随着工业化程度的不断加深，异步电机广泛应用于工业生产行业，其中煤矿行业尤为突出。在煤矿行业中，电机的正常运行可以保证生产过程的高效、安全、便捷。因此，开展异步电机故障诊断技术的研究是目前故障诊断领域的重点。

本文以异步电机故障中常见的转子断条和偏心故障为例，对其故障机理进行分析，通过有限元仿真建立正常与故障状态(转子断条、气隙动静偏心)下的电机模型，提取相对应的定子电流信号，通过分析比较电流频谱，一方面验证提取定子电流数据的正确性；另一方面，为后续的特征提取方法提供可靠有效的数据支撑。

由于矿用异步电机运行过程中电流信号具有非线性、非平稳的特点，本文采用可调Q小波分维数提取电流信号的特征，相比于其他分维数这种方法更能表征信号的特性。经验模态分解同样适合处理非线性信号，但信号进行分解时会出现模态混叠等现象，这一现象极大可能导致诊断识别率降低。基于此，本文采用集合经验模态分解提取电流信号，通过仿真对比不同的集合经验模态分解熵值，发现能量熵作为特征集区分故障更加明显。随后采用支持向量机分类器识别故障，由于传统支持向量机分类器存在参数优化问题，因此将粒子群寻优和灰狼群寻优方法分别应用到支持向量机参数优化中，建立了粒子群参数优化支持向量机和灰狼群参数优化支持向量机故障诊断模型。在仿真实验中，将可调Q小波分维数特征集和集合经验模态分解能量熵特征集分别输入支持向量机分类器和迭代优化后的支持向量机分类器，进行全方面的对比分析，综合得出可调Q小波分维数特征结合灰狼群优化的支持向量故障诊断模型，其诊断识别率在这些方法中是最高的，诊断时间是最短的。

本文最终建立了一种基于可调Q小波分维数和灰狼群参数优化支持向量机的异步电机转子故障诊断新技术，并应用于工程实验中加以验证，得到了较高的故障识别率，且识别时间也较短。

With the continuous deepening of industrialization, asynchronous motors are widely used in industrial production industries, especially in the coal industry. In the coal industry, the normal operation of the motor can ensure high efficiency, safety and speed in the production process. Therefore, the research of asynchronous motor fault diagnosis technology is of great significance.

In this paper, the common rotor broken bar and eccentric faults in asynchronous motor faults are taken as examples to analyze the mechanism of asynchronous motor broken bar faults and eccentric faults. The motor models under normal and fault conditions are established through finite element simulation, and then the corresponding stator current signals are collected. By analyzing and comparing the current spectra under different conditions, on the one hand, it verifies the correctness of the stator current signal data extraction, on the other hand, it provides reliable and effective data support for the subsequent feature extraction methods.

Due to the non-linear and non-stationary characteristics of the current signal during the operation of the mine asynchronous motor, the emergence of fractal geometry opens up new ideas for feature signal extraction. Since wavelet transform and fractal geometry have self-similarity, the tunable Q wavelet fractal dimension is used to extract the characteristics of current signal, which can characterize the characteristics of the signal better than other fractal dimension extraction methods. The empirical mode decomposition is also suitable for processing nonlinear signals, but when the signal is decomposed, there will be phenomena such as modal aliasing, which may greatly reduce the diagnostic recognition rate. Based on this, the ensemble empirical mode decomposition is used to extract the current signal, and the entropy value of different ensemble empirical mode decomposition is compared through simulation, and it is found that energy entropy is more obvious as a feature set to distinguish faults. Then the support vector machine classifier was used to identify the fault, but the traditional support vector machine classifier had parameter optimization problems, so the particle swarm optimization and gray wolf swarm optimization methods were respectively applied to the support vector machine parameter optimization, and the particle swarm parameters were established Optimize the support vector machine and gray wolf pack parameter optimization support vector machine fault diagnosis model. In the simulation experiment, the tunable Q wavelet fractal dimension feature set and ensemble empirical mode decomposition energy entropy feature set were input into the support vector machine classifier and the support vector machine classifier after iterative optimization, respectively, to conduct a comprehensive comparative analysis. Comprehensively obtain the tunable Q wavelet fractal dimension feature, combined with the gray wolf pack optimization support vector fault diagnosis model, the diagnosis recognition rate is the highest among these methods, and the diagnosis time is the shortest.

Finally, a new rotor fault diagnosis technology based on the tunable Q wavelet fractal dimension and GWO-SVM is established, and it is applied to engineering experiments to verify it, and a high recognition rate is obtained.

[1]邵敏珠. 煤炭资源的现状及利用前景分析[J]. 科技传播, 2013, 5(24):85-86.

[2]梁艳萍, 柳杨, 王晨光. 大型水轮发电机定子绕组新型组合换位方法分析[J]. 中国电机工程学报, 2016, 36(11):3092-3100.

[3]周东华, 叶银忠. 现代故障诊断与容错控制(第一版)[M]. 北京:清华大学出版社，2000:1-157.

[4]万书亭, 吴美玲. 基于时域参数趋势分析的滚动轴承故障诊断[J]. 机械工程与自动化, 2010(3): 108-110.

[5]Pandarakone SE, Mizuno Y, Nakamura H. Distinct Fault Analysis of Induction Motor Bearing Using Frequency Spectrum Determination and Support Vector Machine[J]. IEEE Transactions on Industry Applications. 2017, 53(3):3049-3056.

[6]Wang T, Liang M, Li J, et al. Rolling element bearing fault diagnosis via fault characteristic order (FCO) analysis[J]. Mechanical Systems and Signal Processing, 2014, 45(1):139-153.

[7]Wang W, Huang Q, Zhang Y. On the motor fault diagnosis based on wavelet transform and ANN[C]// Control Conference. IEEE, 2012:5339-5346.

[8]刘翌辉, 贾普荣, 王强等. 小波变换在多级传动齿轮箱缺齿故障诊断中的应用[J]. 噪声与振动控制, 2018, 38(S2):303-307.

[9]Ren Z, Zhou S , Chunhui E, et al. Crack fault diagnosis of rotor systems using wavelet transforms[J]. Computers & Electrical Engineering, 2015, 45:33-41.

[10]许允之, 方磊, 张晓春等. 基于小波包变换的转子断条故障特征提取[J]. 煤矿机电, 2016(01):12-15+18.

[11]荆双喜, 赵行宇, 郭松涛等. 异步电机转子断条故障诊断研究[J]. 河南理工大学学报(自然科学版), 2016,35(02):224-229.

[12]魏艳鸣, 海本斋. 基于EMD和改进多重信号分类的感应电机故障检测方法[J]. 微特电机, 2017, 45(07):37-40+53.

[13]蔡艳平, 徐光华, 李艾华等. EMD局部积分均值增密插值改进算法及其在转子故障诊断中的应用[J]. 振动与冲击, 2016, 35(19):81-87.

[14]温华. 基于混沌与分形理论的电机故障诊断方法[J]. 山西焦煤科技, 2014, 38(12):25-28.

[15]许允之, 赵月南, 仝年等. 基于混沌理论的异步电机偏心故障诊断的研究[J]. 大电机技术, 2015(03):4-7+40.

[16]姚亚夫, 张星. 基于瞬时能量熵和 SVM 的滚动轴承故障诊断[J]. 电子测量与仪器学报, 2013, 27(10): 957-962.

[17]郑日晖, 岑健, 陈志豪等. 基于EMD样本熵与改进DS证据理论的故障诊断方法[J]. 自动化与信息工程, 2020,41(02):19-26.

[18]姬志伟, 刘军, 孙益祥等. 基于多方法联合的故障诊断专家系统设计[J]. 航空计算技术, 2015, 45(05): 93-96.

[19]郑秒, 钱虹, 林斯韵等. 核电站高压加热器管系泄漏智能诊断专家知识库的研究[J]. 核动力工程, 2018, 39(01):146-151.

[20]薛寒, 谢利理, 叶留义. 基于模糊推理的电机故障诊断专家系统研究[J].计算机测量与控制，2010, 18(01):8-10.

[21]王伟. 基于神经网络和D-S证据理论的通风机故障诊断[J]. 信息记录材料, 2018, 19(03):75-77.

[22]雷方涛, 王佩, 刘璐. 基于小波分析和BP神经网络的齿轮箱故障诊断技术[J]. 机电信息, 2017(27):77+79.

[23]苟旭丹. 基于Hilbert模量与改进BP神经网络的电机转子断条故障诊断[J]. 电测与仪表, 2018, 55(03):55-58.

[24]朱明飞, 陈兢, 方敏. 基于混合遗传算法与小波神经网络的电机转子断条故障诊断方法[J]. 农村电气化, 2019(06):9-14.

[25]王丽娟. 基于小波变换及SVM的电机故障诊断[J]. 防爆电机, 2014, 49(04):27-32.

[26]Zhu H, Hu J, Lei G, et al. Fault diagnosis of incipient broken rotor bars for squirrel-cage induction motor under continuous variable load condition[J]. Chinese Journal ofScientific Instrument, 2014, 35(7):1646-1653.

[27]张超, 陈建军, 郭迅. 基于EMD能量熵和支持向量机的齿轮故障诊断方法[J]. 振动与冲击, 2010, 29(10):216-220+261.

[28]Zhang X L, Chen W, Wang B J, et al. Intelligent fault diagnosis of rotating machinery using support vector machine with ant colony algorithm for synchronous feature selection and parameter optimization[J]. Neurocomputing, 2015, 167:260-279.

[29]黎康康. 基于小波分析的齿轮箱滚动轴承故障诊断[J]. 机车车辆工艺, 2017(06):37-39..

[30]袁杰, 王福利, 王姝等. 基于D-S融合的混合专家知识系统故障诊断方法[J]. 自动化学报, 2017, 43(09): 1580-1587.

[31]游子跃, 王宁, 李明明等.基于EEMD和BP神经网络的风机齿轮箱故障诊断方法[J]. 东北电力大学学报, 2015, 35(01):64-72.

[32]hu H, Hu J, Lei G, et al. Fault diagnosis of incipient broken rotor bars for squirrel-cage induction motor under continuous variable load condition[J]. Chinese Journal of Scientific Instrument, 2014, 35(07):1646-1653.

[33]贾朱植,徐建英,宋向金等.鼠笼电机转子断条故障的定子电流信号平方解调分析诊断方法[J].仪器仪表学报,2015,36(09):2097-2103.

[34]杨晓光, 许仪勋. 双馈风力发电机故障诊断方法研究[J]. 电测与仪表,2018, 55(2):52-58.

[35]Lee Prognostics and health machinery systems-Reviews, methodology and Systems & Signal Processing, management design for rotary applications[J]. Mechanical 2014, 42(1):314-334.

[36]贾朱植,杨理践,祝洪宇等.基于高分辨率谱估计的早期转子断条故障诊断[J].仪器仪表学报,2017,38(02):279-287.

[37]黄进, 牛发亮, 杨家强. 基于双PQ变换的感应电机转子故障诊断[J]. 中国电机工程学报, 2006, 13(3):135-140．

[38]孙丽玲, 王续, 许伯强. 基于SVD滤波技术与快速四阶累积量ESPRIT算法的异步电动机转子断条故障检测新方法[J]. 电工技术学报, 2015, 30(10):147-156.

[39]许伯强, 朱明飞. 基于精简四阶累积量MUSIC与混合遗传算法的笼型异步电动机转子断条故障检测新方法[J]. 电机与控制应用, 2016, 43(07):73-80+85.

[40]K YUN, WANG T Y, CHU F L. Adaptive TQWT filter based feature extraction method and its application to detection of repetitive transients[J]. Sci China Tech Sci, 2018, 61(10):1556-1574.

[41]Shen Y, Liu X, Yuan X. Fractal Dimension of Irregular Region of Interest Application to Corn Phenology Characterization[J]. IEEE Journal of Selected Topics in Applied Earth Observations & Remote Sensing, 2017, PP(99):1-11.

[42]Sanchez J P A, Alegria O C, Rodriguez M V, et al. Detection of ULF Geomagnetic Anomalies Associated to Seismic Activity Using EMD Method and Fractal Dimension Theory[J]. IEEE Latin America Transactions, 2017, 15(02):197-205.

[43]谢和平编译. 分形几何—数学基础与应用[M]. 重庆:重庆大学出版社, 1991.

[44]杨晓光, 许仪勋. 双馈风力发电机故障诊断方法研究[J]. 电测与仪表, 2018, 55(02):52-58.

[45]惠阿丽, 林辉. 高压绝缘子泄漏电流的线调频小波分形特征[J]. 高电压技术, 2010, 36(05): 1160-1166

[46]Hui A-li, Zheng Jianming, Lin Hui, et al. Wavelet-fractal characteristics of leakage current on HV insulators[C]//3rd International Conference on Deregulation and Restructuring and Power Technologies(DRPT' 2008). Nanjing, China: [s.n.], 2008: 732-736.

[47]Gao J, Shang P. Analysis of complex time serie based on EMD energy entropy plane[J]. Nonlinear Dynamics, 2019, 96(01):1-18.

[48]Varanis M, Pederiva R. Statements on wavelet packet energy-entropy signatures and filter influence in fault diagnosis of induction motor in non-stationary operations[J]. Journal of the Brazilian Society of Mechanical Sciences and Engineering, 2018, 40(02):98.

[49]Wang W, Wang D, Singh V P, et al. Evaluation of information transfer and data transfer models of rain-gauge network design based on information entropy[J]. Environmental Research, 2019, 178.

[50]王振浩, 顾欣然, 孙福军. 基于EMD近似熵的高压断路器故障诊断[J]. 高压电器, 2018, 54(10):151-156+163.

[51]Sokunbi M O, Cameron G G, Ahearn T S, et al. Fuzzy Approximate Entropy Analysis of Resting State FMRI Signal Complexity Across the Adult Life Span[J]. Medical Engineering and Physics, 2015, 37(11): 1082-1090.

[52]杨旭, 邱明, 陈立海, 陈月. 基于PSO-RWE的自适应小波阈值函数滚动轴承振动信号去噪方法[J]. 航空动力学报, 2020, 35(11):2339-2347.

[53]Thi Thom Hoang, Ming-Yuan Cho, Mahamad Nabab Alam, et al. A novel differential particle swarm optimization for parameter selection of support vector machines for monitoring metal-oxide surge arrester conditions[J]. Swarm and Evolutionary Computation, 2018, 38(02): 120-126.

[54]张雷. 基于粒子群算法的神经网络故障诊断方法研究[D]. 广东:华南理工大学, 2016.

[55]Hazlee Azil Illias, Xin Rui Chai, Ab Halim Abu Bakar. Hybrid modified evolutionary particle swarm optimisation-time varying acceleration coefficient-artificial neural, network for power transformer fault diagnosis[J]. Measurement, 2016, 90(04):94-102.

[56]吕明珠, 苏晓明, 陈长征等. 改进粒子群算法优化的支持向量机在滚动轴承故障诊断中的应用[J]. 机械与电子, 2019, 37(01):42-48.

[57]Seyedali Mirjalili, Seyed Mohammad Mirjalili, Andrew Lewis. Grey Wolf Optimizer[J]. Advances in Engineering Software, 2014, 69(03):46-61.

[58]涂强. 灰狼优化算法的改进及其在图像分割中的应用[D]. 河南:河南师范大学, 2017.

[59]邵珠超. 水泥熟料装置工艺故障诊断研究[D]. 长春:长春工业大学, 2018.

[60]左剑, 张程稳, 肖逸等 基于灰狼优化算法的多机电力系统稳定器参数最优设计[J]. 电网技术, 2017, 41(09): 2987-2995.

[61]郭阳,张涛,胡玉蝶,杜航.基于自适应头狼的灰狼优化算法[J].成都大学学报(自然科学版),2020,39(01):60-63+73.