TWI238926B - Singularity analysis method and its processes to apply in the defect diagnosis of mechanical components - Google Patents

Singularity analysis method and its processes to apply in the defect diagnosis of mechanical components Download PDF

Info

Publication number
TWI238926B
TWI238926B TW93115744A TW93115744A TWI238926B TW I238926 B TWI238926 B TW I238926B TW 93115744 A TW93115744 A TW 93115744A TW 93115744 A TW93115744 A TW 93115744A TW I238926 B TWI238926 B TW I238926B
Authority
TW
Taiwan
Prior art keywords
analysis
singularity
value
demodulation
vibration signal
Prior art date
Application number
TW93115744A
Other languages
Chinese (zh)
Other versions
TW200540588A (en
Inventor
Yu-Tai Shen
Original Assignee
Southern Taiwan University Of
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Southern Taiwan University Of filed Critical Southern Taiwan University Of
Priority to TW93115744A priority Critical patent/TWI238926B/en
Application granted granted Critical
Publication of TWI238926B publication Critical patent/TWI238926B/en
Publication of TW200540588A publication Critical patent/TW200540588A/en

Links

Landscapes

  • Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
  • Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)

Abstract

This invention is to propose both a singularity analysis method and its computational algorithm to apply in the defect diagnosis of mechanical components. By applying the singularity analysis to the vibration impulse in time domain, a quantitative feature could be derived from calculating the singularity exponents of the demodulated signal and would become more explicit in defect diagnosis. The quantitative feature would be capable of applying in an automatic diagnosis system to judge the running condition of mechanical systems.

Description

^238926 玖、發明說明: 【發明所屬之技術領域】 本發明係有關一種振動訊號之奇異性診斷處理及執行 方法,主要係藉由取一時域振動訊號進行高頻解調,求取 此脈衝敲擊點的責異性分析值,以可直接判定是否為機械 損壞之脈衝振動,並可利用一數值形式來表示機械系敲 運作狀況,如此將可方便電腦執行判斷,以符合機械損壞 偵測自動化之需求。 ' 【先前技術】 機械系統元件運倾況之鱗在於發現賊之初期損 壞’可在顧發生故障敍早發終訊,㈣產業製造之 調適與安排進行維修。由於機械系統元件的初期損壞將造 成系統運作產生敲擊雜,並由於機礙統巾多數運動元 件均屬於旋轉兀件,若此類元件上發生損壞將會產生週期 性敲擊,因此診斷機械元件之運作狀況則在於侧此敲擊 振動訊號的發生。其分析過程,一般可分為資料擷取、訊 號處理與頻譜分析等三階段。首先利用振動感測器(如位移 σ十速度規或加速規)將振動之物理量轉換為電壓形式,再 經由放大器將電壓適度放大後,透過類比轉數位轉換介面 將電壓輸入電腦中進行數位式訊號處理,以擷取振動訊號 之重要訊息,最後進行頻譜分析以獲取振動頻譜圖以呈現 出振動訊號之頻率特性,並經由頻譜的頻率分布與模式 1238926 (pattern)來判斷機械系統元件的損壞與否。但由於在機械系 統振動訊射滲有相當贿之雜訊,若元件損壞較輕微則 其產生之振動訊號能量較小,因此將會被雜訊所掩蓋而難 以察覺’只有到發生相當嚴重猶時其產生之振動訊號能 量才變大,並較明顯大於雜訊振動能量時才會明顯呈現, 然而此時元件損壞已變嚴重’所以機械系統將隨時可能因 而造成嚴重故障停止運轉。此種現象為一般頻譜分析儀在 分析機械振動訊號時所面臨的困難,亦是在實際運用時無 法發揮損壞診斷預警效果之重大缺失。 對於上述問題,目前最常被採用的訊號處理方法為高 頻解調分析法(high frequency resonance technique),以提高 機械診斷的正確性。而當由元件損壞所產生之振動與機械 系統之共振所造成之爾現象來分析,可贿現機械系統 之元件損壞振動訊號可表示為v(t)=d(t) · s⑴+ n(t)其中d⑴ 為元件損壞所產生之週期性敲擊序列訊號函數,一般其頻 率約為數百Hz ; s⑴為機械系統發生共振之時間函數,一般 其頻率約為數kHz,甚至十幾kHz以上;n⑴則為雜訊之時間 函數’ 一般其頻率能量約分布在為2kHz以下〔如第一圖所 示,敲擊訊號可表示為一指數衰減函數形式〕。當d(t)與s(t) 之乘積將呈現訊號之調變現象,若以簡化機械系統之共振 頻率來看可以表示為如第二圖所示之現象,此調變現象可 將約為數百Hz的損壞振動訊號調變至數kHz。因此若能將機 1238926 械振動訊號K 0以帶通滤波器(選擇2 kH z以上之頻帶)滤除 雜訊n(t),將可僅擷取調變訊號,並大幅降低雜訊所產生之 訊號干擾;然後再將此調變訊號予以解調變則可還原元件 損壞之敲擊訊號。 傳統上,完成咼頻解調分析的方法可以分為硬體與軟 體兩種方法。硬體解調變的方法是將所擷取的機械振動訊 號經過帶通濾波器後,再將此訊號通過lkHz之低通濾波則^ 238926 发明 Description of the invention: [Technical field to which the invention belongs] The present invention relates to a singularity diagnosis processing and execution method of vibration signals, which mainly performs high-frequency demodulation by taking a time-domain vibration signal to obtain the pulse knock. The analysis value of the hit point's liability is based on the pulse vibration that can directly determine whether it is mechanical damage, and a numerical form can be used to indicate the operating status of the mechanical system. This will make it easier for the computer to perform judgments to meet the mechanical damage detection automation. demand. '[Previous technology] The scale of the mechanical system components is the initial damage of the thief'. You can send the final message when the failure occurs, and adjust and arrange for maintenance by the industrial manufacturing. Because the initial damage of the mechanical system components will cause the system to generate knocks, and due to the obstruction of the system, most of the moving components are rotating components. If damage occurs to such components, periodic knocks will occur, so the mechanical components are diagnosed. The operating condition lies in the occurrence of the vibration signal. The analysis process can be generally divided into three stages: data acquisition, signal processing, and spectrum analysis. First use a vibration sensor (such as a displacement σ ten speed gauge or an acceleration gauge) to convert the physical quantity of vibration into a voltage form. After the voltage is moderately amplified by an amplifier, the voltage is input into a computer through an analog-to-digital conversion interface for digital signals. Processing to capture important information of vibration signals, and finally perform spectrum analysis to obtain vibration spectrum diagrams to show the frequency characteristics of vibration signals, and to judge the damage of mechanical system components through the frequency distribution and pattern of the frequency spectrum 1238926 (pattern) . However, due to the relatively brittle noise in the vibration signal penetration of the mechanical system, if the component is damaged slightly, the energy of the vibration signal generated is small, so it will be covered by the noise and difficult to detect. The energy of the vibration signal generated by it will become larger, and it will be obvious when it is larger than the vibration energy of noise. However, at this time, the damage of the components has become serious. Therefore, the mechanical system may stop running at any time due to this. This phenomenon is a difficulty faced by general spectrum analyzers when analyzing mechanical vibration signals, and is also a major deficiency in the failure to exert the effect of damage diagnosis and early warning in actual use. For the above problems, the most commonly used signal processing method is the high frequency resonance technique to improve the accuracy of mechanical diagnosis. When analyzing the phenomenon caused by the vibration caused by the damage of the component and the resonance of the mechanical system, the vibration signal of the damage of the component of the mechanical system can be expressed as v (t) = d (t) · s⑴ + n (t ) Where d⑴ is a periodic tapping sequence signal function generated by component damage, and its frequency is generally about several hundred Hz; s⑴ is a time function of the resonance of a mechanical system, and its frequency is generally about several kHz, even more than a dozen kHz; n⑴ It is a time function of noise. Generally, its frequency energy is distributed below 2 kHz (as shown in the first figure, the tap signal can be expressed as an exponential decay function form). When the product of d (t) and s (t) will show the modulation phenomenon of the signal, if the resonance frequency of the mechanical system is simplified, it can be expressed as the phenomenon shown in the second figure. The damaged vibration signal of hundreds of Hz is modulated to several kHz. Therefore, if the 1238926 mechanical vibration signal K 0 can be used to filter out the noise n (t) with a band-pass filter (select a frequency band above 2 kH z), only the modulation signal can be captured, and the noise generated by the noise can be greatly reduced. Signal interference; then demodulate this modulation signal to restore the impact signal of component damage. Traditionally, the methods for completing the demodulation analysis of audio frequency can be divided into two methods: hardware and software. The method of hardware demodulation is to pass the captured mechanical vibration signal through a band-pass filter, and then pass this signal through a low-pass filter of lkHz.

可獲得解調變訊號,其中所採用之帶通濾波與低通濾波均 是以硬體方式達成;但是此一方法所獲得之解調變訊號將 有相當程度之失真。至於軟體解調變的方法是將所擷取的 機械振動訊號經由類比轉數位方式輸入電腦中,再利用數 位濾、波方法或快速傅立葉轉換將所擷取訊號做帶通濾波, 然後再將帶通濾波後之訊號取Hilbert轉換以獲得最後之解 調變訊號。由上述兩種解調變方法獲致解調訊號後,再以 頻譜分析法分析振動頻率中是否含有機械元件損壞的特徵 頻率’藉以來判斷機械元件的損壞。雖然藉由振動頻譜可 精破判斷出機械損壞的元件,但頻譜的判讀一般需要專業 人員來進行,對於產業大廠機械數量多的情況下,將需相 田數蓋的專業人員來進行機械損壞的分析判讀,如此將會 間接提高生產成本。 也因具有上述之諸等缺失,發明人係於9 2年7月7 曰以「機械系統元件運作狀況診斷之訊號處理及執行方法」 7 ==案提_請(編列為申請第9 2 專),-種機械系統元件 H4號 法,包括冑·· 卿狀,兄錢之訊號處理方 一解調分析函數·· h ^fc yfw (Ο jnt 其中: /c為帶通頻帶之中心頻率; 九為帶通頻帶之頻寬; 為比例參數可用於調整帶通頻帶之衰減斜率(S咖 attenuation); 且解調訊號則可表示為v⑴與U褶積(晴。iuti〇n)之 絕對值,即 匕,/c,/,) 一 L v(r) \/c,4 (ί 一 r) d/r 。 及一種機械系統元件運作狀況診斷之訊號處理及執行方法,包 括有: h a,fc,fw it) -e^)\eilz{fcJY)t ^1η{ί^Ί:)χ 其中: 乂為帶通頻帶之中心頻率; /νν為帶通頻帶之頻寬; 1238926 «為比例參數可用於調整帶通頻帶之衰減斜率(柳e attenuation) i 且解調訊號則可表示為v⑴與u)姆(贈。iuti〇n)之 絕對值,即The demodulated signal can be obtained. The band-pass filtering and low-pass filtering used are both implemented in hardware; however, the demodulated signal obtained by this method will have a considerable degree of distortion. As for the software demodulation method, the captured mechanical vibration signal is input into the computer by analog to digital method, and then the captured signal is band-pass filtered by digital filtering, wave method or fast Fourier transform, and then the band is filtered. The signal after pass filtering is converted by Hilbert to obtain the final demodulated signal. After the demodulation signal is obtained by the above two demodulation methods, a frequency analysis method is used to analyze whether the vibration frequency contains the characteristic frequency of the damage of the mechanical component. Therefore, the damage of the mechanical component is judged. Although the mechanically damaged components can be accurately judged through the vibration spectrum, the interpretation of the frequency spectrum usually requires professionals to perform. For the large number of machinery in the industrial large factory, professionals with the same number of covers will be required to perform the mechanical damage. Analysis and interpretation, this will indirectly increase production costs. Because of the above-mentioned shortcomings, the inventor used the "signal processing and execution method for the diagnosis of the operating status of mechanical system components" on July 7, 1992. 7 == Case _ Please (listed as the 9th ), A kind of mechanical system element H4 method, including 胄 ·, clear, demodulation analysis function of the signal processing side of the brother ... h ^ fc yfw (0 jnt where: / c is the center frequency of the bandpass band; Nine is the bandwidth of the bandpass band; is a proportional parameter that can be used to adjust the attenuation slope of the bandpass band (Scaattenuation); and the demodulation signal can be expressed as the absolute value of v⑴ and U convolution (clear. Iuti〇n) , That is, dagger, / c, /,)-L v (r) \ / c, 4 (ί a r) d / r. And a signal processing and execution method for the diagnosis of the operating status of mechanical system components, including: ha, fc, fw it) -e ^) \ eilz {fcJY) t ^ 1η {ί ^ Ί:) χ where: χ is a bandpass The center frequency of the frequency band; / νν is the bandwidth of the bandpass band; 1238926 «is a proportional parameter that can be used to adjust the attenuation slope of the bandpass band (will e attenuation) i, and the demodulation signal can be expressed as v⑴ and u) .Iuti〇n) absolute value, that is

e-Jc Jw W 1:1 lj(^) haJcJw (t ~ T) dT 上述理論分析以下列演算法實現於電腦系統中·· 基於電腦之規格特性設定資料擷取之取樣時間τ、參數α、頻 見/w與中心頻率/c之範圍,則波形函數長度為2ΚΤ之解調 變函數可表示為 、“幻=」_/苧)V·2冬>,(/c+夸)' j π kT } 其中Α:(_Κ=^=Κ)為解調分析函數圖形中的某一點位置; 就中心頻率/c範圍取等頻率間隔妒,即/0, /0+妒, /〇+2妒,···,/〇+以/’而頻寬忍為定值,則解調變分 矩陣可表示為 Η (η 十 1)χ(2κ+ι) u-K)υ—κ) ^/〇,/w(-K+l) ... haJ0,fw(K)-、/〇博,九(-K+1) ··· \/。缚,人(K)e-Jc Jw W 1: 1 lj (^) haJcJw (t ~ T) dT The above theoretical analysis is implemented in a computer system with the following algorithm. · The sampling time τ, parameter α, See the range of / w and center frequency / c, then the demodulation function with a waveform function length of 2 kt can be expressed as, "Magic =" _ / 苎) V · 2 Winter >, (/ c + quad) 'j π kT} where Α: (_ Κ = ^ = Κ) is the position of a point in the demodulation analysis function graph; take equal frequency intervals for the center frequency / c range, that is, / 0, / 0 + jealousy, / 〇 + 2jealousy , ··· , / 〇 + takes / 'and the bandwidth tolerance is a fixed value, then the demodulation variational matrix can be expressed as Η (η 十 1) χ (2κ + ι) uK) υ—κ) ^ / 〇, / w (-K + l) ... haJ0, fw (K)-, / 〇 博 , 九 (-K + 1) ··· \ /. Binding, person (K)

Kj^ndfjS ^ Kif^ndfJw(-K+l) ·.. Ktf〇^ndf f ^ 9 1238926 \ n+1 h 八 其中\表示第固頻帶之解調分析向量。 對一機械振動訊號序列v(m),0=m =M,其解調變分析則可 表示為 八 一 abs(^ ^ κν) 3^η+1,Μ+2Κ+1 ^2 • • A abs(/z2: • φ 杯v) Len+lJ _abs(in+ 1*ν)^Kj ^ ndfjS ^ Kif ^ ndfJw (-K + l) · .. Ktf〇 ^ ndf f ^ 9 1238926 \ n + 1 h Eight where \ represents the demodulation analysis vector of the fixed band. For a mechanical vibration signal sequence v (m), 0 = m = M, the demodulation analysis can be expressed as Bayi abs (^ ^ κν) 3 ^ η + 1, Μ + 2Κ + 1 ^ 2 • • A abs (/ z2: • φ cup v) Len + lJ _abs (in + 1 * ν) ^

其中^表示由第A:個頻帶所獲致之解調訊號,*則表示向 量褶積運算’而abs〇則是取其向量中每一元素之絕對值; 最後分別對解調訊號做頻譜分析,則三維頻譜可表示為 A 一 3(¾) n+l,M+2K+l ^ Ε2 3(¾) • • 其中4·)表示對向量序列取頻譜分析。 在「機械系統元件運作狀況診斷之訊號處理及執行方 法」中’提出調變訊號的三維頻譜圖〔如第三〜 u _所 10 1238926 不〕,可有祕者在選擇帶雜波_上_择,並 以立體圖現振_譜,將有助轉專業人員進ς 振動頻譜,以確定機械損壞之發生。顧此—診斷: 二維頻譜圖的缺失做改善,但仍簡要人員的檢視判斷。 若能以電腦自動《酬初步狀況,當已有初步損壞時再 由專業人員來進-步確認損壞並判斷機械損壞之元件,這 將更能符合目前製造業自動化與節省製造成本的需求。 【發明内容】Among them, ^ indicates the demodulation signal obtained by the A: frequency band, * indicates the vector convolution operation, and abs〇 takes the absolute value of each element in the vector; finally, perform spectrum analysis on the demodulated signal. Then the three-dimensional spectrum can be expressed as A-3 (¾) n + 1, M + 2K + l ^ E2 3 (¾) • • where 4 ·) means spectrum analysis of the vector sequence. In "Signal Processing and Execution Method for Diagnosis of Mechanical System Element Operation Status", "Three-dimensional spectrum diagram of modulation signal [such as the third ~ u _ 101038926 not], there may be secrets in the selection with clutter _ on _ Selecting and presenting the vibration spectrum in a three-dimensional view will help professionals to enter the vibration spectrum to determine the occurrence of mechanical damage. Consider this—diagnosis: The lack of two-dimensional spectrograms has been improved, but the personnel's inspection and judgment are still brief. If the computer can be used to automatically report the initial status, when there is initial damage, then professionals will further confirm the damage and judge the mechanically damaged components, which will better meet the current needs of manufacturing automation and saving manufacturing costs. [Summary of the Invention]

本發明係錢-縣動職之奇異性診斷處理及執 行方法,主要係藉錄—時域振動訊號,以—解調分析函 數進行訊號處理,並設定該函數帶通頻率之一序列衰減率 (slope attenuation)以分別進行高頻解調,將所獲得的一序列 解凋§fl號分別求取其最大值,最後求取此最大值對應帶通 頻率衰減率(slope attenuation)在全對數座標上之曲線斜The invention is a singularity diagnosis processing and execution method of Qian-Xian County, which mainly borrows a time-domain vibration signal, processes the signal with a demodulation analysis function, and sets a sequence attenuation rate (band attenuation frequency) of the function ( slope attenuation) to perform high-frequency demodulation separately, and obtain the maximum value of the obtained sequence of §fl number, and finally find the maximum value corresponding to the band pass frequency attenuation rate (slope attenuation) on the full logarithmic coordinates. Curve slope

率,此即為脈衝敲擊點的奇異性分析值,其利用解調分析 函數: h a、fc (0 jnt -e 其中: Λ為帶通頻帶之中心頻率; 九為帶通頻帶之頻寬; β為比例參數可用於調整帶通頻帶之衰減率(sl〇pe 11 1238926 au__。而解調訊麵可表示為猶、』 (convolution)之絕對值,即 貝 〜,/c⑺—ioo V(r) \/c,/y (卜 r) ☆ 並且存在一常數C滿足下式 〜jc,/w⑺幺Caa 其中指數α為振動訊號之奇異性分析值; 當取上式心,八,人⑺之最大極限值時,則Rate, which is the singularity analysis value of the pulse striking point, which uses the demodulation analysis function: ha, fc (0 jnt -e where: Λ is the center frequency of the band-pass band; 9 is the bandwidth of the band-pass band; β is a proportional parameter that can be used to adjust the attenuation rate of the band-pass band (slope 11 1238926 au__. And the demodulation signal can be expressed as absolute value of “convolution”, that is, ~~ ioo V (r ) \ / c, / y (卜 r) ☆ and there is a constant C satisfying the following formula ~ jc, / w⑺ 幺 Caa where the index α is the singularity analysis value of the vibration signal; At the maximum limit, then

L,/c,尺(’)W C / 若將上式對應不同比例參數β,並分別求取其最大值,則 此最大值對應比例參數在全對數座標上之曲線斜率,即 為此脈衝敲擊點的奇異性分析值α,可表示為: ,^uW-iogU log «ι - log a2L, / c, ruler (') WC / If the above formula corresponds to different scale parameters β and the maximum value is obtained, this maximum value corresponds to the slope of the curve of the scale parameter on the full logarithmic coordinates, that is, for this pulse. The singularity analysis value α of the hit point can be expressed as: ^ uW-iogU log «ι-log a2

其中fli與A表示兩相鄰比例參數值; 藉此’當振動訊號是由損壞敲擊所獲致者,其奇異性分析 值較低;而對正常機械的振動訊號進行分析,則呈現較高 的奇異性分析值。 【實施方式】 為使本發明使用技術手段、發明特徵、達成目的及功 效易於明白了解,茲配合圖示說明如下: 12 1238926 本發明係有關一種振動訊號之奇異性診斷處理及執行 方法,主要係藉由取一時域振動訊號,以一解調分析函數 進行訊號處理,並設定帶通頻率之一序列衰減率(sl〇pe attenuation)以分別進行高頻解調,將所獲得的一序列解調气Among them, fli and A represent two adjacent proportional parameter values; by this, when the vibration signal is caused by damage, its singularity analysis value is low; and when the vibration signal of normal machinery is analyzed, it shows a higher value Singularity analysis value. [Embodiment] In order to make the technical means, inventive features, objectives and effects of the present invention easy to understand, the following illustrations are illustrated together: 12 1238926 The present invention relates to a method and method for diagnosing and analyzing the singularity of vibration signals. By taking a time-domain vibration signal, performing signal processing with a demodulation analysis function, and setting a sequence attenuation rate (slope attenuation) of a band pass frequency to perform high-frequency demodulation respectively, the obtained sequence is demodulated gas

號兮別求取其最大值,最後求取此最大值對應帶通頻率奢 減率(slope attenuation)在全對數座標上之曲線斜率,此即 為脈衝敲擊點的奇異性分析值,其利用解調分析函數: Kfc,fw (0 = -^~β^\βηπ{ί€-^ JTCt ) 其中: /c為帶通頻帶之中心頻率; 久為帶通頻帶之頻寬; (slope 褶積 α為比例參數可用於調整帶通頻帶之衰減率 attenuati0n)。而解調訊號則可表示為V⑴與、/c‘w (convolution)之絕對值,即Do not find its maximum value, and finally find the maximum value corresponding to the slope of the curve of the bandpass frequency slope attenuation on the full logarithmic coordinates. This is the singularity analysis value of the pulse hit point. Demodulation analysis function: Kfc, fw (0 =-^ ~ β ^ \ βηπ {ί €-^ JTCt) where: / c is the center frequency of the band-pass band; long is the bandwidth of the band-pass band; (slope convolution α is a proportional parameter that can be used to adjust the attenuation rate of the band-pass band (attenuati0n). The demodulated signal can be expressed as the absolute value of V⑴ and / c’w (convolution), that is,

a、fc、fw (叫 {t-τ) άτ\ 並且存在一常數C滿足下式 其令指數0C為雜峨之奇紐分析值; ¥取上式,人(/)之最大極限值時,則 13 1238926 若將上式對應不同比例參數β,並分別求取其最大值,則 此最大值對應比例參數β在全對數座標上之曲線斜率,即 為此脈衝敲擊點的奇異性分析值,可表示為: "=%,,/逆 _ log 〜2,Λ,九(0 1〇g «ι -l〇ga2 當貫際應用於軸承之奇異性分析上印證時,取中心頻率 /c=9 kHz與頻寬4=2 kHz,並取fll=2-8與β2=2·7,則可獲 得正常情況下與損壞情況下之不同奇異性分析值;為將上 述理論分析實現於電腦系統中,以下將提出其執行方法· 1·基於電腦之規格特性設定資料擷取之取樣時間τ、比例 參數α、頻寬/w與中心頻率/c之範圍,則波形函數長 度為2KT之解調變函數可表示為: 又a, fc, fw (called {t-τ) άτ \ and there is a constant C that satisfies the following formula and makes the index 0C be the analysis value of miscellaneous strangeness; ¥ When taking the above formula, the maximum limit of person (/), Then 13 1238926 If the above formula corresponds to different proportional parameters β and find their maximum values respectively, this maximum value corresponds to the slope of the curve of the proportional parameter β on the full logarithmic coordinates, that is, the singularity analysis value of the pulse hit point , Can be expressed as: " =% ,, / inverse_ log ~ 2, Λ, nine (0 1〇g «ι -l〇ga2 when applied to the bearing singularity analysis, the center frequency / c = 9 kHz and bandwidth 4 = 2 kHz, and fll = 2-8 and β2 = 2 · 7, different singularity analysis values under normal conditions and damage conditions can be obtained; in order to realize the above theoretical analysis in In the computer system, the following implementation method will be proposed: 1. Set the sampling time τ, scale parameter α, bandwidth / w and center frequency / c based on the specifications of the computer. The length of the waveform function is 2KT. The demodulation function can be expressed as:

I π kTI π kT

e(^)\ej2"(fc^)kTe (^) \ ej2 " (fc ^) kT

其中灸(-K= A: == K)為解調分析函數圖形中的某一點位置· 2·就一帶通頻率,若中心頻率/c、頻寬忍,並取一序列 比例參數A、化、…知,則解調變分析矩陣可表示為 14 1238926 Η (η)χ(2Κ+1) W-k)W-k) W—K) \,/^(—κ+1)…弋‘九(κ) \,Λ,九(—Κ + 1)…弋‘九(Κ)W—κ+1) ..· WK)Among them, moxibustion (-K = A: == K) is the position of a certain point in the demodulation analysis function graph. 2. It is a bandpass frequency. If the center frequency / c and bandwidth are tolerated, take a sequence ratio parameter A, , ..., the demodulation analysis matrix can be expressed as 14 1238926 Η (η) χ (2Κ + 1) Wk) Wk) W—K) \, / ^ (— κ + 1) ... 弋 'nine (κ) \, Λ, nine (—K + 1) ... 弋 'nine (K) W—κ + 1) .. · WK)

κ 其中4表示第〆個頻帶之解調分析向量,其比例參數為 ,其中1仝灸幺η ;κ where 4 represents the demodulation analysis vector of the 〆th frequency band, and its scale parameter is, where 1 is the same as moxibustion 幺 η;

3. 對一機械振動訊號序列v(m),0= m = Μ,其一序列解調 變分析則可表示為: absA * v) abs(石2 *v) abs(An * v) 其中心表示由第〆個比例參數%所獲致之解調訊號,其中 ΚΑ:’<η,*則表示向量褶積運算,而abs()則是取其向量中每 一元素之絕對值; 4. 取解調振動訊號之最大值為: Λ ,Μ+2Κ+13. For a mechanical vibration signal sequence v (m), 0 = m = Μ, and a sequence of demodulation analysis can be expressed as: absA * v) abs (石 2 * v) abs (An * v) its center Represents the demodulated signal obtained by the second scale parameter%, where κ: '< η, * represents the vector convolution operation, and abs () is the absolute value of each element in its vector; 4. The maximum value of the demodulated vibration signal is: Λ, Μ + 2Κ + 1

AA

15 1238926 V max(A) 一 x2 max(e2) 入 —max(〇 5.最後求讀_參數β在全對數座標上之曲線斜率,此 振動訊號之奇異性分析值為: 钓 l〇g«w -l〇ga%+1 其中&與W為全對數座標上分卿應_參數α邀 β/η+7之相鄰兩點 所’、 6·由機械糸統之臨界值與步驟5 統之運作狀況 中之奇異性分析值判斷機械系15 1238926 V max (A)-x2 max (e2) input—max (〇5. Finally read _ the slope of the curve of the parameter β on the full logarithmic coordinates, the singularity analysis value of this vibration signal is: fishing 10 g « w -l〇ga% + 1 where & and W are full logarithmic coordinates, and the two should be _parameter α invite β / η + 7 adjacent two points', 6. The critical value by the mechanical system and step 5 Singularity analysis value in the operating condition of the system

Ss界值為機械系統之運作正常; α〈臨界值為機械彡統有發生初期損壞; 其中“臨界值’’的設定是依不同機械元件而定; 今為使本發明為達上揭目的及其所採用之技術手段能令: 解,係採—最佳之測量值,其執行方法如下: L取樣關τ取祕電腦頻設備,就—般軸訊號量測6 圍,可以選擇T=l/25000秒,· μ之帶賴帶k解的改_是意味著選擇機齡統之妓 振模式中的不同頻帶,就—般振動訊號量_翻,可以選 16 ^ 1238926 擇[3000, 12000]; 有關,一般可設定範圍 3·丸之帶通頻帶頻寬的設定與系統特性 為[2000, 3000]; 4·解調函數波形長度可取375筆窨粗Β上 军貝料,即為2KT=〇 〇15秒; 至於比例參數與可分別設定 又疋為1〇8與1〇-7用以調整一 適當之帶通頻帶衰減斜率; 以上五項參數在使用時均為選定之固定來數· 6.當應用於軸承之奇異性分析時,取I·麵秒,中心頻率 Θ kHz與頻紅=2此,解調函數波形長度可取π筆資 料’並取㈣與㈣,則可獲得正常情況下與不同元件損 壞情況下之奇異性分(如第八圖所示)。 如第八騎示,由分可知,若設定^2.Wrf7,The boundary value of Ss is the normal operation of the mechanical system; α <the critical value is the initial damage of the mechanical system; the setting of the "critical value" is determined by different mechanical components; to make the present invention achieve the purpose of disclosure and The technical method used by it can make: Solution, take the best measurement value, its execution method is as follows: L sampling off τ to get the secret computer frequency equipment, to measure the 6 range of the normal axis signal, you can choose T = l / 25000 seconds, the change of μ with the k solution means that the different frequency bands in the prostitute mode of the age system are selected, and the amount of vibration signal is turned, and you can choose 16 ^ 1238926. [3000, 12000 ]; Generally, the setting range is 3. The setting of the Maruno Band Pass Bandwidth and the system characteristics are [2000, 3000]; 4. The waveform length of the demodulation function can be 375 strokes, and it is 2KT. = 〇〇15 seconds; As for the proportional parameter and can be set to 108 and 1-7 respectively to adjust an appropriate attenuation slope of the bandpass frequency band; the above five parameters are selected fixed numbers when used · 6. When applied to bearing singularity analysis, take I · plane seconds, center frequency Θ kH z and frequency red = 2 Here, the length of the demodulation function waveform can take π data 'and take ㈣ and ㈣, you can get the singularity score under normal conditions and when different components are damaged (as shown in Figure 8). The eighth riding show, from the points, if you set ^ 2.Wrf7,

备軸承運狀奇賊分減她,,錄為正常運作丨但 :軸承運轉之可異性分析值α小於1()·2時則為軸承發生損 壞,並且_承補的籠重其奇異性分減諸會隨之變 小’因此確實可有關斷軸承之運作狀況。 =上所述’本發明實施例確能達到所預期之使用功 介^、所揭露之具體構造,不僅未曾見諸於職產品 懇請惠予審查,並賜准專利, 公開於中請前,誠已完全符合專利法之規定與要求, 麦依法提出發明專利之申請, 則實感德便。 17 1238926 【圖式簡單說明】 第一圖:為敲擊訊號所表示之指數衰減函數形式 第二圖:為調變現象由數百Hz的損壞診斷訊號調變至數 千Hz的示意圖 第三圖:本發明應用於第二圖調變訊號之結果示意圖 第四圖··正常軸承之三維頻譜圖 第五圖:滾子損壞軸承之三維頻譜圖(一) 第六圖:外環損壞軸承之三維頻譜圖(二) 第七圖:内環損壞軸承之三維頻譜圖(三) 第八圖:本發明所分析之振動訊號之奇異分析值表 18Prepare the bearing for the strange situation and subtract her, and record it as normal operation 丨 But: when the analysis of the bearing's heterogeneity value α is less than 1 () · 2, the bearing is damaged, and the singularity score Minus will become smaller accordingly, so it can indeed be related to the operating condition of the broken bearing. = As stated above, the embodiments of the present invention can indeed achieve the expected use of the power ^, and the specific structure disclosed, not only has not been seen in the product, please be reviewed, and granted a patent, published before the request, sincerely It has fully complied with the provisions and requirements of the Patent Law. Mai Mai's application for an invention patent is a real virtue. 17 1238926 [Brief description of the diagram] The first picture: the exponential attenuation function form represented by the percussion signal. The second picture: the diagram of the modulation from the damage diagnosis signal of hundreds of Hz to the thousands of Hz. : The result of the application of the present invention to the modulation signal of the second figure. The fourth figure. The three-dimensional spectrum chart of the normal bearing. The fifth figure: The three-dimensional spectrum chart of the roller damaged bearing. The sixth figure: The three-dimensional spectrum of the damaged outer ring bearing. Spectrum chart (II) Figure 7: Three-dimensional spectrum chart of inner ring damaged bearing (III) Figure 8: Singular analysis value of vibration signal analyzed by the present invention Table 18

Claims (1)

1238926 拾、申請專利範圍: 1· 一種振動訊號之奇異性診斷處理方法,主要係藉由 取一時域振動訊號,以一解調分析函數進行訊號處 理’並設定帶通頻率之一序列衰減率(slope attenuation)以分別進行高頻解調,將所獲得的一 序列解調訊號分別求取其最大值,最後求取此最大 值對應帶通頻率衰減率(slope attenuation)在全 對數座標上之曲線斜率,此即為脈衝敲擊點的奇異 性分析值; 藉此’利用機械自動診斷系統之發展提出脈衝 振動訊號之奇異性分析法,可由時域振動訊號直接 判定是否為機械損壞之脈衝振動,並可利用一數值 形式來表示機械系統之運作狀況,如此將可方便電 腦執行判斷,以符合機械損壞偵測自動化之需求及 達到方便判讀機械系統運作狀態者。 〜種振動訊號之奇異性分析值之執行方法,其包含: (1) 設定解調分析函數之參數值:依電腦之規格特性 設定資料擷取之取樣時間T、頻寬A與中心頻率 /c之範圍’並設定兩適當選擇之比例參數31、奶, 且取波形函數長度為2KT ; (2) 獲得解調變分析矩陣:解調變分析矩陣可表示 [(2)χ(2Κ+1)1238926 Patent application scope: 1. A singularity diagnosis and processing method of vibration signals, which mainly uses a time-domain vibration signal and a demodulation analysis function to perform signal processing 'and sets a sequence attenuation rate of a bandpass frequency ( slope attenuation) to perform high-frequency demodulation separately, obtain the maximum value of a sequence of demodulated signals obtained, and finally obtain the curve corresponding to the band attenuation frequency attenuation rate (slope attenuation) on the full logarithmic coordinates. Slope, which is the singularity analysis value of the impulse hit point; By this, the singularity analysis method of the pulse vibration signal is proposed using the development of the automatic mechanical diagnosis system. The time-domain vibration signal can directly determine whether it is a mechanical damage pulse vibration. A numerical form can be used to indicate the operating status of the mechanical system. This will facilitate the computer to perform judgments to meet the needs of mechanical damage detection automation and to facilitate the interpretation of the operating status of the mechanical system. ~ A method for performing the singularity analysis value of the vibration signal, including: (1) Setting the parameter value of the demodulation analysis function: setting the sampling time T, bandwidth A, and center frequency / c of the data acquisition according to the specifications and characteristics of the computer Range 'and set two appropriately selected scaling parameters 31, milk, and take the length of the waveform function as 2KT; (2) obtain the demodulation analysis matrix: the demodulation analysis matrix can be expressed as [(2) χ (2Κ + 1) 、八,九(-Κ) \,Λ,九(-Κ+1) &quot;·\,Λ,夂(Κ) \,/c,又(-κ) (-Κ+1) ·•人,Λ,九(Κ)一 (3)獲得機械振動訊號序列 &lt;瓜)之解調變分析訊號: 19 1238926 解調變後之訊號表示為: λ _ ei y 2, Μ +2Κ +1 a e2 abs(&amp; *v) abs(^2 *v) 其中*則表示向量褶積運算,而abs()則是取其向 量中每一元素之絕對值; (4)獲得解調振動訊號之最大值為: χι max(^) max(e2) (5) 獲得振動訊號之奇異性分析值為: —log xr log x2 cx—— log axAog a2 (6) 由奇異性分析值判斷機械系統之運作狀況為: 2臨界值為機械系統之運作正常; 〈臨界值為機械系統有發生初期損壞; 其中“臨界值”的設定是依不同機械元件而定; 藉此,以其分析出之奇異性分析值表現於電腦 中,而自動做出判斷初步狀況,當已有初步損壞時再 由專業人員來進一步確認損壞並判斷機械損壞之元 件,這將更能符合目前製造業自動化與節省製造成本 的需求。 3.如申請專利範圍第2項所述振動訊號之奇異性分析 值之執行方法,其中: 20 (1) 取樣時間T取決於電腦硬體設備,就一般振動訊 號量測的範圍,可以選擇T=l/25000秒; (2) Λ之帶通頻帶中心頻率的改變則是意味著選擇 機械系統之共振模式中的不同頻帶,就一般振動 訊號量測的範圍,可以選擇[3000,12000]; (3) Λ之帶通頻帶頻寬的設定與系統特性·有關,一般 可設定範圍為[2000,3000]; (4) 解調函數波形長度可取375筆資料,即為 2ΚΤ=0. 015 秒; (5) 至於比例參數况與决可分別設定為10_8與10_7 用以調整一適當之帶通頻帶衰減斜率; 以上五項參數在使用時均為選定之固定參數; (6) 當應用於轴承之奇異性分析時,取T=l/25000 秒,中心頻率/c=9 kHz與頻寬焱=2 kHz,解調函 數波形長度可取375筆資料,並取况=2_8與 52=2^,則可獲得正常情況下與不同元件損壞情 況下之奇異性分析值者。 21, Eight, nine (-Κ) \, Λ, nine (-Κ + 1) &quot; · \, Λ, 夂 (Κ) \, / c, and (-κ) (-Κ + 1) · • person, Λ, nine (Κ) one (3) obtained the demodulation analysis signal of the mechanical vibration signal sequence &lt; melon): 19 1238926 The demodulated signal is expressed as: λ _ ei y 2, Μ + 2Κ +1 a e2 abs (&amp; * v) abs (^ 2 * v) where * represents the vector convolution operation, and abs () takes the absolute value of each element in its vector; (4) obtains the maximum value of the demodulated vibration signal The value is: χι max (^) max (e2) (5) The singularity analysis value of the vibration signal is obtained: —log xr log x2 cx—— log axAog a2 (6) The singularity analysis value is used to judge the operating status of the mechanical system For: 2 The critical value is the normal operation of the mechanical system; <The critical value is the initial damage of the mechanical system; the setting of the "critical value" is determined by different mechanical components; by this, the singularity analysis value analyzed by it It is displayed in the computer, and the preliminary condition is automatically judged. When there is preliminary damage, professionals will further confirm the damage and judge the mechanically damaged components, which will be more consistent. Meet the current needs of manufacturing automation and saving manufacturing costs. 3. The method of performing the singularity analysis value of the vibration signal as described in item 2 of the scope of the patent application, where: 20 (1) The sampling time T depends on the computer hardware equipment. For the range of general vibration signal measurement, you can choose T = l / 25000 seconds; (2) The change of the center frequency of the bandpass band of Λ means that different frequency bands in the resonance mode of the mechanical system are selected. For the range of general vibration signal measurement, you can choose [3000, 12000]; (3) The setting of the bandpass bandwidth of Λ is related to the system characteristics. Generally, the setting range is [2000, 3000]; (4) The demodulation function waveform length can take 375 records, which is 2KTT = 0.015 seconds ; (5) As for the ratio parameter condition and must be set to 10_8 and 10_7, respectively, to adjust a proper attenuation bandpass slope; the above five parameters are selected fixed parameters when used; (6) when applied to bearings For the singularity analysis, take T = l / 25000 seconds, center frequency / c = 9 kHz and bandwidth 焱 = 2 kHz, and the demodulation function waveform length can take 375 data, and the condition = 2_8 and 52 = 2 ^, You can get the singularity under normal conditions and when different components are damaged Sex analysis value. twenty one
TW93115744A 2004-06-02 2004-06-02 Singularity analysis method and its processes to apply in the defect diagnosis of mechanical components TWI238926B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
TW93115744A TWI238926B (en) 2004-06-02 2004-06-02 Singularity analysis method and its processes to apply in the defect diagnosis of mechanical components

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
TW93115744A TWI238926B (en) 2004-06-02 2004-06-02 Singularity analysis method and its processes to apply in the defect diagnosis of mechanical components

Publications (2)

Publication Number Publication Date
TWI238926B true TWI238926B (en) 2005-09-01
TW200540588A TW200540588A (en) 2005-12-16

Family

ID=37001137

Family Applications (1)

Application Number Title Priority Date Filing Date
TW93115744A TWI238926B (en) 2004-06-02 2004-06-02 Singularity analysis method and its processes to apply in the defect diagnosis of mechanical components

Country Status (1)

Country Link
TW (1) TWI238926B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8807049B2 (en) 2010-08-25 2014-08-19 Nippon Steel & Sumitomo Metal Corporation Vibration suppression device for railway vehicle

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8807049B2 (en) 2010-08-25 2014-08-19 Nippon Steel & Sumitomo Metal Corporation Vibration suppression device for railway vehicle
TWI449642B (en) * 2010-08-25 2014-08-21 Nippon Steel & Sumitomo Metal Corp Vibration Suppression Device for Railway Vehicles

Also Published As

Publication number Publication date
TW200540588A (en) 2005-12-16

Similar Documents

Publication Publication Date Title
CN108955734B (en) Cavity length demodulation method of optical fiber F-P temperature/pressure composite sensor
CN107941510B (en) Extracting method based on the angularly Rolling Bearing Fault Character of dual sampling
Chen et al. Optimal frequency band selection using blind and targeted features for spectral coherence-based bearing diagnostics: A comparative study
CN109855874B (en) Random resonance filter for enhancing detection of weak signals in vibration assisted by sound
CN103558029A (en) Abnormal engine sound fault on-line diagnostic system and diagnostic method
CN109946081B (en) Fault diagnosis method for rolling bearing slipping under variable rotating speed
CN101290347A (en) Method for obtaining still acoustic source acoustic field image by regular sound array and single video camera
CN104958064A (en) Wearable arteriosclerosis detector and pulse wave velocity detecting method
TWI572327B (en) Apparatus, computer program product and computer readable medium using audio signal for detection and determination of narrowing condition of fluid pipe
CN106983524A (en) A kind of parameter and its measuring method for reflecting that biological tissue is abnormal
Jiang et al. Differential spectral amplitude modulation and its applications in rolling bearing fault diagnosis
CN108398260B (en) Method for quickly evaluating instantaneous angular speed of gearbox based on mixed probability method
Lv et al. Generalized synchroextracting-based stepwise demodulation transform and its application to fault diagnosis of rotating machinery
TWI238926B (en) Singularity analysis method and its processes to apply in the defect diagnosis of mechanical components
CN114486263A (en) Noise reduction and demodulation method for vibration signal of rolling bearing of rotary machine
CN109163894B (en) Running-in state identification method based on friction temperature signal
CN114486252B (en) Rolling bearing fault diagnosis method of vector mode maximum envelope
CN106898031A (en) A kind of method of reeling off raw silk from cocoons for high-precision optical fiber measurement
CN112857806B (en) Bearing fault detection method based on moving window time domain feature extraction
CN114858453A (en) Bearing composite fault diagnosis method
CN112036270B (en) Vibration monitoring method for early damage state of rolling bearing
CN211478167U (en) Fault diagnosis device and fault diagnosis system
CN107271184A (en) The kernel regression decomposition method and system of a kind of rolling bearing fault diagnosis
JPH10216096A (en) Biological signal analyzing device
CN117292706B (en) Pipeline gas leakage sound diagnosis method, system, equipment and medium

Legal Events

Date Code Title Description
MM4A Annulment or lapse of patent due to non-payment of fees