[go: up one dir, main page]

TWI300889B - An envelope detection method and a sideband reduction method - Google Patents

An envelope detection method and a sideband reduction method Download PDF

Info

Publication number
TWI300889B
TWI300889B TW94134443A TW94134443A TWI300889B TW I300889 B TWI300889 B TW I300889B TW 94134443 A TW94134443 A TW 94134443A TW 94134443 A TW94134443 A TW 94134443A TW I300889 B TWI300889 B TW I300889B
Authority
TW
Taiwan
Prior art keywords
signal
frequency
envelope
vibration
function
Prior art date
Application number
TW94134443A
Other languages
Chinese (zh)
Other versions
TW200715082A (en
Inventor
Yu Tai Shen
Original Assignee
Univ Southern Taiwan Tech
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 Univ Southern Taiwan Tech filed Critical Univ Southern Taiwan Tech
Priority to TW94134443A priority Critical patent/TWI300889B/en
Publication of TW200715082A publication Critical patent/TW200715082A/en
Application granted granted Critical
Publication of TWI300889B publication Critical patent/TWI300889B/en

Links

Landscapes

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

Description

1300889 1 巇 九、發明說明: 【备明所屬之技術領域】 ▲。本發明係有關於一種包絡訊號之取得方法及其側頻消除 «處理方法,其主要將振動訊號經過包絡線處理後,再以 侧頻请除訊聽理方法來降低、絲包絡訊_譜之側頻現 象,進而運用於損壞機械系統之訊號分析上,來達成 系統損壞與否的判讀之首先發明應用者。 * 【先前技術】 、、二偶锹系既兀件運作狀況之診斷分析過程,一般可 分為資料#貞取、訊號處理錢譜圖等三階段。首先利^ 動感測H(如位移計、速度規或加速規)將振動之物理旦^ 換為電壓形式’再經由放大器將電壓適度放大後,透二 比轉,位轉換介面將電壓輸入電腦中進行數位式訊號戍 理,最後轉動頻譜m現動訊叙頻率特性,= 由頻譜的鮮分布與模式(patteni)來_機 元: 的損壞與否。 騎470件 由於機械系統元件的損壞將造成系統運作產生 動’並由於機械系統巾錄運動元件均屬於旋轉元件, 此類元件上發生損壞將會產生·性敲擊,診機= 元件之運作狀況則在於偵測此—週期性振動訊號的發=械 但由於在機㈣統振動訊號中滲有相# 件損壞較㈣财產生之振祕絲錄小,=將= 雜訊所掩蓋而_察覺,只有到發生相當嚴重損壞時Γί :=?:件損壞已變嚴重’所d 因“成嚴重故障停止運轉。此種視象為—般頻譜分= 1300889 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。 本發明人根據上述之缺失,於先前已研發一種機械系 統元件運作狀況診斷之訊號處理方法,且於鈞局申請獲 准為第I 228646號發明專利,其包括有: 解調分析函數: -(-)2 z a (e 2 -e 2 其中=1300889 1 巇 九, invention description: [provided the technical field of its own] ▲. The invention relates to a method for obtaining an envelope signal and a method for processing the side frequency thereof. The method mainly comprises the following steps: after the vibration signal is processed by the envelope, and then the side frequency is removed, the signal is reduced, and the signal is reduced. The side frequency phenomenon, which is then applied to the signal analysis of the damaged mechanical system, is the first inventor to determine the damage of the system. * [Previous technology] The diagnostic analysis process of the operation status of the two components is generally divided into three stages: data #贞取, signal processing and money spectrum. Firstly, the sensing H (such as a displacement meter, a speed gauge or an accelerometer) converts the physical state of the vibration into a voltage form. Then, after the voltage is moderately amplified by the amplifier, the second conversion is performed, and the bit conversion interface inputs the voltage into the computer. Perform digital signal processing, and finally rotate the spectrum m to report the frequency characteristics, = the damage caused by the fresh distribution of the spectrum and the pattern (patteni). If 470 pieces are damaged due to damage of the mechanical system components, and the moving parts of the mechanical system are all rotating components, damage will occur on such components. Severe knocking, diagnosis machine = operation status of the components It is to detect this - the vibration signal of the periodic vibration signal, but because of the damage in the vibration signal of the machine (four), the damage of the phase is worse than that of the vibration of the fourth (4), the noise will be hidden, and the noise will be covered.只有ί :=?: The damage of the part has become serious. 'The damage caused by the serious failure has stopped. This kind of vision is the general spectrum score = 1300889 The difficulty in analyzing the mechanical vibration signal It is also a major lack of damage diagnosis and early warning effect in actual use. Based on the above-mentioned shortcomings, the inventors have previously developed a signal processing method for the diagnosis of the operation status of mechanical system components, and applied for approval as the first in the bureau. Invention patent No. 228646, which includes: Demodulation analysis function: -(-)2 za (e 2 -e 2 where =

/c為帶通頻帶之中心頻率; Λ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斜率 (slope attenuation); 且解調訊號則可表示為 v(t)與⑴褶積 (convolution)之絕對值,即 eajcjJ〇= iy(T)K/cjw(^T)dT 。 而上述所獲得之結果為一種振動訊號的包絡訊號,而 此包絡訊號之取得僅為其中的一種方式而已,故本發明人 隨科技發展之腳步,經研發製作及實驗之過程,特再提供 一種包絡訊號之取得方法及其側頻消除訊號處理方法,期 可助於機械系統損壞與否的判讀。 【發明内容】 本發明係有關於一種包絡訊號之取得方法及其側頻 消除訊號處理方法,其主要係將振動訊號經過包絡線處理 即得包絡訊號(envelope signal)。又,因該包絡訊號之 1300889 頻譜在其特徵頻率的兩側常會有等頻分佈之侧頻 (sidebands of eqUai S{)acing frequency)出現,在某些 運作情況下其侧頻能量則可能會升高,因而影響特徵頻率 是否明確王現之正確判讀,因此再將此一包絡訊號經過側 頻消除訊號處理後來降低甚至消除包絡訊號頻譜上之側 頻現象,運用於損壞機械系統之訊號分析上,將可有效抑 制該頻譜上側頻之能量以更明確呈現出特徵頻率,如此可 更進一步助於機械系統損壞與否的判讀者。 【實施方式】 而為令本發明之技術手段能夠更完整且清楚的揭露, 兹請一併參閱所附圖式及圖號,並詳細說明如下: 首先,本發明一種包絡訊號之取得方法及其侧頻消除 訊號處理方法,主要是將振動訊號經過包絡線處理而得包 絡訊號,其中: 該包絡線處理主要係將振動訊號經過帶通渡波器 (bandpass filter)予以濾波處理,再將濾波訊號映射至以 共振頻率(resonant frequency)為其振動頻率之三角函數 基底上(sinusoidal function basis),由於敲擊頻率遠小 於共振頻率,因此可以步階函數(stepwise function)來近 似該包絡線訊號,最後以參數估測方法求取其映射之係數 對’所得之係數對之均方根(root mean square)即得該振 動訊號之包絡訊號(envelope signal)。 而當包絡訊號運用於損壞機械系統之訊號分析上,其 包絡訊號之頻譜在其特徵頻率的兩側常會有等頻分佈之側 頻(sidebands of equal spacing frequency)出現,將此 一包絡訊號經過側頻消除訊號處理方法來降低、消除包絡 1300889 訊號頻譜之側頻現象,即可更明確呈現出特徵頻率 該侧頻消除訊號處理方法主要係藉由包絡訊號在損壞,、 中其損壞的特徵頻率遠大於其側頻,因此產生此側頻現= 之函數玎5又為步階函數(stepwise function),故在任—疗 壞敲擊時間週期中步階函數為常數,並由於脈衝敲擊為 數衰減函數且最大值為丨,故將該損壞敲擊時間週期中之包 絡訊號除以在該敲擊週期之最大值即可得脈衝敲擊函數匕 如此即可消除損壞元件頻譜之侧頻現象。 。 藉由上述,將振動訊號經過包絡線處理而得包絡訊 5虎’再利用包絡訊號經過側頻消除訊號處理後來降低甚至 =除,說頻譜上之侧頻現象,當運用於損壞機械系統 呈虎刀析上’可有效抑制該頻譜上側頻之能量以更明確 其特徵頻率’對於正常機械系統之訊號分析,則發現 處理^之頻率間的消長並無明顯影響;由於侧頻消除訊號 步 去可突顯損壞機械系統之特徵頻率,因而可更進一 =於機械系統損壞與否的判讀者。 、取得句故却缺朴令找人> 〇上上一…匕。、…义、办/c is the center frequency of the bandpass band; Λ is the bandwidth of the bandpass band; β is the proportional parameter which can be used to adjust the slope attenuation of the bandpass band; and the demodulation signal can be expressed as v(t) and (1) The absolute value of convolution, ie eajcjJ〇= iy(T)K/cjw(^T)dT. The result obtained above is an envelope signal of a vibration signal, and the acquisition of the envelope signal is only one of them. Therefore, the inventor has provided a kind of process along with the development of technology and the process of research and development and experiment. The method of obtaining the envelope signal and the method for processing the sideband cancellation signal can help the interpretation of the mechanical system damage. SUMMARY OF THE INVENTION The present invention relates to an acquisition method of an envelope signal and a method for processing a sideband cancellation signal, which mainly performs an envelope signal by an envelope signal to obtain an envelope signal. Moreover, since the 1300889 spectrum of the envelope signal often has sidebands of eqUai S{) aging frequency on both sides of its characteristic frequency, the sideband energy may rise in some operating conditions. High, thus affecting whether the characteristic frequency is clear and correct. Therefore, after the envelope signal is processed by the sideband cancellation signal, the side frequency phenomenon on the spectrum of the envelope signal is reduced or even eliminated, and the signal analysis applied to the damaged mechanical system will be Effectively suppressing the energy of the side frequency on the spectrum to more clearly present the characteristic frequency, which can further help the reader of the mechanical system to be damaged or not. [Embodiment] In order to make the technical means of the present invention more complete and clear, please refer to the drawings and drawings, and explain in detail as follows: First, the method for obtaining an envelope signal and the method thereof The sideband cancellation signal processing method mainly includes the envelope signal processed by the vibration signal, wherein: the envelope processing mainly filters the vibration signal through a bandpass filter, and then maps the filtered signal. To the sinusoidal function basis with the resonant frequency as the vibration frequency, since the tapping frequency is much smaller than the resonant frequency, the stepwise function can be used to approximate the envelope signal, and finally The parameter estimation method obtains the coefficient of the mapping pair and the root mean square of the obtained coefficient is the envelope signal of the vibration signal. When the envelope signal is applied to the signal analysis of the damaged mechanical system, the spectrum of the envelope signal often has sidebands of equal spacing frequency on both sides of the characteristic frequency, and the envelope signal passes through the side. Frequency cancellation signal processing method to reduce and eliminate the side frequency phenomenon of the envelope 1300889 signal spectrum, the characteristic frequency can be more clearly presented. The side frequency cancellation signal processing method mainly relies on the envelope signal being damaged, and the characteristic frequency of the damage is large. At its side frequency, the function 玎5 that produces this side frequency = is again a stepwise function, so the step function is constant in the any-time-shock time period, and the attenuation function is due to the pulse tapping. And the maximum value is 丨, so the envelope signal in the damage tapping time period is divided by the maximum value of the tapping period to obtain a pulse tapping function, so that the side frequency phenomenon of the spectrum of the damaged component can be eliminated. . By the above, the vibration signal is processed by the envelope to obtain the envelope signal 5 tiger's reuse envelope signal through the sideband cancellation signal processing and then reduce or even = divide, said the side frequency phenomenon on the spectrum, when applied to the damaged mechanical system is the tiger The knife analysis can effectively suppress the energy of the side frequency of the spectrum to more clearly define its characteristic frequency. For the signal analysis of the normal mechanical system, it is found that the frequency between the processing and the frequency has no significant effect; since the side frequency cancellation signal step can be It highlights the characteristic frequency of the damaged mechanical system, and thus can be further improved to the reader of the mechanical system damage. I got the sentence but I was short of the order to find someone. 〇上上一...匕. Righteousness

〜⑺ ㈣ c〇s(2#(卜r) + <9,J dr +~(7) (4) c〇s(2#(卜r) + <9,J dr +

----(1) 方程式中第一項積分與第二項積分是分別用來 1300889 描述糸統中損壞元彼 Η矣-件與正常元件之振動訊號。 md疋表不損壞數量; dm(t)是表示損壞的脈衝; 細擊相_能量因素; h表示振動傳遞路㈣函數; nr疋表不正常滾子的數量. 办⑺是表示滾子的幾何形狀; H表示與幾何形狀相關的能量因素; …如疋表示振動傳遞路徑的函數; 另外,σ /與/公它丨 統振動模的數量=示其指數頻f與共振頻率、,ζ則為系 ^ ^ 〜⑺與θ/η⑺疋振幅调變的初始相角。 上第7模之振動之包絡訊號則為 .⑺+Σ (Ο (2) ^〇=E«j〇,w(0 ’----(1) The first integral and the second integral in the equation are used to describe the vibration signal of the damaged element and the normal component in the system. Md疋 table does not damage the quantity; dm(t) is the pulse indicating damage; fine stroke phase_energy factor; h means vibration transmission path (4) function; nr疋 indicates the number of abnormal rollers. (7) is the geometry of the roller Shape; H represents the energy factor associated with the geometry; ... such as 函数 represents the function of the vibration transmission path; in addition, the number of σ / / / public system vibration mode = its exponential frequency f and resonance frequency, ζ is The initial phase angle of the amplitude modulation of ^ ^ ~ (7) and θ / η (7) 。. The envelope signal of the vibration of the seventh mode is .(7)+Σ (Ο (2) ^〇=E«j〇,w(0 ’

Sn*^rt *aln '、中me ^,厂二⑽邮,1//“),而mod(t,1//办)則 表示’、餘數運异’ 4則為損壞敲擊的頻率。 一而由上述所得之包絡訊號,其包絡訊號之頻譜在其特 徵頻率的兩侧常會有等頻分布之侧頻(sidebands of q Pacing frequency)出現,在某些運作情況下盆侧 Π:可能會升高,因而影響特徵頻率是否明確呈現之 21 J此需要再將包絡訊號頻譜上之侧頻能量加以 後上之㈣明確呈現出特徵頻率;因此,需將包絡訊號頻 :消除=象以侧頻消除訊號處理方法予以消除,該側 知,二;二,理方法,其主要由包絡訊號之理論分析可 在才貝展料承中艺U相對遠小於gu 幾可勿政· tk n=1 W=1 Ubn\l)。人 心令,由於%(〇為損壞敲擊函數,在頻譜上呈現出損 9 1300889 壞特徵^率,而以〇與_則可呈現出其側頻,故⑽_ ίϊ:ΐίΓΓ)與㈣的頻率值,因此可假設u〇與⑽ 為步階函數(stepwlse function),所以在任一損壞 時間週⑺與可視為常數,並由於_為指數衰減 函數且最大值A 1’故在魏擊㈣之最大值即應為 ί”) ’故將該損壞敲擊時間週期中之包絡訊號除 即可得損壞敲擊函數W,如此即可消^ 之側頻現象°然而’對於正常機械系統之訊號分析 現其頻4之頻率間的消長並無明顯影響。因而, 步突顯損壞機械系統之特徵頻率,將有助於機械系統壞鱼 否的^買,如第一圖所示為滾子損壞的振動頻譜示意圖Ί (a)為=絡訊號處理之振動頻譜,(_為包絡 過側頻消除訊號處理後之頻譜,可發現_可幾乎消除。 再對照第二圖所示,為正常軸承的頻譜示意圖,⑷為 經包絡訊號處理之振動頻譜,(_為包絡訊號經過側頻消 後之頻谱’可發現頻譜現象未有明顯側頻或側 頻減小現象。 另,對於上述第⑵式振動訊號之包絡線處理所取得 ===乂除可根據本發明人於先前之-種機械系統元 =作狀況衫斷之訊號處理方法來獲得之外,於本發明中 =出c訊號之取得方法,以更簡易且快速之包絡 線處理=讀得振動訊號之包絡訊號,其: 由yi)式經擷取其第/模之振動訊號,則可表示為 = i+,J+£^9a<(〇cos(2;r// ί+,η) = 〇1(Οο〇Β(2π/ι t) + c2(t)s^fl t) H=1 (3) 10 1300889 其中,與則為將振動訊號”⑼映射至以 共振頻率//為其振動頻率之三角函數基底上(sinus〇idal function basis)的係數。 由於C/( ί)與Cz( ί)退小於共振頻率,因此可設其 為一步階函數(stepwise function),故以離散的取樣點表 不則為 V/(〇 = ^(7)008(2^ ---) + ^(7)8111(2^ ~) ( * )Sn*^rt *aln ', medium me ^, factory two (10) post, 1//"), and mod (t, 1 / / office) means ', the remainder of the difference' 4 is the frequency of damage knock. As a result of the above-mentioned envelope signal, the spectrum of the envelope signal often has sidebands of q Pacing frequency on both sides of its characteristic frequency. In some operating situations, the basin side Π: may Elevation, thus affecting whether the characteristic frequency is clearly presented 21 J. This needs to add the side frequency energy on the spectrum of the envelope signal to (4) to clearly show the characteristic frequency; therefore, the envelope signal frequency needs to be: eliminate = like the side frequency Eliminate the signal processing method to eliminate, the side knows, two; two, the rational method, which is mainly analyzed by the theoretical analysis of the envelope signal, which can be relatively far less than the gu in the 贝 展 、 、 、 t t n n n =1 Ubn\l). The human heart orders, because of the % (〇 is the damage tap function, the spectrum shows a loss of 9 1300889 bad feature ^ rate, and 〇 and _ can show its side frequency, so (10)_ ίϊ: ΐίΓΓ) and (iv) the frequency value, so it can be assumed that u〇 and (10) are step functions (stepwlse function), In any damage time week (7) and can be regarded as a constant, and since _ is the exponential decay function and the maximum value A 1 ', the maximum value of the Wei strike (four) should be ί") 'So the damage is struck in the time period In addition to the envelope signal, the tapping function W can be damaged, so that the side frequency phenomenon can be eliminated. However, the signal analysis of the normal mechanical system has no significant influence on the frequency variation between the frequencies of the frequency 4. Therefore, the step of highlighting the characteristic frequency of the damaged mechanical system will help the mechanical system to eliminate the bad fish. For example, the vibration spectrum of the roller damage is shown in the first figure. (a) The vibration spectrum of the signal processing (_ is the envelope after the sideband cancellation signal processing, it can be found that _ can be almost eliminated. Then compare the second diagram, as shown in the spectrum diagram of the normal bearing, (4) is the vibration spectrum processed by the envelope signal, (_ The spectrum of the envelope signal after the side frequency cancellation can be found that the spectrum phenomenon has no obvious side frequency or side frequency reduction. In addition, the envelope processing of the vibration signal of the above (2) is obtained by === The inventor obtained the method for processing the signal signal in the prior art, and in the present invention, the method for obtaining the c signal is processed by the simpler and faster envelope = the vibration signal is read. The envelope signal, which is obtained by the yi) type, is represented by = i+, J+£^9a<(〇cos(2;r// ί+,η) = 〇1 (Οο〇Β(2π/ι t) + c2(t)s^fl t) H=1 (3) 10 1300889 where, and then The vibration signal "(9) is mapped to a coefficient of the sinus 〇idal function basis at the resonance frequency // as its vibration frequency. Since C/( ί) and Cz( ί) are smaller than the resonance frequency, it can be set It is a stepwise function, so the discrete sampling point table is not V/(〇= ^(7)008(2^ ---) + ^(7)8111(2^ ~) ( * )

其中/與Λ分別為振動訊號之取樣點與取樣頻率,而「則為包絡訊號上的求取點。將第(4)式表示成矩陣...... 邮-1)+1/ ^ ~Is~~ 咐一 1) + 2、 cos{2nft ο〇8(2π/ι h{kt -1)+1, h(k「l) + 2、 Is> -1)+3、~7s"""; sinQTr/j fs c\ (^/) c2 {ki) cos(2^· ft hkx Ύ V/W/-1)+1) νι(Κ^-1) + 2) ν/(Λ(勾〜l) + 3) v/(^)Where / is the sampling point and sampling frequency of the vibration signal, and "is the point of the envelope signal. The equation (4) is expressed as a matrix... Post-1) + 1 / ^ ~Is~~ 咐一1) + 2, cos{2nft ο〇8(2π/ι h{kt -1)+1, h(k"l) + 2, Is>-1)+3,~7s"""; sinQTr/j fs c\ (^/) c2 {ki) cos(2^· ft hkx Ύ V/W/-1)+1) νι(Κ^-1) + 2) ν/( Λ (hook ~ l) + 3) v/(^)

(5) 其中,力表示振動訊號F/(〇在第1個半週期上的點 其包絡訊號上的求取點對應示意圖,則如第三圖所示。 若簡化表示式,則可表示為 c2(f〇 \y^ Lxi ―⑹ 以線性表小平方法(linear least square analysis)求取 上式的係數對,則為 for h-2 f〇vh>2 (7)(5) where the force represents the vibration signal F / (〇 in the first half cycle of the point on the envelope signal corresponding to the point of view corresponding to the map, as shown in the third figure. If the simplified expression, it can be expressed as C2(f〇\y^ Lxi ―(6) Calculate the coefficient pair of the above formula by linear least square analysis, then for h-2 f〇vh>2 (7)

1300889 [cAL·)! r 1 JLwj·丄抓, cx (kt) .^) J ~{[Mk L· [Mk L2 ri[K ^ [Vik ^ ^ 。所得係數對之均方根(root mean s_e)為該振動訊 號之包絡訊號(envelope signal),即 (8) 請再參閱第四圖所示,為振動訊號之解調變訊號示意 圖,其(a)為振動訊號 ,(b)為所獲得的包絡訊號, (c)則為利用包絡訊號經由第(4)式所反求之還原的振動 訊號。 藉由上述,可知利用包絡訊號之頻譜在一般之特徵頻 率的兩側會有等頻分佈之側頻出現,而在某些運作情況下 其侧頻能量則可能會升高,因而影響特徵頻率所呈現之判 讀的正確性,因此,再將包絡訊號經過側頻消除訊號處理 後,對損壞機械系統之訊號分析上,可有效抑制該頻譜上 侧頻之能量以更明確呈現出特徵頻率,由於玎突顯損壞機 械系統之特徵頻率,因而可更進一步助於機械系統損壞與 否的判讀者。 綜上所述,本發明實施例確能達到所預期之使用功 效,又其所揭露之具體構造,不僅未曾見諸於同類產'"口 中,亦未曾公開於申請前,誠已完全符合專利法之規定與 要求,爰依法提出發明專利之申請,懇請惠予審查,並賜 准專利,則實感德便。 12 1300889 【圖式簡單說明】 第一圖:本發明之滾子損壞的振動頻譜示意圖 第二圖:本發明之正常軸承的頻譜示意圖 第三圖:本發明之包絡訊號上的求取點對應示意圖 第四圖:本發明之振動訊號之解調變訊號示意圖1300889 [cAL·)! r 1 JLwj·丄, cx (kt) .^) J ~{[Mk L· [Mk L2 ri[K ^ [Vik ^ ^ . The root mean s_e of the obtained coefficient is the envelope signal of the vibration signal, that is, (8) Please refer to the fourth figure, which is a schematic diagram of the demodulation signal of the vibration signal, which is (a) ) is a vibration signal, (b) is the obtained envelope signal, and (c) is a vibration signal restored by the envelope signal via the equation (4). From the above, it can be seen that the spectrum of the envelope signal has a side frequency of equal frequency distribution on both sides of the general characteristic frequency, and in some operating conditions, the sideband energy may increase, thereby affecting the characteristic frequency. The correctness of the interpretation is presented. Therefore, after the envelope signal is processed by the sideband cancellation signal, the signal analysis of the damaged mechanical system can effectively suppress the energy of the sideband on the spectrum to more clearly present the characteristic frequency. It highlights the characteristic frequency of damage to the mechanical system, and thus can further help the reader of the mechanical system damage. In summary, the embodiments of the present invention can achieve the expected use efficiency, and the specific structure disclosed therein has not been seen in the similar products, and has not been disclosed before the application, and has completely complied with the patent. The provisions and requirements of the law, the application for invention patents in accordance with the law, and the application for review, and the grant of patents, are really sensible. 12 1300889 [Simplified description of the drawings] First: Schematic diagram of vibration spectrum of roller damage according to the present invention. Second diagram: Schematic diagram of the spectrum of the normal bearing of the present invention. Third diagram: Corresponding diagram of the point of acquisition on the envelope signal of the present invention Fourth: Schematic diagram of the demodulation signal of the vibration signal of the present invention

1313

Claims (1)

1300889 十、申請專利範圍: 1· 一種包絡訊號之取得方法’主要是將振動訊號經過包絡 線處理而得包絡訊號,其包絡線處理係將振動訊號經過 帶通濾波器(bandpass filter)予以濾波處理,再將濾 波訊號映射至以共振頻率(resonant frequency)為其振 動頻率之三角函數基底上(sinusoidal function bas i s ),由於敲擊頻率遠小於共振頻率’因此可以步階 函數(stepwise function)來近似該包絡線訊號,最後 以參數估測方法求取其映射之係數對,所得之係數對之 均方根(root mean square)即為該振動訊號之包絡訊號 (envelope signal)者。 2. —種包絡訊號之侧頻消除訊號處理方法,其主要係將振 動訊號經過包絡線處理,而得包絡訊號(envelope signal);該包絡訊號之頻譜在其特徵頻率的兩侧常會 有等頻分佈之側頻(sidebands of equal spacing frequency)出現,因此需經過側頻消除訊號處理方法來 降低甚至消除包絡訊號頻譜上之侧頻現象,其中: ® 該侧頻消除訊號處理方法,主要藉由包絡訊號在損 壞轴承中其損壞的特徵頻率遠大於其侧頻,因此產生此 側頻現象之函數可設為步階函數(stepwise function),故在任一損壞敲擊時間週期中步階函數為 常數,並由於脈衝敲擊為指數衰減函數且最大值為1, 故將該損壞敲擊時間週期中之包絡訊號除以在該敲擊 週期之最大值即可得脈衝敲擊函數,如此即可消除損壞 元件頻譜之側頻現象者; 藉此’將其運用於損壞機械系統之訊號分析上,將 14 1300889 可有效抑制該頻譜上側頻之能量以更明確呈現出特徵 頻率,由於可突顯損壞機械系統之特徵頻率,而助於機 械系統損壞與否的判讀者。 3.如申請專利第1項所述之包絡訊號之取得方法,其中, 以數學模式來說明機械振動訊號之調變現象,其演算式 可表示為 L md v(,)=Σ (Σ Ιλ ⑺‘⑺〜⑺ ^ _—7) cos(2;r//(,-r)+0j d,+ /=1 m=l Σ tjn^ β-^^Ιπ^ί-^ + θ^ dr) n-l °° 而方程式中第一項積分與第二項積分來描述系統中 損壞元件與正常元件之振動訊號; md是表示損壞數量; dm(t)是表示損壞的脈衝; ςμ(1:)是表示與敲擊相關的能量因素; 是表示振動傳遞路徑的函數; nr是表示正常滾子的數量; gn(t)是表示滾子的幾何形狀; Qn(t)是表示與幾何形狀相關的能量因素; flln(t)是表示振動傳遞路徑的函數; 另外,與f i分別表示其指數頻率與共振頻率,L· 則為系統振動模的數量,而Ui)與(9 z„(〇是振幅調變的初 始相角; 由機械振動訊號之調變現象經擷取其第/模之振動,則可 表不為 151300889 X. Patent application scope: 1. The method for obtaining the envelope signal is mainly to process the vibration signal through the envelope to obtain the envelope signal, and the envelope processing system filters the vibration signal through a bandpass filter. And then mapping the filtered signal to a sinusoidal function bas is with a resonant frequency as the vibration frequency, since the tapping frequency is much smaller than the resonant frequency', it can be approximated by a stepwise function. The envelope signal is finally obtained by the parameter estimation method, and the root mean square of the obtained coefficient is the envelope signal of the vibration signal. 2. A sideband cancellation signal processing method for an envelope signal, which mainly processes the vibration signal through an envelope to obtain an envelope signal; the spectrum of the envelope signal often has an equal frequency on both sides of its characteristic frequency. The sidebands of equal spacing frequency appear, so the sideband cancellation signal processing method is needed to reduce or even eliminate the sideband phenomenon on the envelope signal spectrum. Among them: ® The sideband cancellation signal processing method mainly relies on the envelope The characteristic frequency of the signal in the damaged bearing is much larger than its side frequency, so the function of generating this side frequency phenomenon can be set as a stepwise function, so the step function is constant in any damage tapping time period. And since the pulse tapping is an exponential decay function and the maximum value is 1, the pulse tapping function can be obtained by dividing the envelope signal in the damage tapping period by the maximum value of the tapping period, thus eliminating the damage. The side frequency phenomenon of the component spectrum; by this, it can be applied to the signal analysis of the damaged mechanical system, and 14 1300889 can be effective. Frequency spectrum of the energy system on the side of the reader to judge more clearly exhibit the characteristic frequency, since distinguishing feature of the system frequency mechanical damage, and damage to facilitate mechanical system or not. 3. The method for obtaining an envelope signal according to claim 1, wherein the mathematical mode is used to describe the modulation phenomenon of the mechanical vibration signal, and the calculation formula can be expressed as L md v(,)=Σ (Σ Ιλ (7) '(7)~(7) ^ _—7) cos(2;r//(,-r)+0j d,+ /=1 m=l Σ tjn^ β-^^Ιπ^ί-^ + θ^ dr) nl °° The first integral and the second integral in the equation describe the vibration signal of the damaged component and the normal component in the system; md is the number of damages; dm(t) is the pulse indicating damage; ςμ(1:) is the representation The energy factor associated with the tap; is a function of the vibration transfer path; nr is the number of normal rollers; gn(t) is the geometry of the roller; Qn(t) is the energy factor associated with the geometry ; flln(t) is a function representing the vibration transmission path; in addition, fi represents its exponential frequency and resonance frequency, respectively, L· is the number of system vibration modes, and Ui) and (9 z„(〇 is amplitude modulation The initial phase angle; the vibration of the mechanical vibration signal is obtained by taking the vibration of the first mode, then it can be expressed as 15 1300889 咕)=Σ〜(0 心 W 〜Wc〇s(2;r/"+&) + f {t)cos(2nft ί+θ1η) m=l ^ gntqn>aln V / v w=l = ^(^)008(2^// 0 + ^(08111(2^ t) 其C;u)與oU)則為映射至以共振頻率為其振動頻 率之二角函數基底上(sinusoidal function basis)的係 數;若以離散的取樣點表示則為 V/(0 = q(/)008(2^ γ) + €1(])ύη(2π/ι γ) 其中7與Λ分別為振動訊號之取樣點與取樣頻率,而 /則為包絡訊號上的求取點,將所得之係數以離散的取樣 點表示成矩陣式,即1300889 咕)=Σ~(0 heart W ~ Wc〇s(2;r/"+&) + f {t)cos(2nft ί+θ1η) m=l ^ gntqn>aln V / vw=l = ^(^)008(2^// 0 + ^(08111(2^ t) C;u) and oU) is mapped to the sinusoidal function basis on the octave function of the resonant frequency Coefficient; if expressed as discrete sampling points, it is V/(0 = q(/)008(2^ γ) + €1(])ύη(2π/ι γ) where 7 and Λ are respectively samples of the vibration signal Point and sampling frequency, and / is the point of the envelope signal, the resulting coefficient is represented as a matrix by discrete sampling points, ie zo^{2nft cos{2nft cos(2^/z h{k—Y) + 1 ~~Is""" h(k-l) + 2^ """"Is~~> h(k-l)+3 ~~Is~~ ήϊί{2π^ Λ(灸一 1)+1) ~I~~ 如伽/严):)+¾峰/严严) ci (灸) C2 (众) vz(A(A:-1)+1) v,师-1) + 2) v^Kk-1)+3) vt (hk) r hk、 Q〇s(2nft —-) J s J s 其中,力表示濾波訊號〇在第左個半週期上的點;若 簡化表示式,則可表示為 [从丄2 :),丄 以線性最小平方法(linear least square analysis) 進行參數估測,可求取上式的係數對,則為 16 1300889 forh=2 < ::3 = (WL K L)_1 [A L K L,f〇rh>2 所得係數對之均方根(root mean square)為該振動訊 號之包絡訊號(envelope signal),I^P + c2{jf 〇Zo^{2nft cos{2nft cos(2^/zh{k—Y) + 1 ~~Is""" h(kl) + 2^ """"Is~~> h( Kl)+3 ~~Is~~ ήϊί{2π^ Λ(moxibustion-1)+1) ~I~~ 如伽/严):)+3⁄4峰/严严) ci (moxibustion) C2 (zhong) vz( A(A:-1)+1) v, division-1) + 2) v^Kk-1)+3) vt (hk) r hk, Q〇s(2nft —-) J s J s where force Indicates the point at which the filtered signal 〇 is on the left half of the cycle; if the expression is simplified, it can be expressed as [from 丄2:), and the parameter estimation is performed by linear least square analysis. For the coefficient pair of the above formula, it is 16 1300889 forh=2 < ::3 = (WL KL)_1 [ALKL,f〇rh>2 The root mean square of the obtained coefficient is the envelope of the vibration signal. Envelope signal, I^P + c2{jf 〇 1717
TW94134443A 2005-10-03 2005-10-03 An envelope detection method and a sideband reduction method TWI300889B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
TW94134443A TWI300889B (en) 2005-10-03 2005-10-03 An envelope detection method and a sideband reduction method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
TW94134443A TWI300889B (en) 2005-10-03 2005-10-03 An envelope detection method and a sideband reduction method

Publications (2)

Publication Number Publication Date
TW200715082A TW200715082A (en) 2007-04-16
TWI300889B true TWI300889B (en) 2008-09-11

Family

ID=45070080

Family Applications (1)

Application Number Title Priority Date Filing Date
TW94134443A TWI300889B (en) 2005-10-03 2005-10-03 An envelope detection method and a sideband reduction method

Country Status (1)

Country Link
TW (1) TWI300889B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI421484B (en) * 2009-06-23 2014-01-01 Univ Southern Taiwan Tech Envelope signals derived from application of morlet wavelet in gain of mechanical vibration signals and method for defect diagnosis of machines using exponential decay frequency of envelope signals

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI421484B (en) * 2009-06-23 2014-01-01 Univ Southern Taiwan Tech Envelope signals derived from application of morlet wavelet in gain of mechanical vibration signals and method for defect diagnosis of machines using exponential decay frequency of envelope signals

Also Published As

Publication number Publication date
TW200715082A (en) 2007-04-16

Similar Documents

Publication Publication Date Title
CN109946081B (en) A kind of fault diagnosis method for rolling bearing slip under variable speed
JPH10504107A (en) Digital signal processing of encoder signals for detecting resonance of rotating machinery
TW200936999A (en) Knock signal detection in automotive systems
WO2019200624A1 (en) Comprehensive evaluation method for pump flow induced vibration performance
CN108760310B (en) Fault Diagnosis Method of Stochastic Resonance Rolling Bearing Based on Novel Signal-to-Noise Ratio Index
CN110160765A (en) A kind of shock characteristic recognition methods and system based on sound or vibration signal
CN110763462A (en) A Time-varying Vibration Signal Fault Diagnosis Method Based on Synchronous Compression Operator
CN104729591B (en) A kind of signal processing method for the strong transient impact vibration of the anti-low frequency of vortex-shedding meter replaced based on data
CN102998119B (en) A kind of Method for Bearing Fault Diagnosis based on compound Q factor base algorithm
TW201221929A (en) Real-time detection system and the method thereof
JP2017129583A (en) Vibration monitoring system
CN113865867A (en) A Bearing Fault Diagnosis Method Based on Singular Value Decomposition of Amplitude Characteristics
CN103116032A (en) Method and device for acquiring rotating speed of wind generating set
CN110032776A (en) Blade dynamic strain field and displacement field reconstructing method and its system based on Modes Decoupling
CN111537881A (en) Fault diagnosis method, device and equipment for asynchronous motor and readable storage medium
CN114545079A (en) Resonance detection method, device and equipment for photovoltaic grid-connected system and storage medium
CN114027825B (en) Respiratory signal acquisition method and device and computer equipment
CN117109923A (en) A rolling bearing fault diagnosis method and system
CN104677486B (en) The aero-engine vibration signal Method for Phase Difference Measurement reconstructed based on tacho-pulse
TWI300889B (en) An envelope detection method and a sideband reduction method
CN114812794A (en) Method for correcting blade tip timing parameter identification error caused by frequency sweep rate
CN103913271B (en) Method for extracting dynamic unbalance signals of rotor at non-stable rotational speed
JP3152247B2 (en) Reactor power monitoring method and apparatus
CN115014765A (en) Method for extracting fault characteristics of rolling bearing retainer through acoustic signals
CN114486252B (en) Rolling bearing fault diagnosis method of vector mode maximum envelope

Legal Events

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