带式输送机托辊故障检测方法

武国平

doi:10.13272/j.issn.1671-251x.2022100022

带式输送机托辊故障检测方法

doi: 10.13272/j.issn.1671-251x.2022100022

武国平

国家能源集团准能集团有限责任公司, 内蒙古鄂尔多斯　017000

基金项目: 国家自然科学基金项目（51905459）。

详细信息

作者简介:
武国平（1966—），男，内蒙古凉城人，教授级高级工程师，主要从事科技创新管理工作，E-mail:10570948@chnenergy.com

中图分类号: TD634
计量
- 文章访问数: 274
- HTML全文浏览量: 66
- PDF下载量: 43
- 被引次数: 0
出版历程
- 收稿日期: 2022-10-11
- 修回日期: 2023-02-08
- 网络出版日期: 2023-02-27

Fault detection method for belt conveyor idler

WU Guoping

CHN Energy Zhunneng Energy Group Co., Ltd., Ordos 017000, China

摘要

摘要: 针对现有输煤传送机托辊故障检测方法存在识别精度较低、抗环境干扰能力较差、无法长期稳定运行等问题，提出了一种基于融合信号（TFM）及多输入一维卷积神经网络（MI−1DCNN）的输煤传送机托辊故障检测方法。首先，通过拾音器采集输煤传送机沿线托辊运行的音频信号，采用dB4小波无偏风险估计阈值降噪法对信号进行预处理，消除背景噪声，提高信噪比。然后，对降噪音频信号的时域、频域和梅尔频率倒谱系数（MFCC）及其一阶二阶差分系数进行归一化处理，并进行拼接，得到特征TFM。最后，将TFM输入到多尺度卷积核的MI−1DCNN模型，在网络通道末端进行特征融合，通过Softmax函数完成对正常托辊和故障托辊的分类识别。以某煤矿实际采集的输煤传送机托辊音频信号样本对TFM−MI−1DCNN模型进行试验，结果表明：故障托辊平均识别准确率达98.65%，较改进小波阈值降噪−反向传播−径向基函数网络、MFCC−K 邻近方法−支持向量机的平均识别准确率分别提高了1.50%和1.03%。现场应用结果表明：该方法下故障托辊平均识别准确率为98.4%，说明该方法适用于现场应用。
- 输煤传送机 /
- 智能巡检机器人 /
- 托辊 /
- 音频信号 /
- 小波阈值降噪 /
- MFCC /
- 多输入一维卷积神经网络
Abstract: The existing fault detection methods for belt conveyor idler have the problems of low recognition precision, poor anti-interference capability and inability to operate stably over a long period of time. In order to solve the above problems, a fault detection method for belt conveyor idler based on time-frequency-MFCC(TFM) and multi-input one-dimensional convolutional neural network (MI-1DCNN) is proposed. Firstly, the pickup collects the audio signal of the coal conveyor idler running along the line. The dB 4 wavelet unbiased risk estimation threshold noise reduction method is used to preprocess the signal to eliminate the background noise and improve the signal-to-noise ratio. Secondly, the time domain, frequency domain and Mel frequency cepstrum coefficient (MFCC), and the first and second order difference coefficient of the noise reduction audio signal are normalized respectively, and finally assembled to obtain the feature TFM. Finally, that TFM signals are input into a MI-1DCNN model with a multi-scale convolution kernel. The feature fusion is carried out at the end of a network channel. The classification and identification of the normal idler and the fault idler are completed through a Softmax function. The TFM-MI-1DCNN model is tested with the audio signal samples of coal conveyor idler collected in a coal mine. The results show that the average recognition accuracy of the fault idler is 98.65%. The average recognition accuracy is improved by 1.50% and 1.03% compared to the improved wavelet threshold denoising-backpropagation-radial basis function network and MFCC-K-nearest neighbor algorithm-support vector machine. The false detection rate is only 0.194%. The results of field application show that the average dentification accuracy of the proposed method is 98.4%, indicating that the proposed method is suitable for field application.
- belt conveyor /
- intelligent inspection robot /
- idler /
- audio signal /
- wavelet threshold denoising /
- MFCC /
- multi-input one-dimensional convolutional neural network

HTML全文

图 1 基于TFM及MI−1DCNN的输煤传送机托辊故障诊断流程

Figure 1. Fault diagnosis process of belt conveyor idle based on time-frequency-MFCC and multi-input one-dimensional convolutional neural network

下载: 全尺寸图片幻灯片

图 2 小波阈值降噪过程

Figure 2. The wavelet threshold denoising process

下载: 全尺寸图片幻灯片

图 3 MFCC特征提取流程

Figure 3. MFCC feature extraction process

下载: 全尺寸图片幻灯片

图 4 MI−1DCNN模型结构

Figure 4. MI-1DCNN model structure

下载: 全尺寸图片幻灯片

图 5 小波阈值降噪结果

Figure 5. Wavelet threshold denoising results

下载: 全尺寸图片幻灯片

图 6 故障托辊帧数、维度与MFCC关系

Figure 6. Relationship among frame number, dimension and MFCC of fault idler

下载: 全尺寸图片幻灯片

图 7 正常托辊帧数、维度与MFCC关系

Figure 7. Relationship among frame number, dimension and MFCC of normal idler

下载: 全尺寸图片幻灯片

图 8 TFM

Figure 8. Time-Frequency-MFCC

下载: 全尺寸图片幻灯片

图 9 现场测试流程

Figure 9. Field test process

下载: 全尺寸图片幻灯片

表 1 不同方法识别结果

Table 1. Identification results of different methods

方法	识别类型	识别准确率/%	平均识别准确率/%
改进小波阈值降噪−BP−RBF	正常托辊	98.90	97.15
改进小波阈值降噪−BP−RBF	故障托辊	95.40	97.15
MFCC−KNN−SVM	正常托辊	99.25	97.62
MFCC−KNN−SVM	故障托辊	96.00	97.62
本文方法	正常托辊	99.93	98.65
本文方法	故障托辊	97.38	98.65

下载: 导出CSV

表 2 机器人现场巡检测试结果

Table 2. Test results of robot on-site inspection

巡检日期	巡检总托辊数/组	人工巡检真实故障数/组	方法报出故障		故障托辊识别准确率/%
巡检日期	巡检总托辊数/组	人工巡检真实故障数/组	真实故障数/组	误报故障数/组	故障托辊识别准确率/%
第1周	12796	36	35	27	97.2
第2周	12121	43	42	28	97.6
第3周	14063	39	39	26	100.0
第4周	11259	36	35	21	97.2
第5周	12653	38	38	22	100.0
现场测试故障托辊平均识别准确率 /%					98.4

下载: 导出CSV

参考文献(21)

[1]	蒋卫良,韩东劲. 我国煤矿带式输送机现状与发展趋势[J]. 煤矿机电,2008,39(1):1-6. JIANG Weiliang,HAN Dongjin. Development situation and tendency of native mining belt conveyor[J]. Colliery Mechanical & Electrical Technology,2008,39(1):1-6.
[2]	杨孝新,刘婷婷,刘福明,等. 露天煤矿带式输送机智能巡检机器人[J]. 露天采矿技术,2022,37(4):48-51. YANG Xiaoxin,LIU Tingting,LIU Fuming,et al. Intelligent inspection robots for belt conveyors in open-pit coal mines[J]. Opencast Mining Technology,2022,37(4):48-51.
[3]	周立辉,张永生,孙勇,等. 智能变电站巡检机器人研制及应用[J]. 电力系统自动化,2011,35(19):85-88,96. ZHOU Lihui,ZHANG Yongsheng,SUN Yong,et al. Development and application of equipment inspection root for smart substations[J]. Automation of Electric Power Systems,2011,35(19):85-88,96.
[4]	裴文良,张树生,李军伟. 矿用巡检机器人设计及其应用[J]. 制造业自动化,2017,39(2):73-74,94. PEI Wenliang,ZHANG Shusheng,LI Junwei. The design and application of inspection robot for mine[J]. Manufacturing Automation,2017,39(2):73-74,94.
[5]	朱剑锋. 煤矿带式输送机输送带撕裂红外检测系统设计[J]. 新型工业化,2017,7(7):58-61. ZHU Jianfeng. Design conveyor belt tear infrared detection system of coal mine ribbon conveyer[J]. The Journal of New Industrialization,2017,7(7):58-61.
[6]	韩涛,胡英贝,张蕾,等. 信息融合技术在托辊轴承故障诊断中的应用[J]. 轴承,2012,508(6):57-59. HAN Tao,HU Yingbei,ZHANG Lei,et al. Application of information fusion technology in fault diagnosis of roller bearings[J]. Bearing,2012,508(6):57-59.
[7]	孙维,刁冬梅. 基于φ−OTDR技术的带式输送机托辊故障检测[J]. 工矿自动化,2016,42(8):9-12. SUN Wei,DIAO Dongmei. Roller fault detection of belt conveyor based on φ-OTDR technology[J]. Industry and Mine Automation,2016,42(8):9-12.
[8]	曹贯强. 带式输送机托辊故障检测方法[J]. 工矿自动化,2020,46(6):81-86. CAO Guanqiang. Fault detection method for belt conveyor roller[J]. Industry and Mine Automation,2020,46(6):81-86.
[9]	郝洪涛,苏耀瑞,丁文捷,等.托辊非接触式故障识别方法研究[J/OL].机械科学与技术:1-8[2022-08-25]. DOI: 10.13433/j.cnki.1003-8728.20220008. HAO Hongtao,SU Yaorui,DING Wenjie,et al.Study on the method of non-contact fault identification of rollers[J/OL]. Mechanical Science and Technology for Aerospace Engineering:1-8[2022-08-25]. DOI:10.13433/j.cnki.1003-8728.20220008.
[10]	伊鑫,杨明锦,杨林顺,等. 基于KNN与SVM两级综合健康指标的托辊故障诊断方法[J]. 选煤技术,2020(5):94-102. YI Xin,YANG Mingjin,YANG Linshun,et al. The KNN and SVM-based 2-level comprehensive health indicators diagnosis method for detecting the failure of belt conveyor's idlers[J]. Coal Preparation Technology,2020(5):94-102.
[11]	蔡安江,李涛,王洪波,等. 带式输送机故障准确诊断方法[J]. 金属矿山,2020,48(4):130-134. CAI Anjiang,LI Tao,WANG Hongbo,et al. Accurate diagnosis method for belt conveyor fault[J]. Metal Mine,2020,48(4):130-134.
[12]	葛江华,刘奇,王亚萍,等. 支持张量机与KNN−AMDM决策融合的齿轮箱故障诊断方法[J]. 振动工程学报,2018,31(6):1093-1101. GE Jianghua,LIU Qi,WANG Yaping,et al. Fault diagnosis method of gearbox supporting tension machine and KNN-AMDM decision fusion[J]. Journal of Vibration Engineering,2018,31(6):1093-1101.
[13]	ZHAO Ruimei, CUI Huimin. Improved threshold denoising method based on wavelet transform[C]. 7th International Conference on Modelling, Identification and Control （ICMIC）, Sousse, 2015: 1354-1359.
[14]	蔡铁,朱杰. 小波阈值降噪算法中最优分解层数的自适应选择[J]. 控制与决策,2006,21(2):217-220. CAI Tie,ZHU Jie. Adaptive selection of optimal decomposition level in threshold de-noising algorithm based on wavelet[J]. Control and Decision,2006,21(2):217-220.
[15]	王维,张英堂,任国全. 小波阈值降噪算法中最优分解层数的自适应确定及仿真[J]. 仪器仪表学报,2009,30(3):526-530. WANG Wei,ZHANG Yingtang,REN Guoquan. Adaptive selection and simulation of optimal decomposition level in threshold de-noising algorithm based on wavelet transform[J]. Chinese Journal of Scientific Instrument,2009,30(3):526-530.
[16]	李虹,徐小力,吴国新,等. 基于MFCC的语音情感特征提取研究[J]. 电子测量与仪器学报,2017,31(3):448-453. LI Hong,XU Xiaoli,WU Guoxin,et al. Research on speech emotion feature extraction based on MFCC[J]. Journal of Electronic Measurement and Instrumentation,2017,31(3):448-453.
[17]	吕霄云. 基于MFCC和GMM的异常声音识别算法研究[D]. 成都: 西南交通大学, 2010. LYU Xiaoyun. Research on abnormal audio recognition algorithm based on MFCC and GMM[D]. Chengdu: Southwest Jiaotong University, 2010.
[18]	曲建岭,余路,袁涛,等. 基于一维卷积神经网络的滚动轴承自适应故障诊断算法[J]. 仪器仪表学报,2018,39(7):134-143. doi: 10.19650/j.cnki.cjsi.J1803286 QU Jianling,YU Lu,YUAN Tao,et al. Adaptive fault diagnosis algorithm for rolling bearings based on one-dimensional convolutional neural network[J]. Chinese Journal of Scientific Instrument,2018,39(7):134-143. doi: 10.19650/j.cnki.cjsi.J1803286
[19]	吴春志,江鹏程,冯辅周,等. 基于一维卷积神经网络的齿轮箱故障诊断[J]. 振动与冲击,2018,37(22):51-56. WU Chunzhi,JIANG Pengcheng,FENG Fuzhou,et al. Faults diagnosis method for gearboxes based on a 1-D convolutional neural network[J]. Journal of Vibration and Shock,2018,37(22):51-56.
[20]	叶壮,余建波. 基于多通道一维卷积神经网络特征学习的齿轮箱故障诊断方法[J]. 振动与冲击,2020,39(20):55-66. YE Zhuang,YU Jianbo. Gearbox fault diagnosis based on feature learning of multi-channel one-dimensional convolutional neural network[J]. Journal of Vibration and Shock,2020,39(20):55-66.
[21]	黄华,姚嘉靖,王永和,等. 基于多通道一维卷积神经网络的刀具磨损动态预测模型[J]. 振动与冲击,2023,42(2):60-67. HUANG Hua,YAO Jiajing,WANG Yonghe,et al. Dynamic prediction model for tool wear based on a multi-channel one-dimensional convolutional neural network[J]. Journal of Vibration and Shock,2023,42(2):60-67.