某海洋平台在多次维修中发现在生活楼与甲板连接的角隅处的裂纹有扩展现象，提出了根据裂纹周边多维应变进行裂纹长度识别的思想。搭建了含有初始裂纹的海洋平台有限元模型，以多维应变数据和对应裂纹长度分别作为机器学习模型特征输入与输出，通过梯度回归提升树（GBRT）模型对裂纹长度进行预测。测试结果表明，该模型对裂纹长度预测MSE（均方误差）值可达0.0006，R2可达0.9991，且该模型对噪声有良好的抗干扰性。
After many times of maintenance of an offshore platform it is found that there is crack propagation at the corner which connects the living building to the deck. The idea of crack length identification based on the multi-dimensional strain around the crack is proposed in this paper. A finite element model of offshore platform with initial crack is built. Multi-scale strain data and corresponding crack length are used as feature input and output for the machine learning model respectively. The crack length is predicted by gradient boosting regression tree (GBRT) model. Test results show that the value of MSE and R2 can reach 0.0006 and 0.9991, respectively. At the same time, the model is proved to have good anti-interference to noise.
[1] Yang D, Wang J Q, Li D, et al. Fatigue crack monitoring using plastic optical fibre sensor[J]. Procedia Structural In-tegrity, 2017, 5:1168-1175.
[2] Jin X, Yuan S F, Chen J. On crack propagation monitor-ing by using reflection spectra of AFBG and UFBG sen-sors[J]. Sensors and Actuators A:Physical, 2019, 285:491-500.
[3] 何弯弯,曾捷,夏裕彬,等.空间柔性充气结构分布式光纤裂纹损伤监测方法[J].南京航空航天大学学报, 2019, 51(3):297-304.
[4] 方桂华.基于FBG的CFRP加固钢结构裂纹检测方法研究[D].武汉:武汉理工大学, 2017.
[5] 黄博,白生宝,宁宁,等.基于FBG动态应变监测的金属结构损伤识别方法研究[J].航空制造技术, 2017, 60(19):67-71.
[6] Gao L, Gong Y H, Liu H L, et al. Experiment and numeri-cal study on deformation measurement of cast-in-place concrete large-diameter pipe pile using optical frequency domain reflectometer technology[J]. Applied Sciences, 2018, 8(9):1450.
[7] Sounthararajah A, Wong L, Nguyen N, et al. Evaluation of flexural behaviour of cemented pavement material beams using distributed fibre optic sensors[J]. Construction and Building Materials, 2017, 156:965-975.
[8] Li X, Wang C, Ju H J, et al. Surface defect detection model for aero-engine components based on improved YOLOv5[J]. Applied Sciences, 2022, 12(14):7235.
[9] Su X, Jia Z G, Ma G D, et al. Image-based crack detec-tion method for FPSO module support[J]. Buildings, 2022, 12(8):1147.
[10] Jia Z G, Su X, Ma G D, et al. Crack identification for marine engineering equipment based on improved SSD and YOLOv5[J]. Ocean Engineering, 2023, 268:113534.
[11] 杨文忠,张志豪,吾守尔·斯拉木,等.基于时间序列关系的GBRT交通事故预测模型[J].电子科技大学学报, 2020, 49(4):615-621.
[12] Cai W T, Wei R H, Xu L H, et al. A method for model-ling greenhouse temperature using gradient boost deci-sion tree[J]. Information Processing in Agriculture, 2022, 9(3):343-354.
[13] Raja D P, Vasudevan V. A novel thinking to enhance the gradient boost decision tree classifier for identifying path in autonomous vehicle[J]. Turkish Journal of Com-puter And Mathematics Education, 2021, 12(7):1282-1288.
[14] Wang J D, Li P, Ran R, et al. A short-term photovoltaic power prediction model based on the gradient boost deci-sion tree[J]. Applied Sciences, 2018, 8(5):689.
[15] Mishra S. An optimized gradient boost decision tree us-ing enhanced African buffalo optimization method for cy-ber security intrusion detection[J]. Applied Sciences, 2022, 12(24):12591.
[16] 周晓敏,郝勇凯,丛文韬,等.基于梯度提升决策树模型的冷连轧机颤振研究[J].振动与冲击, 2021, 40(13):154-158.
[17] Zulfiqar H, Yuan S S, Huang Q L, et al. Identification of cyclin protein using gradient boost decision tree algo-rithm[J]. Computational and Structural Biotechnology Journal, 2021, 19:4123-4131.
[18] 谷雨.基于XFEM的梁柱节点断裂分析及裂纹扩展研究[D].兰州:兰州理工大学, 2016.
[19] Chen T, Shang H, Bi Q Z. A prediction method of fiveaxis machine tool energy consumption with GBRT algo-rithm[C]//Proceedings of IEEE 5th International Confer-ence on Mechatronics System and Robots (ICMSR). Pis-cataway, NJ:IEEE, 2019.
[20] Bonaccorso G. Mechine learning algorithms[M].北京:机械工业出版社, 2018.
