基于IWOA-BERT的磨煤机故障预警

PDF(2173 KB)

振动与冲击 ›› 2025, Vol. 44 ›› Issue (11) : 288-294.

故障诊断分析

基于IWOA-BERT的磨煤机故障预警

段明达*

作者信息 +

Coal mill fault warning based on IWOA-BERT

DUAN Mingda*

Author information +

文章历史 +

摘要

实现磨煤机的故障预警技术可以降低事故发生率，针对其运行中随机扰动多，且故障早期阶段不易判断的特点，提出了一种基于改进鲸鱼算法优化BERT模型(IWOA-BERT)的故障预警方法｡首先,通过改进传统鲸鱼算法的收敛因子和引入高斯变异算子来增强算法的寻优能力;其次,选取与磨煤机故障相关的特征参数作为建模变量,利用改进鲸鱼算法优化BERT模型的超参数,建立故障预警模型;然后,计算正常状态数据中每个滑动窗口的相似度均值，选取最小值乘以阈值系数确定预警阈值;最后,根据专家系统推理预警时刻的故障类型并给出检修指导｡将所提方法应用于某350MW机组磨煤机的运行中,结果表明模型的预测准确率高,且能提前24秒给出预警信息,为工程应用提供了参考｡

Abstract

The implementation of fault warning technology for coal mills can reduce the incidence of accidents. In response to the characteristics of frequent random disturbances during operation and difficulty in identifying faults in the early stages, a fault warning method based on the improved whale algorithm optimized BERT model (IWOA-BERT) is proposed. Firstly, by improving the convergence factor of the traditional whale algorithm and introducing a Gaussian mutation operator, the optimization ability of the algorithm is enhanced; Secondly, select characteristic parameters related to coal mill faults as modeling variables, use the improved whale algorithm to optimize the hyperparameters of the BERT model, and establish a fault warning model; Then, calculate the average similarity of each sliding window in the normal state data, select the minimum value and multiply it by the threshold coefficient to determine the warning threshold; Finally, based on the expert system, infer the type of fault at the warning time and provide maintenance guidance. The proposed method was applied to the operation of a coal mill in a 350MW unit, and the results showed that the model had high prediction accuracy and could provide warning information 24 seconds in advance, providing a reference for engineering applications.

导出引用

段明达. 基于IWOA-BERT的磨煤机故障预警[J]. 振动与冲击, 2025, 44(11): 288-294

DUAN Mingda. Coal mill fault warning based on IWOA-BERT[J]. Journal of Vibration and Shock, 2025, 44(11): 288-294

参考文献

[1] Yin W, Hu Y, Ding G, et al. Health indicator construction and application of coal mill based on the dynamic model. Structural Health Monitoring. 2024, 23(2): 686-700.
[2] 孙同敏. 基于ISAE的磨煤机故障预测与诊断方法研究[J].煤炭工程, 2021, 53(05): 148-155.
SUN Tongmin. Prediction and diagnosis of coal mill fault based on ISAE[J]. Coal Engineering, 2021, 53(05): 148-155.
[3] 向玲, 朱浩伟, 丁显, 等. 基于CAE与BiLSTM结合的风电机组齿轮箱故障预警方法研究[J]. 动力工程学报, 2022, 42(06): 514-521.
XIANG Ling, ZHU Haowei, DING Xian, et al. Research on fault warning method of wind turbine gearbox based on CAE and BiLSTM[J]. Journal of Chinese Society of Power Engineering, 2022, 42(06): 514-521.
[4] 郑小霞, 陈广宁, 任浩翰, 等. 基于改进VMD和深度置信网络的风机易损部件故障预警[J]. 振动与冲击, 2019, 38(08): 153-160+179.
ZHENG Xiaoxia, CHEN Guangning, REN Haohan, et al. Fault detection of vulnerable units of wind turbine based on improved VMD and DBN[J]. Journal of Vibration and Shock, 2019, 38(08): 153-160+179.
[5] 王朋飞, 刘长良, 徐健, 等. 基于图注意力和时间卷积网络的风电齿轮箱故障预警方法[J]. 电子测量与仪器学报, 2023, 37(08): 204-213.
WANG Pengfei, LIU Changliang, XU Jian, et al. Wind turbine gearbox fault warning method based on graph attention and temporal convolutional network[J]. Journal of Electronic Measurement and Instrumentation, 2023, 37(08): 204-213.
[6] 杨婷婷, 李浩千, 陈晓峰, 等. 基于WPT-Transformer的磨煤机故障预警研究[J]. 热力发电, 2023, 52(12): 180-189.
YANG Tingting, LI Haoqian, CHEN Xiaofeng, et al. Research on fault early warning of coal mill based on WPT and Transformer[J]. Thermal Power Generation, 2023, 52(12): 180-189.
[7] 张建亮, 李洋, 朱春山, 等. 基于双域稀疏Transformer的变电站设备故障预警方法[J]. 计算机科学, 2024, 51(05): 62-69.
ZHANG Jianliang, LI Yang, ZHU Chunshan, et al. Substation Equipment Malfunction Alarm Algorithm Based on Dual-domain Sparse Transformer[J]. Computer Science, 2024, 51(05): 62-69.
[8] WU X, FAN C, TANG J, et al. Forecast of global ionospheric TEC using an improved transformer model[J]. Advances in Space Research, 2024, 73(9): 4519-4538.
[9] Saman J, Mahin M, Saeed B, et al. Effective text classification using BERT, MTM LSTM, and DT[J]. Data & Knowledge Engineering, 2024, 151: 102306.
[10] QIN H, KANG S, TANG W, et al. Study of flow stress in Mg-Gd-Y-Nd-Zr alloys based on IWOA-BPNN model[J]. 2024, 30: 2848-2857.
[11] CHANG H Y, MOZAFARI S H, CHEN C, et al. PipeBERT: High-throughput BERT Inference for ARM Big.LITTLE Multi-core Processors[J]. Journal of Signal Processing Systems, 2023, 95(7): 877-894.
[12] YU Z, LEE F, CHEN Q. HCT-net: hybrid CNN-transformer model based on a neural architecture search network for medical image segmentation[J]. Applied Intelligence, 2023, 53(17): 19990-20006.
[13] 王伟平, 钟锐, 王青山. 基于改进鲸鱼算法的复合材料层合板优化设计[J]. 振动与冲击, 2023, 42(19): 294-300.
WANG Weiping, ZHONG Rui, WANG Qingshan. Optimization design of composite laminate based on improved whale optimization algorithm[J]. Journal of Vibration and Shock, 2023, 42(19): 294-300.
[14] 朱敏, 臧昭宇, 胥子豪, 等. 气动调节阀最优分数阶PID控制器设计[J]. 振动与冲击, 2022, 41(22): 267-274.
ZHU Min, ZANG Zhaoyu, XU Zihao, et al. Design of the optimal fractional order PID controller for a pneumatic control valve[J]. Journal of Vibration and Shock, 2022, 41(22): 267-274.
[15] 杨宇辰. 基于深度学习的燃煤电站磨煤机故障诊断技术研究[D]. 华北电力大学(北京), 2023.
YANG Yuchen. Research on fault diagnosis of coal mill in thermal power station based on Deep Learning[D]. North China Electric Power University (Beijing), 2023.
[16] 李晓彬, 牛玉广, 葛维春, 等. 基于改进堆叠自编码网络的电站辅机故障预警[J]. 仪器仪表学报, 2019, 40(06): 39-47.
LI Xiaobin, NIU Yuguang, GE Weichun, et al. Early fault warning of power plant auxiliary engine based on improved stacked autoencoder network[J]. Chinese Journal of Scientific Instrument, 2019, 40(06): 39-47.
[17] Hernando A, Galán-García JL, Aguilera-Venegas G. A novel way to build expert systems with infinite-valued attributes[J]. 2024, 9(2): 2938-2963.
[18] ZHANG H, LI Y, ZHAO Z, et al. An Online safety monitoring system of hydropower station based on expert system[J]. Energy Reports, 2022, 8: 1552-1567.