| 11 | 0 | 77 |
| 下载次数 | 被引频次 | 阅读次数 |
为解决国产数控机床可靠性不足及传统故障预测方法对故障演化规律挖掘深度不够的问题，该研究创造性地引入地震学古登堡-里克特(G-R)模型，探寻机床故障演化新规律。研究将机床故障等级与次数分别类比为地震震级与震次，验证了G-R模型的b值可作为表征机床健康状态及预测严重故障风险的动态指标。为预测b值趋势，构建并对比了自适应神经模糊系统(ANFIS)与神经网络(BP)模型。结果表明，两种模型均能有效捕捉b值动态，经深度特征工程优化的BP神经网络模型表现更优。分析表明，模型的卓越性能主要归功于对“故障等级”和“故障次数”的非线性特征的构建。该研究证实了跨学科模型在工业故障分析中的潜力，为数控机床智能运维提供了新预测框架，并强调了特征工程的核心驱动作用。
Abstract：To address the problems of insufficient reliability in domestic CNC machine tools and the shallow pattern mining of traditional fault prediction methods, the present study creatively introduces the Gutenberg-Richter(G-R)model from seismology to explore new laws in machine tool fault evolution.By analogizing machine tool fault grades and counts to earthquake magnitudes and frequencies, the study validates that the G-R model′s b-value can serve as a dynamic indicator for characterizing machine health and predicting severe failure risks.To forecast the b-value trend, both ANFIS and BP neural network models were developed and compared.The results show that both models effectively capture the b-value′s dynamics, with the BP neural network model, enhanced by in-depth feature engineering, exhibiting superior performance.Analysis reveals that the model′s excellent performance is mainly attributed to the construction of non-linear features from “fault grade” and “fault count”.The study confirms the potential of interdisciplinary models in industrial fault analysis, offers a new predictive framework for the intelligent operation and maintenance of CNC machine tools, and underscores the pivotal role of feature engineering as the core driver.
[1] 杨海鱼.可靠性模型及其在机床可靠性评估中的应用[D].兰州：兰州理工大学，2018.
[2] 熊福意.数控系统可靠性研究[D].武汉：武汉工程大学，2017.
[3] 刘玉敏，承姿辛，胡涛.基于FTA和Markov过程的数控机床可靠性分析[J].工程与试验，2025,65(4):28-30,64.
[4] 陈学忠，吕晓健，王慧敏.中国大陆地震活跃期和平静期的b值与地震趋势研究[J].地震，2001(1):59-62.
[5] 张双凤，张小涛，赵志远.利用地震活动性参数b值判断张渤地震带中西段地震危险性[J].地震地磁观测与研究，2018,39(4):51-57.
[6] Yang Congbin,Wang Yongqi,Yan Jun,et al.A fault hierarchical propagation reliability improvement method for CNC machine tools based on spatiotemporal factors coupling[J].Reliability Engineering and System Safety,2025,253:110488.
[7] Chen Chuanhai,Li Guanyu,Liu Zhifeng,et al.A Comprehensive Evaluation Method for Generalized Reliability of CNC Machine Tools Based on Improved Entropy-Weighted Extensible Matter-Element Method[J].International Journal of Precision Engineering and Manufacturing,2023,24(8):3469-3485.
[8] Li Junfa,Li Yulong,Wen Shutao,et al.A novel method of key meta-action unit integrated identification for CNC machine tool reliability[J].Computers & Industrial Engineering,2023,177:109062.
[9] Jie Pengfei,Zhou Yuan,Zhang Zhijie,et al.Heating energy consumption prediction based on improved GA-BP neural network model[J].Energy,2022,254:124314.
[10] Wu Yuan,Chen Xiaolu ,Xie Qian,et al.Pyrolysis behavior of Sabina chinensis leaves:kinetic and thermodynamic analysis based on TG-FTIR and BP neural network[J].Energy,2023,282:128913.
[11] Zhou Yongxi.An Improved BP Neural Network Credit Default Prediction Model Based on SMOTE and Random Forests[J].Forum on Research and Innovation Management,2025,3(7)115-125.
[12] 王光明，付虹，孔德育，等.小滇西地区M≥5地震前b值时空变化特征 [J].地震研究，2020,43(2):363-374.
[13] Zhang Haiping,Zhao lei,Yang Song,et al.Fatigue evaluation of Orthotropic steel deck welds based on WIM data and UD-BP neural network[J].Structures,2023,56:05007.
[14] Yin XiaoJu,Mu Qizheng,Li Bo,et al.Vibration modeling and simulation analysis of wind turbine flexible tower based on BP neural network[J].Measurement and Control,2023,58(7/8):1143-1152.
[15] Sun Yuanquan,Wang Zhongli,Wang Jiahui,et al.Short-Term Solar Photovoltaic Power Prediction Utilizing the VMD-BKA-BP Neural Network[J].Symmetry,2025,17(5):784-901.
[16] 于捷，石耀霖.基于G-R曲线的数控机床故障率分析[J].机床与液压，2009,37(10):257-258,251.
基本信息:
中图分类号:TG659;TP18
引用信息:
[1]王铁彬,于捷,邓凤玲,等.基于G-R模型b值的数控机床故障趋势预测研究[J].梧州学院学报,2026,36(01):39-47.
基金信息:
梧州市科技计划项目(202402018)
2025-08-09
2025
2026-03-23
2026
1
2026-02-15
2026-02-15