作者：田园1, 潘宏侠1,2, 陈玉青1, 潘龙1
作者单位：1. 中北大学机械与动力工程学院, 山西 太原 030051;
2. 中北大学系统辨识与诊断技术研究所, 山西 太原 030051
关键词：自动机;局部窄带分解;广义分形;特征提取;支持向量机
摘要：
针对某型高射机枪自动机振动信号低信噪比、干扰多的特点,提出利用S.L.Peng的局部窄带分解理论对信号进行分解和重构,并用支持向量机对故障模式进行识别。通过对自动机故障机理分析,找到易发生故障的位置,并设置3种故障后进行振动信号采集。将信号通过基于局部窄带信号的分解和重构后通过广义维数计算获得各种工况的盒维数、信息维数、关联维数、广义分形维数谱均值,将其供给支持向量机进行故障分类。所得诊断结果准确率达93.75%,具有一定的参考及实用价值。
Automaton fault diagnosis based on S.L.Peng local narrow-band decomposition and generalized fractal theory
TIAN Yuan1, PAN Hongxia1,2, CHEN Yuqing1, PAN Long1
1. School of Mechanical and Power Engineering, North University of China, Taiyuan 030051;
2. System Identification and Diagnosis Technology Research Institute, North University of China, Taiyuan 030051, China
Abstract: As the vibration signals of a certain type of antiaircraft gun automatons are featured by low signal-to-noise ratio(SNR) and multi-disturbances, a S.L.Peng-based local narrow-band decomposition method has been proposed to decompose and reconstruct the signals. Particularly, a support vector machine(SVM) has been used to identify the failure mode. First, the failure mechanism of the automaton was analyzed to find the location prone to failures and the vibration signals were collected after three kinds of failures were set. Second, the signals were decomposed and reconstructed by means of local narrow-band signal decomposition. Third, the box dimension, information dimension, correlation dimension, and the mean average of generalized fractal dimension spectrum were obtained and put into the SVM to classify the failure. The accuracy rate of the diagnosis is as high as 93.75%, which proves that this method has some reference and practical value.
Keywords: local narrow-band decomposition;generalized fractal;feature extraction;SVM
2016, 42(2): 100-104 收稿日期: 2015-8-21;收到修改稿日期: 2015-10-30
基金项目: 国家自然科学基金(51175480)
作者简介: 田 园(1991-),男,山西太原市人,硕士研究生,专业方向为信号识别与处理、装备系统检测与诊断。
参考文献
[1] 康郦,冯德朝. 某中口径舰炮自动机设计思想与特点分析[J].火炮发射与控制学报,2010(2):61-65.
[2] 冯辅周,张超省,张丽霞,等. 红外热波技术在装甲装备故障诊断和缺陷检测中的应用[J]. 应用光学,2012(5):827-831.
[3] RENSHA W, CHEN J C, HOLLAND S D, et al. The sources of heat generation in vibrothermography[J]. NDT&E International,2011(44):736-739.
[4] 潘宏侠,崔云鹏,王海瑞. 基于混沌理论的自动机故障诊断研究[J]. 火炮发射与控制学报,2014(2):50-54.
[5] 张向前,关小伟,陈力子. BP神经网络在武器火控系统故障诊断中的应用[J]. 计算机测量与控制,2010(11):2576-2578.
[6] 刘春林,潘宏侠,史斐娜,等. 基于MF-DFA和SVM的齿轮箱故障诊断[J]. 机械传动,2014(12):119-123.
[7] FERNÁNDEZz F D, MARTÍNEZ R D, FONTENLA R O, et al. Automatic bearing fault diagnosis based on one-class ν-SVM[J]. Computers & Industrial Engineering,2013, 64(1):357-365.
[8] 郝研,王太勇,万剑,等. 分形盒维数抗噪研究及其在故障诊断中的应用[J]. 仪器仪表学报,2011(3):540-545.
[9] 窦丹丹,姜洪开,何毅娜. 基于信息熵和SVM多分类的飞机液压系统故障诊断[J]. 西北工业大学学报,2012(4):529-534.
[10] 赵勇,杨慎华,郑祺峰,等. 激光预制裂纹槽的断裂分析[J].吉林大学学报(工学版),2011,41(4):984-987.
[11] GULICH D, ZUNINO L. The effects of observational correlated noises on multifractal detrended fluctuation analysis[J]. Physica A:Statistical Mechanics and its Applications,2012,391(16):4100-4110.
[12] CHEN G, WANG Z. A signal decomposition theorem with Hilbert transform and its application to narrowband time series with closely spaced frequency components[J]. Mechanical Systems and Signal Processing,2012(28):258-279.
[13] 舒小红.基于分形理论的双洞隧道洞口段拱顶抗震反应特征分析[J].重庆交通大学学报(自然科学版),2013,32(4):629-634.
[14] DU B, TANG G. Vibration signal fault diagnosis based on generalized fractal dimension[J]. Advances in Information Sciences and Service Sciences,2013,5(6):36.
[15] LI B, ZHANG P, WANG Z, et al. Morphological covering based generalized dimension for gear fault diagnosis[J]. Nonlinear Dynamics,2012,67(4):2561-2571.
[16] 王晓峰,毛德强,冯尚聪. 现代故障诊断技术研究综述[J]. 中国测试,2013,39(6):93-98.
