矢谱分析关键技术与实践研究
|
摘要
随着技术的进步，机械设备的结构越来越复杂，运行条件的要求越来越高，因此需要对设备进行精确的实时监控，尽早发现故障隐患，准确判别故障类型，找出故障根源。但是长期以来，机械状态监测和故障诊断大都以单通道信号分析为基础。转子的振动形式是个空间的概念，单通道信号只是转子在传感器方向上振动的反映，所能提供的转子振动信息并不十分准确、完整。因此监测和诊断结果的可信度并不能满足实际要求。解决这一矛盾的根本所在就是改变单通道信号分析的现状，运用全新的信息融合技术对设备进行监测和诊断。
多传感器信息融合与旋转机械动态特征和回转机理相结合产生了矢谱技术。多传感器信息融合是指动态信号在一定准则下加以自动分析、综合以完成所需的决策和评估而进行的信息处理过程。矢谱技术以多传感器信息融合为基础，致力于转子空间振动的全方位描述。因此，矢谱技术从根本上改变了单通道信号分析割裂信号内在联系的弊端，使监测和诊断更加完善准确。
矢谱分析是针对旋转机械矢量信号的一系列分析方法的总称。矢谱范畴包括针对平稳信号的矢量谱、矢功率谱、矢量倒谱以及针对非平稳信号的短时矢谱等众多分析方法。本文中仅对处于基础地位的矢量谱和矢功率谱进行详细的理论及应用研究，并有代表性的探讨了非平稳信号的短时矢谱分析。还针对不同信息融合程度的矢谱理论，从双通道、三通道到转子不同截面间的信息融合，精确地反映了转子振动的全貌。
在理论探讨的基础上，应用C++ Builder5.0构造了基于Browser/Server模式的矢量信号远程分析系统，并选取实际的矢量故障信号进行了传统分析方法、矢谱、全谱和全息谱等分析方法的比较，在对比中体现了矢谱技术的优越性。
通过本文的理论探讨和矢谱实际应用的研究可以证实：矢谱是一项具有广阔发展前景和较高实用价值的技术。
With the development of technology, the increasing complication of the structure of machinery and its running requirements need accirate real time monitoring, in order to discover the hidden malfunction as soon as possible, identify fault correctly and find out the source of it. But in a long period of time, condition monitoring of the machinery and fault diagnosis are based on the analysis of the signal coming from single channel The vibration of the rotor is a dimensional conception. The signal coming from single channel only reflects the vibration of the rotor in the sensor orientation, so it can't accurately offer integral information. As a result, the reliability of monitoring and diagnosis can't meet the demand of the practice. The key of settling the contradiction is to change status quo that the signal analysis is based on the single channel, and apply bran-new technology, information fusion, in the monitoring and diagnosis.
The vector spectrum descends from the combination of multiple sensor information fusion and the rotary theory of rotor and its dynamic characters. Multiple sensor information fusion means the fusion process of dynamic signals with the same source. Through the process, a deal of information of rotor vibration is picked up, accordingly the vibration of rotor can be characterized in ML Therefore, the vector spectrum can radically change the discontented state and make the monitoring and diagnosis accirate much more.
In fact, vector spectrum comprises a' series of analysis methods: fundamental vector spectrum, vector power spectrum, vector spectrum that used to analysis stationary signals, and short time vector spectrum to non-stationary signals. Being the base of vector spectrum, fundamental vector spectrum and vector power spectrum are studied detailedly in this paper The work of short time vector spectrum is also involved in this paper What's more, this paper also dealt with different degree of information fusion. From double censors to treble censors, to multi-sensor of the two sections in the same rotor, the vibration of the rotary machine is reflected more and more accurately.
After the rigorous study of the theory, using the C++ Builder5.0 as the developing tool, a vector signals analysis system was visualized, which based on the mode of Browser/Server and can run on the Internet With this system, some real fault signals were analyzed and the value of vector spectrum was approved
From the study of the vector spectrum theory and its applications, a conclusion can be affirmed: the vector spectrum has high practical value and a developing future.
多传感器信息融合与旋转机械动态特征和回转机理相结合产生了矢谱技术。多传感器信息融合是指动态信号在一定准则下加以自动分析、综合以完成所需的决策和评估而进行的信息处理过程。矢谱技术以多传感器信息融合为基础，致力于转子空间振动的全方位描述。因此，矢谱技术从根本上改变了单通道信号分析割裂信号内在联系的弊端，使监测和诊断更加完善准确。
矢谱分析是针对旋转机械矢量信号的一系列分析方法的总称。矢谱范畴包括针对平稳信号的矢量谱、矢功率谱、矢量倒谱以及针对非平稳信号的短时矢谱等众多分析方法。本文中仅对处于基础地位的矢量谱和矢功率谱进行详细的理论及应用研究，并有代表性的探讨了非平稳信号的短时矢谱分析。还针对不同信息融合程度的矢谱理论，从双通道、三通道到转子不同截面间的信息融合，精确地反映了转子振动的全貌。
在理论探讨的基础上，应用C++ Builder5.0构造了基于Browser/Server模式的矢量信号远程分析系统，并选取实际的矢量故障信号进行了传统分析方法、矢谱、全谱和全息谱等分析方法的比较，在对比中体现了矢谱技术的优越性。
通过本文的理论探讨和矢谱实际应用的研究可以证实：矢谱是一项具有广阔发展前景和较高实用价值的技术。
With the development of technology, the increasing complication of the structure of machinery and its running requirements need accirate real time monitoring, in order to discover the hidden malfunction as soon as possible, identify fault correctly and find out the source of it. But in a long period of time, condition monitoring of the machinery and fault diagnosis are based on the analysis of the signal coming from single channel The vibration of the rotor is a dimensional conception. The signal coming from single channel only reflects the vibration of the rotor in the sensor orientation, so it can't accurately offer integral information. As a result, the reliability of monitoring and diagnosis can't meet the demand of the practice. The key of settling the contradiction is to change status quo that the signal analysis is based on the single channel, and apply bran-new technology, information fusion, in the monitoring and diagnosis.
The vector spectrum descends from the combination of multiple sensor information fusion and the rotary theory of rotor and its dynamic characters. Multiple sensor information fusion means the fusion process of dynamic signals with the same source. Through the process, a deal of information of rotor vibration is picked up, accordingly the vibration of rotor can be characterized in ML Therefore, the vector spectrum can radically change the discontented state and make the monitoring and diagnosis accirate much more.
In fact, vector spectrum comprises a' series of analysis methods: fundamental vector spectrum, vector power spectrum, vector spectrum that used to analysis stationary signals, and short time vector spectrum to non-stationary signals. Being the base of vector spectrum, fundamental vector spectrum and vector power spectrum are studied detailedly in this paper The work of short time vector spectrum is also involved in this paper What's more, this paper also dealt with different degree of information fusion. From double censors to treble censors, to multi-sensor of the two sections in the same rotor, the vibration of the rotary machine is reflected more and more accurately.
After the rigorous study of the theory, using the C++ Builder5.0 as the developing tool, a vector signals analysis system was visualized, which based on the mode of Browser/Server and can run on the Internet With this system, some real fault signals were analyzed and the value of vector spectrum was approved
From the study of the vector spectrum theory and its applications, a conclusion can be affirmed: the vector spectrum has high practical value and a developing future.
引文
[1] 韩捷，张瑞林等．旋转机械故障机理及诊断技术[M]．北京：机械工业出版社，1996．
[2] 孙惠国，张瑞林．设备状态监测与故障诊断[M]．郑州：河南科学技术出版社，1992．
[3] 张贤达，保铮．非平稳信号分析与处理[M]．北京：国防工业出版社，1998．12～31．
[4] 何正嘉，孟庆丰．机械设备非平稳信号的故障诊断原理及应用[M]．北京：高等教育出版社，2000．147～160．
[5] 王顺奎．国外信息融合技术概述[J]．红外与激光技术，1995，24(1)：1～4．
[6] 袁南儿，杨东勇等．多传感器信息融合及其在工业控制中的应用[J]．浙江工业大学学报，1999，27(4)：28 1～286．
[7] 张彦铎，姜兴渭．多传感器信息融合及在智能故障诊断中的应用[J]．传感器技术，1999，18(2)：18～22．
[8] 王耀南，李树涛．多传感器信息融合及其应用综述[J]．控制与决策，2001，16(5)：158～162．
[9] 李建勋．决策信息融合研究[J]．西安科技大学学报，2000，27(1)：83～87．
[10] 王璇，李春升．多传感器信息融合技术[J]．北京航空航天大学学报，1994，20(4)：402～406．
[11] 郁文贤，雍少为．多传感器信息融合技术述评[J]．国防科技大学学报，1994，16(3)：1～11．
[12] 袁小宏，屈梁生．机械故障诊断中的信息融合利用问题研究[J]．振动、测试与诊断，1999，19(3)：188～193．
[13] 王志武，丁国清．多分辨率多传感器动态数据的融合和应用[J]．上海交通大学学报，2002，22(7)：953～957．
[14] Bill Laws. ToolBox Tips-When you use spectrum, don't use it halfway [J]. Orbit, 1998, 18: 23～26.
[15] Don Southwick. Using Full Spectrum Plots [J]. Orbit, 1993, 14: 19～21.
[16] Don Southwick. Using Full Spectrum Plots Part2 [J]. Orbit, 1993, 15:11～15.
[17] 屈梁生，史东锋．全息谱十年：回顾与展望[J]．振动、测试与诊断，1998，18(4)：235～243．
[18] 孟建，屈梁生．全息谱的分解及其在机械诊断中的应用[J]．机械工程学报，1997，33(2)：106～110．
[19] 钟一谔，何衍宗．转子动力学[M]．北京：清华大学出版社，1987．165～172．
[20] 韩捷，关惠玲，梁川等．矢谱——一种实用的旋转机械故障诊断分析方法[J]．机械强度，1998，20(3)：212～215．
[21] 韩捷，关惠玲，梁川等．旋转机械故障诊断的进动谱研究[J]．振动工程学报，1998，13(11)：67～71．
[22] 韩捷，张琳娜．旋转机械振动的谱强度计量[J]．计量技术，1997，6：22～25．
[23] 宗孔德，胡广书．数字信号处理[M]．北京：清华大学出版社，1988．
[24] 李圣怡，吴学忠，范大鹏．多传感器信息融合理论及在智能制造系统中的应用[M]．长沙：国防科技大学出版社，2000，23～52．
[25] 韩捷，石来德．全谱分析技术及其在故障诊断中的应用[J]．郑州大学学报，2002，
23(3): 44～47.
[26] Bon Bently. How to handle( and some cases of) forward and reverse orbits[J]. Orbit, 1995, 9:4～5.
[27] Donald, Paul Goldman, Jing Yuan. Rotor Dynamics of Centrifugal Compressors in Rotating Stall [J]. Orbit, 2001, 40～59.
[28] Paul Goldman, Agnes Muszynska. Application of full spectrum to rotating machinery diagnostics[J]. Orbit, 1999, 1:17～21.
[29] 胡广书．数字信号处理理论、算法与实现[M]．北京：清华大学出版社，1997，57～63，320～341．
[30] Qu LiangSheng, Liu Xiong, et al. The holospectrom: a new method for rotor surveillance and diagnosis. Mechanical Systrms and Signal Processing, 1989, 3(3): 255～2671
[31] Qu LiangSheng, Liu Xiong, et al. Discovering the holospectrum. J.Noise & Vibration Vorldwide, 1989, 20(2): 58～63.
[32] 刘雄，屈梁生．基于二维全息谱的回转机械亚同步振动分析[J]．化工机械，1994，21(1)：33～37．
[33] 屈梁生，何正嘉．机械故障诊断学[M]．上海：上海科技出版社，1987．
[34] 吴湘淇．信号、系统与信号处理(下)[M]．北京：电子工业出版社，1986，187～197．
[35] 李志华．ARMA序列及其功率谱估计若干新的理论和方法[J]．大连海事大学学报，1997，23(4)：93～97．
[36] 毕果，韩捷，梁川．基于矢量振动信号的AR功率谱分析及应用[J]．郑州大学学报，2003，24(2)．
[37] 何正嘉，赵纪元，孟庆丰等．径向监测诊断中非平稳信号处理的研究[J]．西安交通大学学报，1996，30(7)：36～42．
[38] He Zhengjia, Meng Qingfeng, Zhao Jiyuan, et al. Time-frequency(scale) analysis and diagnosis for nonstationary dynamic of machinery. International Journal of Plant Engineering and Management, 1996, 5(6): 40～47.
[39] 杨福生．小波变换的工程分析与处理[M]．北京：科学出版社，1998．
[40] Leon Cohen: Time-Frequency Analysis: Theory and Applications[M]. Prentice Hall, 1995, 77～85.
[41] Newland D E. Wavelet analysis of vibration, Part 1: Theory. Journal of Vibration and Acoustics, 1994,116(Oet.): 409～416.
[42] 郑建明，李言．伸缩窗口短时Fourier分析[J]．振动、测试与诊断，2000，20(4)：254～230．
[43] 张贤达，保铮．非平稳信号分析与处理[M]．国防工业出版社,1998．20～26．
[44] Mark W D. Spectral analysis of the convolution and filtering of non-stationary stochastic processes[J]. Sound Vib., 1970,11(1): 19～36.
[45] Eshleman. The Role of Sum and Difference Frequencies in Rotating Machinery Fault Diagnosis[J]. International Mechanical Engineering, 1980.
[46] 史东锋，屈梁生．三维全息谱分解在回转机械诊断中的应用研究[J]．西安交通大学学报，1998，32(7)：93～97．
[47] Roger Harker. Understanding Machinery Management-Using Machine Condition and Process Information for Maximum Benefit[J]. Orbit, 2000, 3:24～32.
[48] 周仁睦．转子动平衡——原理、方法和标准[M]．北京：化学工业出版社，1993，
10～12．
[49] 贾民平，许飞云，钟秉林等．远程分布式汽轮机组故障诊断网络系统的研究[J]．中国电机工程学报，1998，18(2)：60～63．
[50] 冯志鹏，宋希庚，薛冬新．旋转机械振动故障诊断理论与技术进展综述[J]．振动与冲击，2001,20(4)：36～41．
[51] 张金玉．基于网络的远程诊断与处理支持中心的研究：[博士学位论文]．西安交通大学，2000．
[52] Donald, Malcolm Wemer. Extending Machine Life[J]. Orbit, 1990,9:5～9.
[53] Donald, Charles. Fundamentals of Rotating Machinery Diagnostics[M], Bently Pressurized Bearing Press, 2002.
[54] George Murray, John Mucci, Scott Brier. Analysis of generator rotor unbalance[J]. Orbit,1993,3:25～29.
[55] Mountford, Lee. Machinery Management-Review of Advances in Diagnostic Methodologies, Proceedings of Power-Gen, 1999,9.
[56] 周绪，韦文斌等．SQL Server7.0入门与提高[M]．北京：清华大学出版社，2000．
[57] Ron Bosmans. Petrochemichal company uses online information to keep critical machine running[J]. Orbit, 1995,9:9～11.
[58] 杨光，杨拥民，袁杰红等．基于B/S模型的设备远程诊断系统的研究[J]．测控技术，2002，20(6)：16～18．
[59] Bill Laws. Remote machinery diagnostics: A cost-effective way to improve plant safety, productivity and profitability[J]. Orbit, 1995,7:13～15.
[60] 黎洪生，何岭松，史铁林等．基于B／S的远程故障诊断专家系统研究[J]．武汉工业大学学报，1999，21(4)：39～41．
[61] Nee A.Y.C. Web-based fault diagnostic and learning system[J]. International Journal of Advanced Manufacturing Technology, 2001, 18(7): 502～511.
[62] 陈翔，胡建华，王清心．AetiveX技术与动态Web页面的开发[J]．昆明理工大学学报，1999,24(2)：147～152．
[2] 孙惠国，张瑞林．设备状态监测与故障诊断[M]．郑州：河南科学技术出版社，1992．
[3] 张贤达，保铮．非平稳信号分析与处理[M]．北京：国防工业出版社，1998．12～31．
[4] 何正嘉，孟庆丰．机械设备非平稳信号的故障诊断原理及应用[M]．北京：高等教育出版社，2000．147～160．
[5] 王顺奎．国外信息融合技术概述[J]．红外与激光技术，1995，24(1)：1～4．
[6] 袁南儿，杨东勇等．多传感器信息融合及其在工业控制中的应用[J]．浙江工业大学学报，1999，27(4)：28 1～286．
[7] 张彦铎，姜兴渭．多传感器信息融合及在智能故障诊断中的应用[J]．传感器技术，1999，18(2)：18～22．
[8] 王耀南，李树涛．多传感器信息融合及其应用综述[J]．控制与决策，2001，16(5)：158～162．
[9] 李建勋．决策信息融合研究[J]．西安科技大学学报，2000，27(1)：83～87．
[10] 王璇，李春升．多传感器信息融合技术[J]．北京航空航天大学学报，1994，20(4)：402～406．
[11] 郁文贤，雍少为．多传感器信息融合技术述评[J]．国防科技大学学报，1994，16(3)：1～11．
[12] 袁小宏，屈梁生．机械故障诊断中的信息融合利用问题研究[J]．振动、测试与诊断，1999，19(3)：188～193．
[13] 王志武，丁国清．多分辨率多传感器动态数据的融合和应用[J]．上海交通大学学报，2002，22(7)：953～957．
[14] Bill Laws. ToolBox Tips-When you use spectrum, don't use it halfway [J]. Orbit, 1998, 18: 23～26.
[15] Don Southwick. Using Full Spectrum Plots [J]. Orbit, 1993, 14: 19～21.
[16] Don Southwick. Using Full Spectrum Plots Part2 [J]. Orbit, 1993, 15:11～15.
[17] 屈梁生，史东锋．全息谱十年：回顾与展望[J]．振动、测试与诊断，1998，18(4)：235～243．
[18] 孟建，屈梁生．全息谱的分解及其在机械诊断中的应用[J]．机械工程学报，1997，33(2)：106～110．
[19] 钟一谔，何衍宗．转子动力学[M]．北京：清华大学出版社，1987．165～172．
[20] 韩捷，关惠玲，梁川等．矢谱——一种实用的旋转机械故障诊断分析方法[J]．机械强度，1998，20(3)：212～215．
[21] 韩捷，关惠玲，梁川等．旋转机械故障诊断的进动谱研究[J]．振动工程学报，1998，13(11)：67～71．
[22] 韩捷，张琳娜．旋转机械振动的谱强度计量[J]．计量技术，1997，6：22～25．
[23] 宗孔德，胡广书．数字信号处理[M]．北京：清华大学出版社，1988．
[24] 李圣怡，吴学忠，范大鹏．多传感器信息融合理论及在智能制造系统中的应用[M]．长沙：国防科技大学出版社，2000，23～52．
[25] 韩捷，石来德．全谱分析技术及其在故障诊断中的应用[J]．郑州大学学报，2002，
23(3): 44～47.
[26] Bon Bently. How to handle( and some cases of) forward and reverse orbits[J]. Orbit, 1995, 9:4～5.
[27] Donald, Paul Goldman, Jing Yuan. Rotor Dynamics of Centrifugal Compressors in Rotating Stall [J]. Orbit, 2001, 40～59.
[28] Paul Goldman, Agnes Muszynska. Application of full spectrum to rotating machinery diagnostics[J]. Orbit, 1999, 1:17～21.
[29] 胡广书．数字信号处理理论、算法与实现[M]．北京：清华大学出版社，1997，57～63，320～341．
[30] Qu LiangSheng, Liu Xiong, et al. The holospectrom: a new method for rotor surveillance and diagnosis. Mechanical Systrms and Signal Processing, 1989, 3(3): 255～2671
[31] Qu LiangSheng, Liu Xiong, et al. Discovering the holospectrum. J.Noise & Vibration Vorldwide, 1989, 20(2): 58～63.
[32] 刘雄，屈梁生．基于二维全息谱的回转机械亚同步振动分析[J]．化工机械，1994，21(1)：33～37．
[33] 屈梁生，何正嘉．机械故障诊断学[M]．上海：上海科技出版社，1987．
[34] 吴湘淇．信号、系统与信号处理(下)[M]．北京：电子工业出版社，1986，187～197．
[35] 李志华．ARMA序列及其功率谱估计若干新的理论和方法[J]．大连海事大学学报，1997，23(4)：93～97．
[36] 毕果，韩捷，梁川．基于矢量振动信号的AR功率谱分析及应用[J]．郑州大学学报，2003，24(2)．
[37] 何正嘉，赵纪元，孟庆丰等．径向监测诊断中非平稳信号处理的研究[J]．西安交通大学学报，1996，30(7)：36～42．
[38] He Zhengjia, Meng Qingfeng, Zhao Jiyuan, et al. Time-frequency(scale) analysis and diagnosis for nonstationary dynamic of machinery. International Journal of Plant Engineering and Management, 1996, 5(6): 40～47.
[39] 杨福生．小波变换的工程分析与处理[M]．北京：科学出版社，1998．
[40] Leon Cohen: Time-Frequency Analysis: Theory and Applications[M]. Prentice Hall, 1995, 77～85.
[41] Newland D E. Wavelet analysis of vibration, Part 1: Theory. Journal of Vibration and Acoustics, 1994,116(Oet.): 409～416.
[42] 郑建明，李言．伸缩窗口短时Fourier分析[J]．振动、测试与诊断，2000，20(4)：254～230．
[43] 张贤达，保铮．非平稳信号分析与处理[M]．国防工业出版社,1998．20～26．
[44] Mark W D. Spectral analysis of the convolution and filtering of non-stationary stochastic processes[J]. Sound Vib., 1970,11(1): 19～36.
[45] Eshleman. The Role of Sum and Difference Frequencies in Rotating Machinery Fault Diagnosis[J]. International Mechanical Engineering, 1980.
[46] 史东锋，屈梁生．三维全息谱分解在回转机械诊断中的应用研究[J]．西安交通大学学报，1998，32(7)：93～97．
[47] Roger Harker. Understanding Machinery Management-Using Machine Condition and Process Information for Maximum Benefit[J]. Orbit, 2000, 3:24～32.
[48] 周仁睦．转子动平衡——原理、方法和标准[M]．北京：化学工业出版社，1993，
10～12．
[49] 贾民平，许飞云，钟秉林等．远程分布式汽轮机组故障诊断网络系统的研究[J]．中国电机工程学报，1998，18(2)：60～63．
[50] 冯志鹏，宋希庚，薛冬新．旋转机械振动故障诊断理论与技术进展综述[J]．振动与冲击，2001,20(4)：36～41．
[51] 张金玉．基于网络的远程诊断与处理支持中心的研究：[博士学位论文]．西安交通大学，2000．
[52] Donald, Malcolm Wemer. Extending Machine Life[J]. Orbit, 1990,9:5～9.
[53] Donald, Charles. Fundamentals of Rotating Machinery Diagnostics[M], Bently Pressurized Bearing Press, 2002.
[54] George Murray, John Mucci, Scott Brier. Analysis of generator rotor unbalance[J]. Orbit,1993,3:25～29.
[55] Mountford, Lee. Machinery Management-Review of Advances in Diagnostic Methodologies, Proceedings of Power-Gen, 1999,9.
[56] 周绪，韦文斌等．SQL Server7.0入门与提高[M]．北京：清华大学出版社，2000．
[57] Ron Bosmans. Petrochemichal company uses online information to keep critical machine running[J]. Orbit, 1995,9:9～11.
[58] 杨光，杨拥民，袁杰红等．基于B/S模型的设备远程诊断系统的研究[J]．测控技术，2002，20(6)：16～18．
[59] Bill Laws. Remote machinery diagnostics: A cost-effective way to improve plant safety, productivity and profitability[J]. Orbit, 1995,7:13～15.
[60] 黎洪生，何岭松，史铁林等．基于B／S的远程故障诊断专家系统研究[J]．武汉工业大学学报，1999，21(4)：39～41．
[61] Nee A.Y.C. Web-based fault diagnostic and learning system[J]. International Journal of Advanced Manufacturing Technology, 2001, 18(7): 502～511.
[62] 陈翔，胡建华，王清心．AetiveX技术与动态Web页面的开发[J]．昆明理工大学学报，1999,24(2)：147～152．