基于随机森林算法的行星滚柱丝杠副摩擦力矩预测

doi:10.3969/j.issn.1004-132X.2025.07.013

摘要/Abstract

摘要： 摩擦力矩增大会加剧行星滚柱丝杠副磨损，严重影响其使用及寿命。探讨了利用机器学习算法预测行星滚柱丝杠副摩擦力矩的可行性，分析了行星滚柱丝杠副摩擦力矩与磨损状态的关系。采用基于随机森林、支持向量回归和BP神经网络的机器学习算法预测了不同转数行星滚柱丝杠副的摩擦力矩变化。研究结果表明，基于随机森林算法对行星滚柱丝杠副摩擦力矩的预测准确率达到97%。

关键词: 行星滚柱丝杠, 摩擦力矩, 机器学习, 随机森林算法

Abstract: The increase in frictional torques of planetary roller screw pairs led increased wear of planetary roller screw mechanisms(PRSM), which seriously affected the use and service life of PRSM. The feasibility of using ML algorithms to predict the frictional torques of PRSM was explored, and the relationship between PRSM frictional torques and wear states was analyzed. Random forest algorithm, support vector regression, and BP neural network were used to predict the changes in frictional torques of PRSM at different rotational speeds. The results demonstrate that the random forest algorithm achieves the prediction accuracy of 97% for the frictional torques of PRSM.

Key words: planetary roller screw, frictional moment, machine learning(ML), random forest algorithm

中图分类号:

TH132

徐洋, 祖莉, 李伟龙, 刘晓玲, 何建樑, 刘俊. 基于随机森林算法的行星滚柱丝杠副摩擦力矩预测[J]. 中国机械工程, 2025, 36(07): 1505-1511.

XU Yang, ZU Li, LI Weilong, LIU Xiaoling, HE Jianliang, LIU Jun. Prediction of Frictional Torques of Planetary Roller Screw Pairs Based on Random Forest Algorithm[J]. China Mechanical Engineering, 2025, 36(07): 1505-1511.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: https://www.cmemo.org.cn/CN/10.3969/j.issn.1004-132X.2025.07.013

https://www.cmemo.org.cn/CN/Y2025/V36/I07/1505

参考文献

［1］乔冠. 行星滚柱丝杠副摩擦力矩及热特性的理论与实验研究［D］. 西安:西北工业大学, 2019.
QIAO Guan. Theoretical and Experimental Research on the Friction Torque and Thermal Characteristics of the Planetary Roller Screw Mechanism［D］. Xi’an:Northwestern Polytechnical University, 2019.
［2］解志杰, 张传伟, 薛其河, 等. 行星滚柱丝杠动态摩擦力矩和传动效率分析［J］. 摩擦学学报, 2019, 39(4):489-496.
XIE Zhijie, ZHANG Chuanwei, XUE Qihe, et al. Analysis of Dynamic Friction Torque and Transmission Efficiency of Planetary Roller Screw［J］. Tribology, 2019, 39(4):489-496.
［3］濮建荣. 行星滚柱丝杠副接触、摩擦与工作特性的理论与试验研究［D］. 南京:南京理工大学, 2015.
PU Jianrong. Theoretical and Experimental Study on Contact, Friction and Working Characteristics of Planetary Roller Screw Pair［D］. Nanjing:Nanjing University of Science and Technology, 2015.
［4］徐强, 王水铭, 赵国平, 等. PWG型差动丝杠摩擦转矩分析与效率求解［J］. 现代制造工程, 2016(5):138-142.
XU Qiang, WANG Shuiming, ZHAO Guoping, et al. The Analysis of Friction Torque and Calculation of Efficiency of PWG Planetary Roller Screw［J］. Modern Manufacturing Engineering, 2016(5):138-142.
［5］XING Mingcai, ZHANG Bohong, DENG Pan, et al. A Novel Wear Prediction Model for Planetary Roller Screw Based on Universal Sliding Distance Model［J］. Tribology International, 2022, 175:107851.
［6］GU Weikai, LI Chaoyang, MIAO Jiacheng, et al. Wear Prediction Method of Differential Planetary Roller Screws Considering the Ambient Temperature Variations［J］. Applied Sciences, 2023, 13(19):10609.
［7］AURGAN G, FRIDRICI V, KAPSA P, et al. Experimental Simulation of Rolling-Sliding Contact for Application to Planetary Roller Screw Mechanism［J］. Wear, 2015, 332:1176-1184.
［8］MIAO Jiacheng, DU Xing, LI Chaoyang, et al. Lubrication and Wear Analysis of Planetary Roller Screw Mechanism with Threaded Surface Roughness in Thermal Elastohydrodynamic Lubrication［J］. Tribology Transactions, 2022, 65(6):1069-1087.
［9］MENG Junjie, DU Xing, ZHAO Xin, et al. Research on Contact and Wear Characteristics of the Planetary Roller Screw Mechanism with Screw Misalignments［J］. Lubricants, 2022, 10(6):115.
［10］BA瘙塁 Hasan, KARABACAK Y E. Machine Learning-based Prediction of Friction Torque and Friction Coefficient in Statically Loaded Radial Journal Bearings［J］. Tribology International, 2023, 186:108592.
［11］CHENG Ganlin, XIANG Chong, GUO Fei, et al. Prediction of the Tribological Properties of a Polymer Surface in a Wide Temperature Range Using Machine Learning Algorithm Based on Friction Noise［J］. Tribology International, 2023, 180:108213.
［12］PATURI U M R, PALAKURTHY S T, REDDY N S. The Role of Machine Learning in Tribology:a Systematic Review［J］. Archives of Computational Methods in Engineering, 2023, 30(2):1345-1397.
［13］ZHAO Honghao, E Jiming, CHEN Shengshan, et al. Prediction of Friction Coefficient of Polymer Surface Using Variational Mode Decomposition and Machine Learning Algorithm Based on Noise Features［J］. Tribology International, 2024, 191:109184.
［14］KOLEV M. COF-RF-Tool:a Python Software for Predicting the Coefficient of Friction of Open-cell AlSi10Mg-SiC Composites Using Random Forest Model［J］. Software Impacts, 2023, 17:100520.
［15］周亮. 基于滚珠丝杠副性能检测平台的故障智能诊断研究［D］. 济南:山东建筑大学, 2020.
ZHOU Liang. Research on Intelligent Fault Diagnosis Based on Performance Detection Platform of Ball Screw Pair［D］. Jinan:Shandong Jianzhu University, 2020.
［16］LIU Chao, HE Yan, LI Yufeng, et al. Predicting Residual Properties of Ball Screw Raceway in Whirling Milling Based on Machine Learning［J］. Measurement, 2021, 173:108605.
［17］HU Rui, WEI Peitang, DU Xuesong, et al. Investigation of Loaded Contact Characteristics of Planetary Roller Screw Mechanism Based on Influence Coefficient Method and Machine Learning［J］. Advanced Engineering Informatics, 2023, 58:102146.
［18］HU Rui, WEI Peitang, LIU Huaiju, et al. Investigation on Load Distribution among Rollers of Planetary Roller Screw Mechanism Considering Machining Errors:Analytical Calculation and Machine Learning Approach［J］. Mechanism and Machine Theory, 2023, 185:105322.
［19］WU Hanlin, WEI Peitang, HU Rui, et al. Study on the Relationship between Machining Errors and Transmission Accuracy of Planetary Roller Screw Mechanism Using Analytical Calculations and Machine-learning Model［J］. Journal of Computational Design and Engineering, 2023, 10(1):398-413.
［20］马尚君, 刘更, 佟瑞庭. 行星滚柱丝杠副摩擦力矩及传动效率研究［J］. 哈尔滨工业大学学报, 2013, 45(11):74-79.
MA Shangjun, LIU Geng, TONG Ruiting. The Frictional Moment and Transmission Efficiency of Planetary Roller Screw［J］. Journal of Harbin Institute of Technology, 2013, 45(11):74-79.
［21］HOJJAT Y,MAHDI AGHELI M. A Comprehensive Study on Capabilities and Limitations of Roller-Screw with Emphasis on Slip Tendency［J］. Mechanism and Machine Theory, 2009, 44(10):1887-1899.

[1]	李涛, 邓林辉, 莫彬, 石非凡, 刘伟嵬. 基于混合遗传蚁群优化随机森林算法的激光熔覆Ni60裂纹预测与工艺参数优化[J]. 中国机械工程, 2025, 36(06): 1322-1328,1337.
[2]	惠生猛1, 毛晓博4, 湛利华1, 2, 3. 铝锂合金回弹预测的机器学习及有限元仿真与实验[J]. 中国机械工程, 2024, 35(12): 2114-2121.
[3]	尤勇1, 2, 孟云龙1, 2, 吴景涛1, 2, 王长青3. 基于鲸鱼优化算法支持向量回归的汽车运动状态估计[J]. 中国机械工程, 2024, 35(06): 973-981，992.
[4]	陈昱, 项薇, 龚川. 基于数据挖掘的注塑产品质量在线故障检测及预测[J]. 中国机械工程, 2023, 34(14): 1749-1755.
[5]	吴翰林, 魏沛堂, 蔡磊, 胡瑞, 刘怀举, 杜雪松, 朱才朝, 刘浪. 基于加工误差敏感度与模糊层次分析法的行星滚柱丝杠公差匹配优化方法[J]. 中国机械工程, 2022, 33(22): 2693-2703.
[6]	李方坡, 路彩虹, 赵靖宵, 李秀程, 尚成嘉. 基于机器学习的石油装备用大截面高强韧马氏体钢智能设计与性能研究[J]. 中国机械工程, 2022, 33(19): 2325-2330.
[7]	王顶, 吴德宇, 杨达亮, 陈立挺, 叶锦华, 吴海彬. 具有压力和位置检测功能的气囊型触觉传感器[J]. 中国机械工程, 2022, 33(14): 1691-1696,1706.
[8]	刘永兴, 唐小琦, 钟靖龙, 钟震宇, 周向东. 基于轻量级梯度提升机的非对称风险注塑成形产品尺寸预测模型[J]. 中国机械工程, 2022, 33(08): 965-969.
[9]	翁剑, 庄可佳, 浦栋麟, 丁汉, . 基于机器学习和多目标算法的钛合金插铣优化[J]. 中国机械工程, 2021, 32(07): 771-777.
[10]	陈光耀, 李恒, 贺子芮, 马俊, 李光俊, 付颖. [成形过程的数据挖掘与深度学习方法]基于机器学习的管材弯曲回弹有效预测与补偿[J]. 中国机械工程, 2020, 31(22): 2745-2752.
[11]	姜少飞;邬天骥;彭翔;李吉泉;李治;孙涛. 基于XGBoost特征提取的数据驱动故障诊断方法[J]. 中国机械工程, 2020, 31(10): 1232-1239.
[12]	隗寒冰1;曹旭1;赖锋2. 智能汽车环境感知算法测试评价系统开发[J]. 中国机械工程, 2018, 29(19): 2298-2305,2311.
[13]	张大伟, 赵升吨. 行星滚柱丝杠副滚柱塑性成形的探讨[J]. 中国机械工程, 2015, 26(3): 385-389.
[14]	刘庭伟, 张宁, 梁伟, 张亮. 星载活动部件用轴承的失效仿真分析及试验验证[J]. 中国机械工程, 2014, 25(21): 2864-2868.
[15]	马尚君, 刘更, 佟瑞庭, 张文杰. 考虑滚柱节圆偏移的反向式行星滚柱丝杠副运动学分析[J]. 中国机械工程, 2014, 25(11): 1421-1426.