發(fā)布時間：2018-01-05 18:06

  本文關鍵詞：頻帶熵方法及其在滾動軸承故障診斷中的應用　出處：《上海交通大學》2012年碩士論文　論文類型：學位論文

  更多相關文章： 故障診斷 狀態(tài)監(jiān)測 時頻分析 頻帶熵 帶通濾波 遺傳算法 滾動軸承

【摘要】：隨著科學技術的日益進步與現(xiàn)代工業(yè)的飛速發(fā)展,機械設備不斷向大型、復雜、高速、高效及重載的方向發(fā)展;與此同時,其工作和運行環(huán)境也更加復雜和苛刻。這些設備一旦突然發(fā)生故障,不僅會增加企業(yè)的維護成本,降低企業(yè)的生產效率,還可能造成巨大的經濟損失,甚至導致嚴重的人員傷亡,產生不良的社會影響。因此,如何對設備進行有效的狀態(tài)監(jiān)測和故障診斷,是當前亟需解決的問題。 如何有效提取反映設備運行狀態(tài)的特征,以及準確判斷故障類別,一直是故障診斷領域的研究熱點,新方法和新理論的研究也層出不窮,對豐富和完善機械故障診斷技術起到了重要作用。本文以滾動軸承為研究對象,提出頻帶熵方法,并對其在故障診斷中的應用進行了研究,旨在為滾動軸承狀態(tài)監(jiān)測提供一個新指標,為故障診斷信號預處理提供一種新方法,論文主要包括以下幾個方面的內容: (1)從理論分析與工程應用的角度出發(fā),闡述了論文的選題背景和研究意義。分析了機械設備故障診斷方法、滾動軸承故障診斷、時頻分析與信息熵理論等方面的國內外研究現(xiàn)狀,確立了本文的研究內容。 (2)介紹了作為本文理論基礎的幾種時頻分析方法及信息熵理論,借鑒譜峭度方法提出頻帶熵概念,定義頻帶熵為某一頻率上(頻帶內)信號的復雜度,或者說不確定性,給出了頻帶熵的基本算法,最后從濾波的角度對頻帶熵概念進行了擴展。 (3)介紹了滾動軸承的振動機理和故障特征。討論了頻帶熵指標用于滾動軸承狀態(tài)監(jiān)測的可能性,對其魯棒性進行了研究,證明其對奇異點的不敏感性�；�于頻帶熵的上述特性,將其應用于滾動軸承全壽命周期數(shù)據(jù)分析,探討了頻帶熵指標在性能退化各階段的表現(xiàn)。介紹了為上述理論提供數(shù)據(jù)支撐的滾動軸承故障試驗和加速疲勞壽命試驗,通過對試驗數(shù)據(jù)的分析,表明頻帶熵可作為狀態(tài)監(jiān)測指標的有效補充。 (4)針對共振解調帶通濾波中心頻率難以確定的問題,提出了頻帶熵確定中心頻率的方法。對基于STFT的頻帶熵,討論了頻率離散點數(shù)、時頻分析窗長、窗函數(shù)類型等參數(shù)對頻帶熵的影響;對基于小波包變換的頻帶熵,討論了小波包分解層數(shù)和小波包函數(shù)的選擇對頻帶熵的影響。最后將兩種方法應用于仿真和實際的滾動軸承故障診斷。分析結果證明,頻帶熵能夠準確確定信號的共振頻帶,提升帶通濾波和包絡解調后的診斷效果。 (5)提出頻帶熵與遺傳算法相結合的方法,用于共振解調帶通濾波器的優(yōu)化設計。以頻帶熵最小為遺傳算法的優(yōu)化目標,通過選擇、交叉、變異等操作,在取值范圍內尋找最優(yōu)的中心頻率和帶寬組合,設計優(yōu)化濾波器。通過對仿真信號和不同信噪比實驗數(shù)據(jù)的分析,證明此方法能夠有效確定濾波中心頻率和帶寬,從而提高信號的信噪比,實現(xiàn)對軸承故障的診斷。
[Abstract]:Along with the rapid development of science and technology and the rapid development of modern industry , the machinery and equipment continue to develop in the direction of large , complex , high speed , high efficiency and heavy load . At the same time , its work and operating environment are more complex and severe . Once the equipment suddenly fails , it will not only increase the maintenance cost of the enterprise , reduce the production efficiency of the enterprise , but also can cause great economic loss , even lead to serious casualties and bad social impact . Therefore , how to carry on the effective state monitoring and fault diagnosis of the equipment is a problem that needs to be solved urgently . How to efficiently extract the characteristics reflecting the running state of the equipment and accurately judge the fault category has been a hot topic in the field of fault diagnosis . The new method and the new theory have been studied in this paper . The application of the new method and the new theory has been studied in this paper . The application of the new method and the new theory in fault diagnosis has been studied . The purpose of this paper is to provide a new index for the monitoring of rolling bearing status . ( 1 ) From the angle of theoretical analysis and engineering application , the background and significance of the thesis are described . The research status of fault diagnosis method , fault diagnosis , time - frequency analysis and information entropy theory are analyzed , and the research contents of this paper are established . ( 2 ) This paper introduces several time - frequency analysis methods and information entropy theory , which is the base of this paper . The concept of frequency band entropy is proposed by using the method of spectrum . The frequency band entropy is defined as the complexity of the signal in a certain frequency ( in - band ) signal , or the uncertainty is given . The basic algorithm of frequency band entropy is given . Finally , the concept of frequency band entropy is expanded from the angle of filtering . ( 3 ) The vibration mechanism and fault characteristics of rolling bearing are introduced . The possibility of using frequency band entropy index to monitor the state of rolling bearing is discussed , and its robustness is studied . Based on the above mentioned characteristics of frequency band entropy , the performance of the frequency band entropy index in all stages of performance degradation is discussed . The fault test and accelerated fatigue life test of the rolling bearing supported by the theory are introduced . The analysis of the test data shows that the frequency band entropy can be used as the effective complement of the state monitoring index . ( 4 ) Based on the frequency band entropy of STFT , the influence of frequency discrete point number , time frequency analysis window length and window function type on frequency band entropy is discussed . The frequency band entropy based on STFT is discussed . ( 5 ) The method of combining the frequency band entropy with the genetic algorithm is proposed for the optimization design of the resonant demodulation band - pass filter . The optimal filter is designed by selecting , crossing , mutation and so on . The optimal filter is designed by selecting , crossing , mutation and so on . By analyzing the experimental data of the simulation signal and the different signal - to - noise ratio , this method can effectively determine the center frequency and bandwidth of the filter , thereby improving the signal - to - noise ratio of the signal and realizing the diagnosis of the bearing fault .

【學位授予單位】：上海交通大學
【學位級別】：碩士
【學位授予年份】：2012
【分類號】：TH165.3

【參考文獻】

相關期刊論文 前10條

1 于德介,雷慧;用BP神經網絡診斷結構破損[J];工程力學;2001年01期

2 鄒紅星,周小波,李衍達;時頻分析:回溯與前瞻[J];電子學報;2000年09期

3 朱冰蓮,楊磊;心音信號的短時傅立葉變換分析[J];重慶大學學報(自然科學版);2004年08期

4 王志剛,李友榮,呂勇,李方;基于諧波小波變換的低速滾動軸承故障診斷[J];軸承;2004年10期

5 王平,廖明夫;滾動軸承故障診斷的自適應共振解調技術[J];航空動力學報;2005年04期

6 曾宇清,王衛(wèi)東,賀啟庸;按頻帶順序排列的小波包新算法及應用[J];力學學報;1998年02期

7 伍星,羅德?lián)P,郭之璜;快速頻率細化 ZOOM 軟件的開發(fā)研究[J];昆明理工大學學報;1997年02期

8 鄒紅星,戴瓊海,李衍達,盧旭光;不含交叉項干擾且具有WVD聚集性的時頻分布之不存在性[J];中國科學E輯;2001年04期

9 申_",黃樹紅,韓守木,楊叔子;旋轉機械振動信號的信息熵特征[J];機械工程學報;2001年06期

10 林京,屈梁生;信號時頻熵及其在齒輪裂紋識別中的應用[J];機械傳動;1998年02期

相關博士學位論文 前1條

1 謝平;故障診斷中信息熵特征提取及融合方法研究[D];燕山大學;2006年

，

本文編號：1384232