基于GM(1,1)-ISSA-SVM的车轮廓形磨耗演变预测方法

doi:10.3901/JME.260719

摘要/Abstract

摘要： 针对传统机器学习方法在小样本条件下难以准确预测车轮磨耗演变的问题，提出一种非等间距灰色预测模型GM(1,1)与改进麻雀搜索算法(Improved sparrow search algorithm，ISSA)优化支持向量机(Support vector machine，SVM)相结合的车轮廓形磨耗预测方法(GM(1,1)-ISSA-SVM)。首先，采用基于背景值优化的改进GM(1,1)模型对车轮型值点磨耗序列进行初步预测；随后，提出一种改进的麻雀搜索算法，对SVM的惩罚系数与核参数进行自动寻优，利用优化SVM模型对车轮型值点初始预测序列进行残差修正，并结合考虑最优权重因子组合的NURBS曲线拟合方法获得最终的车轮廓形磨耗预测结果。最后，利用某地铁线路车轮廓形磨耗跟踪数据对所提模型的有效性和优越性进行验证。试验结果表明，ISSA有效提升了传统SSA的全局寻优能力，在此基础上构建的GM(1,1)-ISSA-SVM模型可实现对小样本车轮磨耗数据集的高精度拟合与预测。与单一GM(1,1)模型、LSTM模型及GM(1,1)-SVM组合模型相比，GM(1,1)-ISSA-SVM模型在各项性能指标上均表现出更优的预测能力，车轮型值点的预测结果与实测值之间的平均绝对误差和相对误差仅为0.005 mm和0.5%。此外，最终预测得到的车轮型面与实测廓形之间的法向误差不超过0.05 mm，且轮轨接触特性与实测型面高度吻合。研究成果有助于推动车轮运维策略由计划修向状态修转变。

关键词: 车轮磨耗预测, 数据驱动, GM(1,1)模型, 麻雀搜索算法, 支持向量机

Abstract: To address the challenge that traditional machine learning methods face in accurately predicting wheel profile wear under small-sample conditions, a wheel profile wear prediction method (GM(1,1)-ISSA-SVM) is proposed. This method integrates a non-equidistant grey prediction model GM(1,1) and a support vector machine(SVM) optimized by an improved sparrow search algorithm(ISSA). First, an enhanced GM(1,1) model with optimized background values is used to generate an initial prediction of the wear sequence at wheel profile points. Next, an improved sparrow search algorithm is proposed to automatically optimize the penalty coefficient and kernel parameter of the SVM model. The optimized SVM is then applied to perform residual correction on the initially predicted wear sequence. The corrected results are further integrated with a NURBS curve fitting approach, which uses an optimal weighting factor combination, to obtain the final prediction of wheel profile wear. Finally, the effectiveness and superiority of the proposed model are validated using wheel profile wear tracking data collected from a specific metro line. Experimental results show that the proposed ISSA significantly enhances the global optimization capability of the original SSA. With this improvement, the developed GM(1,1)-ISSA-SVM model achieves high-precision fitting and prediction on small-sample wheel wear datasets. Compared with the standalone GM(1,1) model, the LSTM model, and the GM(1,1)-SVM combined model, the GM(1,1)-ISSA-SVM model demonstrates superior predictive performance across all evaluation metrics. The mean absolute and relative errors between the predicted and measured wheel profile data are only 0.005 mm and 0.5%, respectively. Furthermore, the normal error between the predicted and measured wheel profiles does not exceed 0.05 mm, and the wheel-rail contact characteristics closely match those derived from the measured profile. These results could help facilitate the transition of wheel maintenance strategies from scheduled maintenance to condition-based maintenance.

Key words: wheel wear prediction, data driven, GM (1,1) model, sparrow search algorithm, support vector machine

中图分类号:

U211

温炳光, 李文杰, 陶功权, 温泽峰. 基于GM(1,1)-ISSA-SVM的车轮廓形磨耗演变预测方法[J]. 机械工程学报, 2026, 62(16): 369-382.

WEN Bingguang, LI Wenjie, TAO Gongquan, WEN Zefeng. Wheel Profile Wear Evolution Prediction Method Based on GM(1,1)-ISSA-SVM[J]. Journal of Mechanical Engineering, 2026, 62(16): 369-382.

导出引用管理器 EndNote|Reference Manager|ProCite|BibTeX|RefWorks

链接本文: http://www.cjmenet.com.cn/CN/10.3901/JME.260719

http://www.cjmenet.com.cn/CN/Y2026/V62/I16/369

参考文献

[1] 李霞，温泽峰，金学松. 地铁车轮踏面异常磨耗原因分析[J]. 机械工程学报，2010，46(16)：60-66. LI Xia，WEN Zefeng，JIN Xuesong. Analysis of abnormal wear on metro wheel tread[J]. Journal of Mechanical Engineering，2010，46(16)：60-66.
[2] TAO G，TAN M，YUAN X，et al. Influence of wheel wear on the rolling contact fatigue of metro wheels in the central area of the wheel tread[J]. Wear，2025，571：205777.
[3] 金学松，赵国堂，梁树林，等. 高速铁路轮轨磨损特征、机理、影响和对策—车轮踏面横向磨耗[J]. 机械工程学报，2018，54(4)：3-13. JIN Xuesong，ZHAO Guotang，LIANG Shulin，et al. Characteristics，mechanisms，influences and counter measures of highspeed wheel/rail wear：Transverse wear of wheel tread[J]. Journal of Mechanical Engineering，2018，54(4)：3-13.
[4] ZOBORY I. Prediction of wheel/rail profile wear[J]. Vehicle System Dynamics，1997，28(2-3)：221-259.
[5] JENDEL T. Prediction of wheel profile wear-comparisons with field measurements[J]. Wear，2002，253(1-2)：89-99.
[6] BRAGHIN F，BRUNI S，RESTA F. Wear of railway wheel profiles：A comparison between experimental results and a mathematical model[J]. Vehicle System Dynamics，2003，37(Suppl.)：478-489.
[7] SANG H，ZENG J，QI Y，et al. Study on wheel wear mechanism of high-speed train in accelerating conditions[J]. Wear，2023，516：204597.
[8] SANG H，ZENG J，WANG Q，et al. Theoretical study on wheel wear mechanism of high-speed train under different braking modes[J]. Wear，2024，540：205262.
[9] YANG Y，LING L，WANG J，et al. A numerical study on tread wear and fatigue damage of railway wheels subjected to anti-slip control[J]. Friction，2023，11(8)：1470-1492.
[10] WANG R，YANG Y，LIU P，et al. A comparative study of WSP control methods：Effects on wear and RCF evolution in railway wheels[J]. Wear，2025，572：206049.
[11] 陈帅，吴磊，陶功权，等. 基于摩擦温升效应的地铁车轮磨耗特性研究[J]. 机械工程学报，2023，59(4)：213-220. CHEN Shuai，WU Lei，TAO Gongquan，et al. Study of wheel wear characteristic of subway vehicle based on the effect of friction temperature rising[J]. Journal of Mechanical Engineering，2023，59(4)：213-220.
[12] WEN B，TAO G，WEN Z. Prediction of locomotive wheel wear evolution considering thermo-mechanical coupling：Wear model and validation[J]. Wear，2025，571：205805.
[13] WANG P，YANG X，TAO G，et al. Development of a wheel wear prediction model considering the interaction of abrasive block-wheel and wheel-rail[J]. Wear，2024，550：205418.
[14] XIE Y，CAI W，LI Y，et al. A case study of the economic reprofiling method for a high-speed train’s wheel tread in the frequent contact area[J]. Wear，2024，556：205244.
[15] 朱爱华，张财胜，杨建伟，等. 制动时车轮型面演变及表面硬化对车轮磨耗的影响[J]. 北京交通大学学报，2025，49(2)：174-182. ZHU Aihua，ZHANG Caisheng，YANG Jianwei，et al. Influence of wheel profile evolution and surface hardening on wheel wear during braking[J]. Journal of Beijing Jiaotong University，2025，49(2)：174-182.
[16] SHEBANI A，IWNICKI S. Prediction of wheel and rail wear under different contact conditions using artificial neural networks[J]. Wear，2018，406-407：173-184.
[17] 王美琪，王艺，陈恩利，等. 基于恒等映射多层极限学习机的高速列车踏面磨耗预测模型[J]. 力学学报，2022，54(6)：1720-1731. WANG Meiqi，WANG Yi，CHEN Enli，et al. High-speed train tread wear prediction model based on I-ML-ELM[J]. Chinese Journal of Theoretical and Applied Mechanics，2022，54(6)：1720-1731.
[18] 殷冬年，解乃军，纪有旺. 基于改进GA-BP神经网络的FDM式3D打印机故障诊断[J]. 印刷与数字媒体技术研究，2025(4)：169-181. YIN Dongnian，XIE Naijun，JI Youwang. Fault diagnosis of FDM-style 3D printer based on improved GA-BP neural network[J]. Printing and Digital Media Technology Study，2025(4)：169-181.
[19] 王美琪，贾思贤，陈恩利，等. 基于SQPSO优化DELM的踏面磨耗测量模型[J]. 摩擦学学报，2021，41(1)：65-75. WANG Meiqi，JIA Sixian，CHEN Enli，et al. Measurement model of tread wear based on SQPSO optimized DELM[J]. Tribology，2021，41(1)：65-75.
[20] 张霞，邓银强，杨岳，等. 基于改进NARXNN神经网络的动车组车轮磨耗预测[J]. 铁道科学与工程学报，2026，23(2)：833-845. ZHANG Xia，DENG Yinqiang，YANG Yue，et al. Wear prediction of EMU wheels based on improved NARXNN neural network[J]. Journal of Railway Science and Engineering，2026，23(2)：833-845.
[21] 施法中. 计算机辅助几何设计与非均匀有理B样条[M]. 北京：高等教育出版社，2013. SHI Fazhong. Computer-aided geometric design and non-uniform rational B-splines[M]. Beijing：Higher Education Press，2013.
[22] 叶丽，谢明红. 采用积累弦长法拟合3次NURBS曲线[J]. 华侨大学学报(自然科学版)，2010，31(4)：383-387. YE Li，XIE Minghong. Fitting of cubic NURBS curves using the cumulative chord length method[J]. Journal of Huaqiao University (Natural Science)，2010，31(4)：383-387.
[23] GONG L，LIANG Y，YU X，et al. Damage prediction of hydraulic concrete in severe cold region based on FOTP-GM(1,1) model[J]. KSCE Journal of Civil Engineering，2023，27(7)：2993-3005.
[24] 马洪霞，丛林虎. 基于改进非等间隔灰色预测模型的导弹退化状态预测[J]. 计算机与现代化，2016(3)：5-10. MA Hongxia，CONG Linhu. Prediction of missile degradation condition based on improved unequal interval grey model[J]. Computer and Modernization，2016(3)：5-10.
[25] 郭炳辰，丁勇，左锋，等. 基于BWO- SVM的无损检测电涡流传感器温度补偿模型研究[J]. 自动化仪表，2025，46(8)：16-21. GUO Bingchen，DING Yong，ZUO Feng，et al. Research on temperature compensation model of eddy current sensor for nondestructive testing based on BWO-SVM[J]. Process Automation Instrumentation，2025，46(8)：16-21.
[26] XUE J，SHEN B. A novel swarm intelligence optimization approach：Sparrow search algorithm[J]. Systems Science & Control Engineering，2020，8(1)：22-34.
[27] 毛清华，张强. 融合柯西变异和反向学习的改进麻雀算法[J]. 计算机科学与探索，2021，15(6)：1155-1164. MAO Qinghua，ZHANG Qiang. Improved sparrow algorithm combining cauchy mutation and opposition-based learning[J]. Journal of Frontiers of Computer Science and Technology，2021，15(6)：1155-1164.
[28] 张璐. 基于Sine-SSA-BP算法的光谱反射率重建研究[D]. 杭州：浙江理工大学，2022. ZHANG Lu. Spectral reflectance reconstruction based on Sine-SSA-BP algorithm[D]. Hangzhou：Zhejiang Sci-Tech University，2022.
[29] 奚金明，郑荣艳. 基于自适应权重和莱维飞行的改进海鸥优化算法[J]. 计算机系统应用，2023，32(12)：171-179. XI Jinming，ZHENG Rongyan. Improved seagull optimization algorithm based on adaptive weight and levy flight[J]. Computer Systems & Applications，2023，32(12)：171-179.
[30] 王开文. 车轮接触点迹线迹轮轨接触几何参数的计算[J]. 西南交通大学学报，1984(1)：89-99. WANG Kaiwen. The track of wheel contact points and the calculation of wheel/rail geometric contact parameters[J]. Journal of Southwest Jiaotong University，1984(1)：89-99.
[31] PIOTROWSKI J，KIK W. A simplified model of wheel/rail contact mechanics for non-Hertzian problems and its application in rail vehicle dynamic simulations[J]. Vehicle System Dynamics，2008，46(1-2)：27-48.