基于改进EWT-多尺度熵和KELM的球磨机负荷识别方法

doi:10.11949/0438-1157.20190811

摘要/Abstract

摘要：

针对球磨机在磨矿过程中负荷靠经验难以准确判断的问题，提出了一种基于改进的经验小波变换(empirical wavelet transform, EWT)-多尺度熵和核极限学习机(KELM)的球磨机负荷识别方法。首先，针对筒体振动信号的多样性和复杂性特点，对EWT频谱分割方法进行改进，通过构建信号仿真模型，比较EWT、EMD的分解效果，证明该方法的有效性。再将不同负荷状态下的筒体振动信号用改进的EWT算法进行分解得到内禀模态函数(intrinsic mode function, IMF)，接着，对分解后的IMF分量进行相关性分析得到敏感分量进行重构；最后，将重构信号的多尺度熵作为表征磨机不同负荷状态的特征向量，并计算多尺度熵偏均值。结果表明：三种负荷信号的多尺度熵及多尺度熵偏均值都存在明显的差异，关系表现为:欠负荷>正常负荷>过负荷。将提取的多维特征向量进行归一化处理并作为KELM的输入，磨机负荷状态作为输出，利用核排列(kernel target alignment, KTA)算法优化核参数，建立磨机负荷状态识别最优模型；通过磨矿实验验证了方法的可行性，相比SVM整体识别率提高了3.4%，且对于EMD-多尺度熵、EWT-多尺度熵分别提高了12.3%、8.9%。

关键词: 磨机负荷, 经验小波变换, 优化, KELM, 计算机模拟, 模型预测控制

Abstract:

Aiming at the problem that the load of the ball mill is difficult to accurately judge during the grinding process, a ball mill load identification method based on improved empirical wavelet transform (EWT)-multi-scale entropy and kernel extreme learning machine (KELM) is proposed. Firstly, according to the diversity and complexity of cylinder vibration signal, the EWT spectrum segmentation method is improved. By constructing the signal simulation model, the decomposition effect of EWT, EMD is compared, and the effectiveness of the method is proved. Secondly, the intrinsic modal function (IMF) is obtained by using the improved EWT algorithm to decompose the vibration signal of the cylinder under different load states, and then the effective IMF component is selected for reconstruction by correlation analysis. Finally, the multi-scale entropy of the reconstructed signal is extracted as the eigenvector to characterize the different load states of the mill, and the mean value of multi-scale entropy deviation is calculated. The results show that there are obvious differences in the mean value of multi-scale entropy and multi-scale entropy of the three load signals, and the relationship between the three load signals is underload > normal load > overload. The extracted multidimensional feature vector is normalized and used as the input of KELM and the load state of the mill is used as the output. The kernel target alignment (KTA) algorithm is used to optimize the kernel parameters and the optimal model of mill load state identification is established. The feasibility of the method is verified by grinding experiments. Compared with SVM, the overall recognition rate of EWT- is 3.4% higher, and for EMD-multi-scale entropy, EWT-multi-scale entropy is increased by 12.3% and 8.9%, respectively.

Key words: mill load, EWT, optimal, KELM, computer simulation, model-predictive control

中图分类号:

TP 29

罗小燕, 戴聪聪, 程铁栋, 蔡改贫, 刘鑫, 刘吉顺. 基于改进EWT-多尺度熵和KELM的球磨机负荷识别方法[J]. 化工学报, 2020, 71(3): 1264-1277.

Xiaoyan LUO, Congcong DAI, Tiedong CHENG, Gaipin CAI, Xin LIU, Jishun LIU. Load identification method of ball mill based on improved EWT multi-scale entropy and KELM[J]. CIESC Journal, 2020, 71(3): 1264-1277.

图/表 21

图1 KTA-KELM磨机负荷状态识别模型建立流程图

Fig.1 Flow chart of KTA-KELM mill load state identification model

图2 仿真信号x(t)的时域波形

Fig.2 Time domain waveform of simulation signalx(t)

图3 改进EWT频谱分割图及相应的小波滤波器组

Fig.3 Improvement of EWT spectrum segmentation map and corresponding wavelet filter banks

图4 传统EWT的频谱分割图及相应的小波滤波器组

Fig.4 Spectrum segmentation of traditional EWT and corresponding wavelet filter banks

图5 改进EWT和传统EWT分解结果(红色虚线代表原始信号，蓝色实线代表分解结果)

Fig.5 Improved EWT and traditional EWT decomposition results

图6 EMD分解结果

Fig.6 EMD decomposition results

图7 实验装置

Fig.7 Experimental device diagram

图8 原始信号

Fig.8 Original signal

图9 IMF分量与原始信号的相关系数

Fig.9 Correlation coefficient between IMF component and original signal

图10 重构信号

Fig.10 Reconstructed signal

表1 不同负荷信号去噪效果比较

Table 1 Comparison of denoising effects of different load signals

负荷状态	原始信号SNR/dB	重构信号SNR/dB
欠负荷	9.23	26.68
正常负荷	10.78	24.37
过负荷	9.11	27.41

图11 欠负荷筒体振动信号重构

Fig.11 Vibration signal reconstruction of under-loaded cylinder

表2 不同算法去噪后的信噪比

Table 2 Signal to noise ratio after denoising with different algorithms

算法	信噪比SNR/dB
EMD	12.65
EWT	18.47
改进EWT	26.68

表3 3种负荷状态振动信号的样本熵值

Table 3 Sample entropy values of vibration signals in three load states

数据样本	欠负荷	正常负荷	过负荷
A₁	0.1031	0.0663	0.0442
A₂	0.1017	0.0554	0.0561
A₃	0.0918	0.0535	0.0473
A₄	0.1003	0.0664	0.0531
A₅	0.0974	0.0631	0.0514
均值	0.09886	0.0609	0.05042

图12 不同负荷下振动信号多尺度熵随不同数据长度N变化

Fig.12 Variation of multi-scale entropy of vibration signal with different data lengthN under different loads

图13 不同负荷下振动信号多尺度熵随相似容限r变化

Fig.13 Multi-scale entropy of vibration signal varies with similar tolerancer under different loads

图14 不同算法下三种负荷状态重构信号的多尺度熵值变化

Fig.14 Multiscale entropy variation of three load state reconstruction signals with different algorithms

图15 不同算法下三种负荷状态重构信号的多尺度熵偏均值

Fig.15 Multiscale entropy partial means of three load state reconstruction signals with different algorithms

图16 KTA-KELM和KELM算法稳定性对比

Fig.16 Stability comparison of KTA-KELM and KELM algorithms

图17 负荷识别结果

Fig.17 Load identification results

表4 不同特征提取算法磨机负荷识别结果

Table 4 Recognition results of mill load based on different feature extraction algorithms

特征提取算法	球磨机不同负荷状态识别率/%			总体识别率/%
特征提取算法	欠负荷	正常负荷	过负荷	总体识别率/%
EMD-多尺度熵	86.7	83.3	83.3	84.4
EWT-多尺度熵	90	86.7	86.7	87.8
改进EWT-多尺度熵	100	93.3	96.7	96.7

参考文献 30

1	蔡改贫,宗路,罗小燕,等.基于CEEMDAN-云模型特征熵和LSSVM的磨机负荷预测研究[J].振动与冲击,2019,38(7):128-133.
	Cai G P,Zong L,Luo X Y,et al.Research on mill load forecasting based on CEMDAN-cloud model characteristic entropy and LSSVM[J].Vibration and Impact,2019,38(7):128-133.
2	杜永贵,李思思,阎高伟,等.基于流形正则化域适应湿式球磨机负荷参数软测量[J].化工学报,2018,69(3):1244-1251.
	Du Y G,Li S S,Yan G W,et al.Soft measurement of load parameters of wet ball mill based on Manifold Regularization domain[J].CIESC Journal,2018,69(3):1244-1251.
3	白锐,柴天佑.基于数据融合与案例推理的球磨机负荷优化控制[J].化工学报,2009,60(7):1746-1752.
	Bai R,Chai T Y.Ball mill load optimization control based on data fusion and case-based reasoning[J].CIESC Journal,2009,60(7):1746-1752.
4	胡显能,蔡改贫,罗小燕,等.基于CEEMDAN和多尺度排列熵的球磨机负荷识别方法[J].噪声与振动控制,2018,38(3):146-151.
	Hu X N,Cai G P,Luo X Y,et al.Load identification method of ball mill based on CEMDAN and multi-scale permutation entropy[J].Noise and Vibration Control,2018,38(3):146-151.
5	阎高伟,龚杏雄,续欣莹.基于云模型的球磨机料位概念表示与测量模型[J].中国电机工程学报,2014,34(14):2281-2288.
	Yan G W,Gong X X,Xu X Y.Conceptual representation and measurement model of ball mill level based on cloud model[J].Chinese Journal of Electrical Engineering,2014,34(14):2281-2288.
6	汤健,柴天佑,丛秋梅,等.选择性融合多尺度筒体振动频谱的磨机负荷参数建模[J].控制理论与应用,2015,32(12):1582-1591.
	Tang J,Chai T Y,Cong Q M,et al.Modeling of mill load parameters based on selective fusion of multi-scale cylinder vibration spectrum[J].Control Theory and Application,2015,32(12):1582-1591.
7	赵立杰,柴天佑,汤健,等.基于EMD和选择性集成学习算法的磨机负荷参数软测量[J].自动化学报,2014,40(9):1853-1866.
	Zhao L J,Chai T Y,Tang J,et al.Soft measurement of mill load parameters based on EMD and selective integrated learning algorithm[J].Journal of Automation,2014,40(9):1853-1866.
8	蔡改贫,宗路,刘鑫,等.基于MEEMD-多尺度分形盒维数和ELM的球磨机负荷识别方法[J].化工学报,2019,70(2):764-771.
	Cai G P,Zong L,Liu X,et al.Ball mill load identification method based on MEEMD-multiscale fractal box dimension and ELM[J].CIESC Journal,2019,70(2):764-771.
9	罗小燕,卢小江,熊洋,等.小波分析球磨机轴承振动信号特征提取方法[J].噪声与振动控制,2016,36(1):148-152.
	Luo X Y,Lu X J,Xiong Y,et al.Wavelet analysis method for feature extraction of bearing vibration signal of ball mill[J].Noise and Vibration Control,2016,36(1):148-152.
10	Gilles J.Empirical wavelet transform[J].IEEE Transactions on Signal Processing,2013,61(16):3999-4010.
11	Huimin Z,Shaoyan Z,Jian F,et al.Study on a motor bearing fault diagnosis method using improved EWT based on scale space threshold method[J].International Journal of Emerging Electric Power Systems,2018,19(4):1-13.
12	肖启阳,李健,孙洁娣,等.基于EWT及模糊相关分类器的管道微小泄漏检测[J].振动与冲击,2018,37(14):122-129.
	Xiao Q Y,Li J,Sun J D,et al.Pipeline micro-leakage detection based on EWT and fuzzy correlation classifier[J].Vibration and Impact,2018,37(14):122-129.
13	He Y,Hongru L,Yaolong L,et al.A novel improved full vector spectrum algorithm and its application in multi-sensor data fusion for hydraulic pumps[J].Measurement,2018,133(1):145-161.
14	祝文颖,冯志鹏.基于改进经验小波变换的行星齿轮箱故障诊断[J].仪器仪表学报,2016,37(10):2193-2201.
	Zhu W Y,Feng Z P.Fault diagnosis of planetary gearbox based on improved empirical wavelet transform[J].Chinese Journal of Science Instrument,2016,37(10):2193-2201.
15	辛玉,李舜酩,王金瑞,等.基于迭代经验小波变换的齿轮故障诊断方法[J].仪器仪表学报,2018,39(11):79-86.
	Xin Y,Li S Q,Wang J R,et al.Gear fault diagnosis method based on iterative empirical wavelet transform[J].Chinese Journal of Science Instrument,2018,39(11):79-86.
16	Merainani B,Rahmoune C,Djamel B,et al.A novel gearbox fault feature extraction and classification using Hilbert empirical wavelet transform, singular value decomposition, and SOM neural network[J].Journal of Vibration and Control,2018,24(12):2512-2531.
17	高正.基于EWT和特征融合的钻机钻杆故障识别研究[D].杭州:浙江大学,2019.
	Gao Z.Research on drill pipe fault identification based on EWT and feature fusion [D].Hangzhou:Zhejiang University,2019.
18	He Q,Wang Y W,Du S,et al.Motor imagery based on adaptive parameter less empirical wavelet transform and selective integrated classification[J].Acta Physica Sinica,2018,67(11):185-196.
19	Costa M,Goldberger A L,Peng C K.Multiscale entropy analysis of complex physiologic time series[J].Physical Review Letters,2002,89(6):1-18.
20	Begum S,Barua S,Filla R.Classification of physiological signals for wheel loader operators using multi-scale entropy analysis and case-based reasoning[J].Expert Systems with Application,2014,41(2):295-305.
21	Liu Z,Chai T Y,Yu W,et al.Multi-frequency signal modeling using empirical mode decomposition and PCA with application to mill load estimation[J].Neurocomputing,2015,69(23):392-402.
22	苟先太,李昌喜,金炜东.VMD多尺度熵用于高速列车横向减振器故障诊断[J].振动.测试与诊断,2019,39(2):292-297+442.
	Gou X T,Li C X,Jin W D.VMD multiscale entropy for fault diagnosis of lateral shock absorber of high speed train[J].Vibration. Testing and Diagnosis,2019,39(2):292-297+442.
23	李从志,郑近德,潘海洋,等.基于精细复合多尺度散布熵与支持向量机的滚动轴承故障诊断方法[J/OL].中国机械工程[2019-07-09].
	Li C Z,Zheng J D,Pan H Y,et al.Rolling bearing fault diagnosis method based on fine composite multiscale dispersion entropy and support vector machine[J/OL].China Mechanical Engineering: [2019-07-09].
24	Huang G B,Zhou H,Ding X,et al.Extreme learning machine for regression and multiclass classification[J].IEEE Transactions on Systems,Man, andCybernetics-Part B: Cybernetics,2012,42(2):513-529.
25	范永东.模型选择中的交叉验证方法综述[D].太原:山西大学,2013.
	Fan Y D.Overview of cross-validation methods in model selection[D].Taiyuan:Shanxi University,2013.
26	章勇高,高彦丽,马迪.基于GA-KELM的光伏短期出力预测研究[J].控制工程,2018,25(7):1155-1159.
	Zhang Y G,Gao Y L,Ma D.Study on short-term photovoltaic output prediction based on GA-KELM[J].Control Engineering,2018,25(7):1155-1159.
27	张文涛,马永光,董子健,等.基于粒子群算法优化核极限学习机的磨煤机故障诊断[J].电力科学与工程,2018,34(9):54-58.
	Zhang W T,Ma Y G,Dong Z J,et al.Fault diagnosis of coal mill based on particle swarm optimization for nuclear limit learning machine[J].Electric Power Science and Engineering,2018,34(9):54-58.
28	王裴岩,蔡东风.一种基于核距离的核函数度量方法[J].计算机科学,2014,41(2):72-75.
	Wang P Y,Cai D F.A kernel function measurement method based on kernel distance[J].Computer Science,2014,41(2):72-75.
29	王建国,杨柳,张文兴.核排列优化的支持向量机在齿轮故障诊断中的应用[J].机械设计与制造,2018, (S1):37-40.
	Wang J G,Yang L,Zhang W X.Application of support vector machine with kernel arrangement optimization in gear fault diagnosis[J].Mechanical Design and Manufacture,2018, (S1):37-40.
30	Albert A P,Nii A O.A criterion for selecting relevant intrinsic mode functions in empirical mode decommposition[J].Advances in Adaptive Data Analysis,2010,2(1):1-24.

[1]	杨欣, 王文, 徐凯, 马凡华. 高压氢气加注过程中温度特征仿真分析[J]. 化工学报, 2023, 74(S1): 280-286.
[2]	刘远超, 关斌, 钟建斌, 徐一帆, 蒋旭浩, 李耑. 单层XSe₂（X=Zr/Hf）的热电输运特性研究[J]. 化工学报, 2023, 74(9): 3968-3978.
[3]	何松, 刘乔迈, 谢广烁, 王斯民, 肖娟. 高浓度水煤浆管道气膜减阻两相流模拟及代理辅助优化[J]. 化工学报, 2023, 74(9): 3766-3774.
[4]	陈哲文, 魏俊杰, 张玉明. 超临界水煤气化耦合SOFC发电系统集成及其能量转化机制[J]. 化工学报, 2023, 74(9): 3888-3902.
[5]	齐聪, 丁子, 余杰, 汤茂清, 梁林. 基于选择吸收纳米薄膜的太阳能温差发电特性研究[J]. 化工学报, 2023, 74(9): 3921-3930.
[6]	陈国泽, 卫东, 郭倩, 向志平. 负载跟踪状态下的铝空气电池堆最优功率点优化方法[J]. 化工学报, 2023, 74(8): 3533-3542.
[7]	刘文竹, 云和明, 王宝雪, 胡明哲, 仲崇龙. 基于场协同和耗散的微通道拓扑优化研究[J]. 化工学报, 2023, 74(8): 3329-3341.
[8]	邢雷, 苗春雨, 蒋明虎, 赵立新, 李新亚. 井下微型气液旋流分离器优化设计与性能分析[J]. 化工学报, 2023, 74(8): 3394-3406.
[9]	张曼铮, 肖猛, 闫沛伟, 苗政, 徐进良, 纪献兵. 危废焚烧处理耦合有机朗肯循环系统工质筛选与热力学优化[J]. 化工学报, 2023, 74(8): 3502-3512.
[10]	诸程瑛, 王振雷. 基于改进深度强化学习的乙烯裂解炉操作优化[J]. 化工学报, 2023, 74(8): 3429-3437.
[11]	吴文涛, 褚良永, 张玲洁, 谭伟民, 沈丽明, 暴宁钟. 腰果酚生物基自愈合微胶囊的高效制备工艺研究[J]. 化工学报, 2023, 74(7): 3103-3115.
[12]	汤晓玲, 王嘉瑞, 朱玄烨, 郑仁朝. 基于Pickering乳液的卤醇脱卤酶催化合成手性环氧氯丙烷[J]. 化工学报, 2023, 74(7): 2926-2934.
[13]	徐野, 黄文君, 米俊芃, 申川川, 金建祥. 多源信息融合的离心式压缩机喘振诊断方法[J]. 化工学报, 2023, 74(7): 2979-2987.
[14]	文兆伦, 李沛睿, 张忠林, 杜晓, 侯起旺, 刘叶刚, 郝晓刚, 官国清. 基于自热再生的隔壁塔深冷空分工艺设计及优化[J]. 化工学报, 2023, 74(7): 2988-2998.
[15]	邵远哲, 赵忠盖, 刘飞. 基于共同趋势模型的非平稳过程质量相关故障检测方法[J]. 化工学报, 2023, 74(6): 2522-2537.