M, YOUSEFI M. An op timized model using LSTM network for demand forecas ting[J]. Computers & Industrial Engineering, 2020, 143: 106435. 
[9] CROSTON J D. Forecasting and stock control for inter mittent demands[J]. Journal of the Operational Re search Society, 1972, 23(3): 289-303. 
[10] SYNTETOS A A, BOYLAN J E. The accuracy of inter mittent demand estimates[J]. International Journal of Forecasting, 2005, 21(2): 303-314. 
[11] GUTIERREZ R S, SOLIS A O, MUKHOPADHYAY S. Lumpy demand forecasting using neural networks[J]. In ternational Journal of Production Economics, 2008, 111 (2): 409-420. 
[12]张瑞. 不常用备件需求预测模型与方法研究[D]. 武 汉: 华中科技大学, 2011. 
ZHANG R. Research on forecasting models and methods of rarely used spare parts′ demand[D]. Wuhan: Hua zhong University of Science and Technology, 2011. 
[13] SHI Q Q, YIN J M, CAI J J, et al. Block Hankel tensor ARIMA for multiple short time series forecasting[J]. Proceedings of the AAAI Conference on Artificial Intelli gence, 2020, 34(4): 5758-5766. 
[14] BOUKHTOUTA A, JENTSCH P. Support vector machine for demand forecasting of Canadian armed forces spare parts[C]∥2018 6th International Symposium on Compu tational and Business Intelligence (ISCBI). Piscataway: IEEE, 2018: 59-64. 
[15] YOKOTA T, EREM B, GULER S, et al. Missing slice recovery for tensors using a low-rank model in embedded space[C]∥2018 IEEE/CVF Conference on Computer Vi sion and Pattern Recognition. Piscataway: IEEE, 2018: 8251-8259. 
[16]周晓艳, 唐涛, 张思乾, 等. 多角度SAR图像非目标 遮挡缺失信息重构[J]. 信号处理, 2021, 37(9): 1569-1580. 
ZHOU X Y, TANG T, ZHANG S Q, et al. Missing information reconstruction for multi-aspect SAR image occlusion[J]. Journal of Signal Processing, 2021, 37 (9): 1569-1580. 
[17] CHENG J, DEKKERS J C M, FERNANDO R L. Cross validation of best linear unbiased predictions of breeding values using an efficient leave-one-out strategy[J]. Jour nal of Animal Science, 2020, 98(S4): 10-11.

更新日期/Last Update: 2024-06-14