基于有限元仿真的大规模法兰螺栓组应力神经网络预测研究

Stress neural network prediction for large-scale flange bolt groups based on finite element simulation

摘要

大规模法兰连接螺栓组的应力监测是风电机组等大型设备健康运行的重要需求。本研究以预紧法兰连接螺栓在变化的法兰外载荷作用下应力的有限元模拟结果为数据样本，采用BP、GA-BP和PSO-BP神经网络模型对螺栓组的应力进行预测和分析。结果表明，以相对误差为评价标准，3种模型的预测精度均随螺栓组中最小螺栓应力值的减小以及螺栓组中螺栓的最大应力差的增大而降低；螺栓组处于弹性阶段时的预测精度高于有螺栓进入塑性以后的情况；GA-BP和PSO-BP相对于BP具有更高的预测精度，PSO-BP的预测精度最高；对于更值得关注的高应力水平（400 MPa以上）螺栓，3种模型的预测精度均较高，整体平均误差均在2%以下。

Abstract

Stress monitoring of large-scale flange connection bolt groups is an important requirement for the healthy operation of large equipment such as wind turbines. In this work, the finite element simulation results of the stress on pretensioned flange connection bolts under varying external flange loads are taken as the data sample, and three neural network models, namely, BP, GA-BP and PSO-BP, are employed to predict and analyze the stress of the bolt group. The results indicate that, taking the relative error as the evaluation criterion, the prediction accuracy of the three models decreases with the reduction of the minimum bolt stress value in the bolt group and the increase of the maximum stress difference among the bolts. The prediction accuracy is higher when the bolt group is in the elastic stage than when some bolts have entered the plastic stage. GA-BP and PSO-BP have higher prediction accuracy than BP, while PSO-BP has the highest accuracy. For the high-stress bolts (above 400 MPa) that deserve more attention, the prediction accuracy of all the three models is quite high, all with an overall average error below 2%.

关键词

法兰连接螺栓 / 螺栓应力 / 有限元分析 / 神经网络预测

Key words

flange connection bolts / bolt stress / FEM analysis / neural network prediction

参考文献

[1] 王韦强, 马官兵, 汤建帮, 等. 超声检测技术在核电厂螺栓在役检查中的应用[J]. 核动力工程, 2025, 46(3): 266-270.

[2] WANG Weiqiang, MA Guanbing, TANG Jianbang, et al. Application of ultrasonic technique in in-service inspection of bolts in nuclear power plants[J]. Nuclear power engineering, 2025, 46(3): 266-270(in Chinese).

[3] 栾宇, 关振群, 程耿东, 等. 螺栓法兰连接结构的拉压不同刚度动力学模型[J]. 导弹与航天运载技术, 2021(4): 30-35.

[4] LUAN Yu, GUAN Zhenqun, CHENG Gengdong, et al. Simplified dynamic model with different tensile and compressive stiffness for bolted flange connections[J]. Missiles and space vehicles, 2021(4): 30-35(in Chinese).

[5] 赵排航, 王克印, 黄海英, 等. 扭矩扳手发展现状及展望[J]. 工具技术, 2015, 49(10): 20-24.

[6] ZHAO Paihang, WANG Keyin, HUANG Haiying, et al. Development status and prospect of torque wrench[J]. Tool engineering, 2015, 49(10): 20-24(in Chinese).

[7] 郭永兴, 李志雄. 增敏型光纤布拉格光栅锚索测力传感器[J]. 激光与光电子学进展, 2023, 60(5): 68-75.

[8] GUO Yongxing, LI Zhixiong. Sensitized fiber bragg grating anchor cable force sensor[J]. Laser & optoelectronics progress, 2023, 60(5): 68-75(in Chinese).

[9] 吴强, 俞志伟, 吉爱红, 等. 一种小型电阻应变式三维力传感器的仿真设计[J]. 中国机械工程, 2011, 22(11): 1288-1293.

[10] WU Qiang, YU Zhiwei, JI Aihong, et al. Development of a miniature resistance strain three dimensional force sensor[J]. China mechanical engineering, 2011, 22(11): 1288-1293(in Chinese).

[11] JOHNSON G C, HOLT A C, CUNNINGHAM B. An ultrasonic method for determining axial stress in bolts[J]. Journal of testing and evaluation, 1986, 14(5): 253-259.

[12] 刘家斌, 王雪梅, 倪文波. 螺栓轴向应力-超声波渡越时间自动标定系统研究[J]. 中国测试, 2020, 46(3): 91-96.

[13] LIU Jiabin, WANG Xuemei, NI Wenbo. Research on automatic calibration system of bolt axial stress-ultrasonic time of flight[J]. China measurement & testing technology, 2020, 46(3): 91-96(in Chinese).

[14] 周士栋, 薛扬, 马晓晶, 等. 基于SCADA数据的风电机组关键载荷预测[J]. 农业工程学报, 2018, 34(2): 219-225.

[15] ZHOU Shidong, XUE Yang, MA Xiaojing, et al. Prediction of wind turbine key load based on SCADA data[J]. Transactions of the Chinese Society of Agricultural Engineering, 2018, 34(2): 219-225(in Chinese).

[16] 尹少平, 冯沫, 王灵梅, 等. 大型风电机组塔筒应力监测点的选取与预警[J]. 电网与清洁能源, 2018, 34(10): 76-82.

[17] YIN Shaoping, FENG Mo, WANG Lingmei, et al. Selection and early warning of monitoring points of the tower of the large sized wind turbine unit[J]. Power system and clean energy, 2018, 34(10): 76-82(in Chinese).

[18] 薛磊, 王灵梅, 孟恩隆, 等. 基于SCADA数据和改进BP神经网络的塔筒应力预测[J]. 噪声与振动控制, 2021, 41(1): 95-102.

[19] XUE Lei, WANG Lingmei, MENG Enlong, et al. Stress prediction of wind turbine tower drums based on SCADA data and improved BP neural network[J]. Noise and vibration control, 2021, 41(1): 95-102(in Chinese).

[20] 毛晓敏, 张慧华, 纪晓磊, 等. 基于XFEM与BP神经网络的裂纹智能识别[J]. 应用力学学报, 2022, 39(6): 1158-1167.

[21] MAO Xiaomin, ZHANG Huihua, JI Xiaolei, et al. Intelligent crack identification based on XFEM and BP neural network[J]. Chinese journal of applied mechanics, 2022, 39(6): 1158-1167(in Chinese).

[22] 汤群益, 张栋梁, 陈金军, 等. 基于BP神经网络的风电机组钢混组合式塔架结构频率预测[J]. 固体力学学报, 2021, 42(5): 612-622.

[23] TANG Qunyi, ZHANG Dongliang, CHEN Jinjun, et al. The frequency prediction for the hybrid structure of wind turbine foundation based on the back propagation(BP)neural network[J]. Chinese journal of solid mechanics, 2021, 42(5): 612-622(in Chinese).

[24] 杨超, 张开富. 基于PSO-BiLSTM神经网络的机身筒段应力预测[J]. 航空学报, 2023, 44(7): 261-270.

[25] YANG Chao, ZHANG Kaifu. Stress prediction of fuselage tube section based on PSO-BiLSTM neural network[J]. Acta aeronautica et astronautica Sinica, 2023, 44(7): 261-270(in Chinese).

[26] 余运俊, 衷国瑛, 万晓凤, 等. 基于遗传算法及BP神经网络的混合孤岛检测方法[J]. 可再生能源, 2018, 36(5): 701-706.

[27] YU Yunjun, ZHONG Guoying, WAN Xiaofeng, et al. A hybrid islanding detection method based on genetic algorithm and BP neural network[J]. Renewable energy resources, 2018, 36(5): 701-706(in Chinese).

[28] 高培昱. 风力发电变桨连接螺栓故障诊断和可靠性分析[D]. 太原: 太原理工大学, 2017.

[29] 庞松. 基于超声波的风电机组多螺栓应力在线监测研究[D]. 成都: 西南交通大学, 2021.

[30] 薛磊. 大型风电机组载荷监测与延寿评估技术研究[D]. 太原: 山西大学, 2020.

[31] 徐祉康, 易伟同, 祝磊. 风力发电机组钢塔筒M64螺栓连接法兰受拉承载性能试验研究[J]. 建筑结构, 2021, 51(增刊1): 1522-1527.

[32] XU Zhikang, YI Weitong, ZHU Lei. Experimental study on tensile bearing capacity of M64 bolted flange of wind turbine tower[J]. Building structure, 2021, 51(S1): 1522-1527(in Chinese).

[33] 徐静, 薛欣玮, 卢健. 基于ANSYS软件的螺栓法兰连接结构应力和疲劳分析[J]. 机械制造, 2018, 56(12): 17-20.

[34] XU Jing, XUE Xinwei, LU Jian. Stress and fatigue analysis of bolted flange connection structure based on ANSYS software[J]. Machinery, 2018, 56(12): 17-20(in Chinese).

[35] 唐洲. 风力发电机组高强度螺栓连接有限元模拟研究[D]. 绍兴: 绍兴文理学院, 2021.

[36] 康凯, 孟雨萱, 郑健, 等. 基于BP神经网络的周口店遗址裂缝变形监测数据分析与预测[J]. 安全与环境工程, 2024, 31(6): 169-178.

[37] KANG Kai, MENG Yuxuan, ZHENG Jian, et al. Analysis and prediction of crack deformation monitoring data of Zhoukoudian Site based on back propagation neural network[J]. Safety and environmental engineering, 2024, 31(6): 169-178(in Chinese).

[38] 王睿, 漆泰岳, 冯剑, 等. 基于遗传算法的BP神经网络隧道施工参数正反演分析与应用[J]. 铁道学报, 2016, 38(4): 98-104.

[39] WANG Rui, QI Taiyue, FENG Jian, et al. Application of genetic algorithm based ANN to forward and back analysis on tunnel construction parameters[J]. Journal of the China Railway Society, 2016, 38(4): 98-104(in Chinese).

[40] LIU H M, QIN X P, HUANG S, et al. Geometry characteristics prediction of single track cladding deposited by high power diode laser based on genetic algorithm and neural network[J]. International journal of precision engineering and manufacturing, 2018, 19(7): 1061-1070.

[41] ZHANG Y X, GAO X D, KATAYAMA S. Weld appearance prediction with BP neural network improved by genetic algorithm during disk laser welding[J]. Journal of manufacturing systems, 2015, 34: 53-59.

[42] 郭迎庆, 周敏, 吴景, 等. 基于遗传算法优化的磁流变减震结构模糊控制研究[J]. 应用力学学报, 2020, 37(4): 1500-1507.

[43] GUO Yingqing, ZHOU Min, WU Jing, et al. Study on fuzzy control of magneto-rheological structure based on genetic algorithm optimization[J]. Chinese journal of applied mechanics, 2020, 37(4): 1500-1507(in Chinese).

[44] 李淑芸, 高轩, 郑修鹏, 等. 基于有限测点信息的管路振动载荷识别方法研究[J]. 应用力学学报, 2025, 42(5): 1049-1056.

[45] LI Shuyun, GAO Xuan, ZHENG Xiupeng, et al. Identification method of pipeline vibration loads with limited measurement information[J]. Chinese journal of applied mechanics, 2025, 42(5): 1049-1056(in Chinese).

[46] 张昕阳, 申玉生, 常铭宇, 等. 基于GA-BP神经网络的隧道围岩相似材料配合比设计[J]. 现代隧道技术, 2024, 61(6): 82-91.

[47] ZHANG Xinyang, SHEN Yusheng, CHANG Mingyu, et al. Mix proportion design of similar materials for tunnel surrounding rocks based on GA-BP neural network[J]. Modern tunnelling technology, 2024, 61(6): 82-91(in Chinese).

[48] KENNEDY J, EBERHART R. Particle swarm optimization[C]// Proceedings of ICNN'95-International Conference on Neural Networks. Piscataway, NJ, USA:IEEE, 1995: 1942-1948.

[49] 黄艺航, 蔡凯武, 黄晓智, 等. 基于机器学习的通信电缆故障检测与定位方法[J]. 自动化与信息工程, 2024, 45(1): 35-41.

[50] HUANG Yihang, CAI Kaiwu, HUANG Xiaozhi, et al. Fault detection and localization method of communication cables based on machine learning[J]. Automation & information engineering, 2024, 45(1): 35-41(in Chinese).

[51] 李大权, 李顺才, 吴春力. 粒子群算法优化的车削温度组合预测模型研究[J]. 应用力学学报, 2020, 37(6): 2354-2361.

[52] LI Daquan, LI Shuncai, WU Chunli. Research on combined prediction model of turning temperature optimized by particle swarm optimization[J]. Chinese journal of applied mechanics, 2020, 37(6): 2354-2361(in Chinese).

[53] 张宇鹏, 刘韡. 基于PSO-ANFIS模型的矩形薄板荷载情况反演[J]. 应用力学学报, 2024, 41(6): 1336-1342.

[54] ZHANG Yupeng, LIU Wei. Load inversion of rectangular thin plate based on PSO-ANFIS model[J]. Chinese journal of applied mechanics, 2024, 41(6): 1336-1342(in Chinese).

[55] 李晓娟, 张芳媛, 喻玲. 基于主成分分析-BP神经网络的风电备件需求预测[J]. 科学技术与工程, 2024, 24(1): 281-288.

[56] LI Xiaojuan, ZHANG Fangyuan, YU Ling. Wind power spare parts demand forecasting based on PCA-BP neural network[J]. Science technology and engineering, 2024, 24(1): 281-288(in Chinese).

基金资助

国家自然科学基金资助项目(52401343)

四川省自然科学基金资助项目(2024NSFSC1953)