CN101178841A - Bearing fault self-diagnosis method and apparatus - Google Patents

Abstract

The invention discloses a self-diagnosis method and a device of the bearing fault. A slot is opened on the bearing and a wireless vibrating micro sensor module which is integrated with fault diagnosis algorithm is embedded into the slot on the bearing. Online collection and processing of the bearing status signal is made by wireless vibrating micro sensor module and the analysis result is wirelessly output into the remote monitoring PC computer at the receiving end. The invention is characterized in that firstly, the wireless vibrating micro sensor module can be integrated with the bearing without changing the overall size of the bearing; secondly, real-time and online collection and processing of the bearing status signal can be realized; thirdly, the collected status signal is lightly polluted and more reliable; fourthly, the autonomy of the bearing fault diagnosis is improved and the dependence on people is lowered.

Description

A kind of bearing fault self-diagnosis method and device

Technical field

The present invention relates to the on-line monitoring and the diagnosis of bearing fault, embedded self-diagnosing method of a kind of bearing fault and device are provided especially.

Background technology

Bearing is the important rotary mechanical part of a class, plays a part to bear and transmitted load in plant equipment, and its running status quality will directly influence the performance of whole equipment, and tends to cause catastrophic consequence in case break down.Therefore, the state of real time on-line monitoring bearing has great importance and actual application value to improving its reliability and security.

Often acceleration transducer is measured and the analysis vibration signal is realized by installing at its surface of shell (as bearing seat) for traditional bearing state monitoring and fault diagnosis mode, the quality that the deficiency of this mode maximum is acquired signal not only with rotary part and housing between surface of contact relevant, also with sensor and housing between connection relevant.Just because of the tight contact of sensor is shell structure rather than rotary part itself, therefore the status signal of measuring is easy to be subjected to dependency structure vibration and influence of environmental noise, make that the signal to noise ratio (S/N ratio) of fault signature signal is low, increased the difficulty that fault signature extracts greatly, diagnostic result is often unsatisfactory, even also can produce mistaken diagnosis or fail to pinpoint a disease in diagnosis.On the other hand,, often be difficult to even not allow to install external sensor, cause being difficult to obtain the status signal of bearing in some practical application.

If the sensor of measuring status signal can be directly embedded in the bearing, sensor is tried one's best near the generation source of measured signal (as vibration, stress, temperature etc.), the intermediate medium of bearing state signal transmission will reduce so, the signal of gathering just more can reflect the practical working situation of bearing truly, the quality of status signal also can obtain very big improvement, the signal to noise ratio (S/N ratio) height.Simultaneously,, can gather the status signal of bearing in real time, exactly, and carry out online treatment by integrated signal microprocessing unit, thus the self diagnosis of realization fault.

Summary of the invention

The invention provides a kind of embedded self-diagnosing method and device of bearing fault, with the accuracy of raising fault diagnosis, in linearity and independence.

Bearing fault self-diagnosis method of the present invention, comprise the following steps: on bearing arrangement, to open sulculus, the wireless vibration microsensor module that is integrated with fault diagnosis algorithm is embedded in the sulculus on the bearing arrangement, by wireless vibration microsensor module online acquisition and processing bearing state signal, with the wireless receiving end that outputs to of analysis result.

Bearing fault self-diagnosis device of the present invention comprises being arranged on wireless vibration microsensor module, the remote monitoring PC in the groove on the bearing arrangement; Comprise radio receiving transmitting module and be integrated with the self diagnosis algorithm in the wireless vibration microsensor module, the remote monitoring PC regularly sends the self diagnosis startup command to wireless vibration microsensor module by radio receiving transmitting module, wireless vibration microsensor module receives the state that bearing is gathered in the order back, and utilize integrated self diagnosis algorithm to analyze and online treatment, with the wireless PC that outputs to of result.

The present invention has following characteristics: the one, and wireless vibration microsensor module can become one with bearing, does not change the appearance integral size of bearing; The 2nd, can gather and handle the status signal of bearing in real time, online; The 3rd, the status signal of gathering is contaminated little, truer; The 4th, improved the independence of bearing failure diagnosis process, reduced dependence to the people.The present invention can improve the accuracy of bearing failure diagnosis, in linearity and independence.

Further specify the present invention below in conjunction with accompanying drawing.

Description of drawings

Fig. 1 is apparatus of the present invention synoptic diagram.

Fig. 2 is the graph of a relation of fluting size and bearing load carrying capacity coefficient.

Fig. 3 is wireless vibration microsensor modular structure figure.

Fig. 4 is a wireless vibration microsensor module software process flow diagram.

Fig. 5 is the vibration signal comparison diagram of bearing different measuring points.

Fig. 6 is bearing fault self-diagnosis system interface figure.

Embodiment

Bearing fault self-diagnosis method of the present invention, comprise the following steps: on bearing arrangement, to open sulculus, the wireless vibration microsensor module that is integrated with fault diagnosis algorithm is embedded in the sulculus on the bearing arrangement, by wireless vibration microsensor module online acquisition and processing bearing state signal, with the wireless receiving end that outputs to of analysis result.

The device that adopts described method comprises being arranged on wireless vibration microsensor module 2, the remote monitoring PC 3 in the groove on the bearing arrangement outer ring 1 as shown in Figure 1; Comprise radio receiving transmitting module and be integrated with the self diagnosis algorithm in the wireless vibration microsensor module 2, remote monitoring PC 3 regularly sends the self diagnosis startup command to wireless vibration microsensor module by radio receiving transmitting module, wireless vibration microsensor module 2 receives the state that bearing is gathered in the order back, and utilize integrated self diagnosis algorithm to analyze and online treatment, with the wireless PC 3 that outputs to of result.

Be example now, the best approach that the sulculus size is determined in the inventive method is described with the deep groove ball bearing.

Described sulculus determining dimensions is that the length of sulculus and the degree of depth are then determined according to following step with the thickness of the bearing width as sulculus:

(I) at first according to a Stribeck theory and a hertz contact theory, calculate externally load F of bearing _rUnder internal load distribute, and then the maximum bearing load that obtains bearing steel ball is:

$Q=\frac{5}{Z}{F}_r---\left(1\right)$

Wherein Z is the number of ball.According to the relation of acting force and reacting force, when bearing roller contacts with the bearing outer ring ball track bottom, obtain the maximum load value Q that outer ring raceway bears by formula (1) _MaxFor:

$Q_{\max }=\frac{5}{Z}{F}_r---\left(2\right)$

(II) according to hertz Elastic Contact theory, if the deep groove ball bearing outer ring is carried in the limit of elasticity, then the surface of contact of rolling body and bearing outer ring ball track is oval, and the major semi-axis a of contact ellipse and the value of minor semi-axis b just can be tried to achieve by following formula:

$a=\mathrm{\μ}\sqrt[3]{\frac{3}{8}\×\frac{Q({\mathrm{\θ}}_1+{\mathrm{\θ}}_2)}{\mathrm{\Σ\ρ}}},b=\mathrm{\ν}\sqrt[3]{\frac{3}{8}\×\frac{Q({\mathrm{\θ}}_1+{\mathrm{\θ}}_2)}{\mathrm{\Σ\ρ}}}---\left(3\right)$

Q is a load in the formula, θ ₁And θ ₂Be respectively:

${\mathrm{\&theta;}}_1=\frac{4(1-1/{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="S2007101924810E00011.png,S2007101924810E00021.png"\; state="\{\{state\}\}">m</figure-callout>}}_1}^2)}{{\mathrm{<figure-callout\; id="1"\; label="E"\; filenames="S2007101924810E00011.png,S2007101924810E00021.png"\; state="\{\{state\}\}">E</figure-callout>}}_1},{\mathrm{\&theta;}}_2=\frac{4(1-1/{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="S2007101924810E00011.png"\; state="\{\{state\}\}">m</figure-callout>}}_2}^2)}{E_2}---\left(4\right)$

E in the formula ₁, E ₂Be the elastic modulus (units MPa) of material, 1/m ₁, 1/m ₂Be Poisson ratio, μ and v are the contact region sizes of two articles, can obtain by tabling look-up.

(III) note fluting back is α with the rate of growth of fluting preceding (under identical loading) distortion _d, then can define fluting rear bearing load-bearing capacity coefficient and be:

$\mathrm{\&eta;}=\frac{1}{1+{\mathrm{\&alpha;}}_d}---\left(5\right)$

On (I) and basis (II), utilize the bearing behind the finite element software cross-notching to carry out modeling, according to the finite element simulation result obtain slotting size and bearing load carrying capacity coefficient relation as shown in Figure 2.

(IV) under the prerequisite that guarantees the bearing proper working load, can determine optimum slotting length and depth dimensions according to Fig. 2.

Wireless vibration microsensor Module Design in apparatus of the present invention:

1) hardware components

Wireless vibration microsensor Module Design principle is to satisfy small size groove and low-power consumption requirement on the bearing arrangement.

That vibration transducer is selected for use is the simulation output MEMS 3-axis acceleration sensor LIS3L02AL of ST (STMicw Electronics) company.Radio transmitting and receiving chip is selected for use be Nordic company embedding high-performance single-chip microcomputer kernel, A/D converter and integrate the chip nRF24E1 that receives and dispatches.

The structure of wireless vibration microsensor module as shown in Figure 3.The x axle simulation output of LIS3L02AL is connected to the A/D converting analogue input channel AIN2 of nRF24E1 after dividing potential drop, y axle simulation output is connected to the A/D converting analogue input channel AIN1 of nRF24E1 after dividing potential drop, z axle simulation output is connected to A/D analog input channel AIN0 after dividing potential drop.

The microcontroller of nRF24E1, A/D converter and wireless transceiver all are operated under the same crystal oscillator, and frequency is 16MHz.The program code of nRF24E1 need load from chip external memory, has selected the AT25320 of atmel corp for use.When the SPI interface used, the SO pin of AT25320 was connected with the P1.2/DIN0 pin of nRF24E1, and the SCK pin is connected with the P1.0/DIO0 pin, and the SI pin is connected with the P1.1/D1O1 pin, Pin is connected with the P0.0/DIO2 pin.The Antenna Design of nRF24E1 adopts Nodic company reference design scheme, realizes by apply copper on circuit board.

Because wireless vibration microsensor module satisfies the requirement of low-power consumption, therefore adopt lithium battery CR2032 that it is powered.

2) software section

Wireless vibration microsensor module software mainly is to realize functions such as the obtaining of bearing state signal, processing and wireless transmission.In order to satisfy the low-power consumption requirement, adopted to send the working method that acquisition is come Remote Wake Up microsensor module, promptly the microsensor module is in standby mode always when acquired signal not.

The microsensor module software can select to carry out single shaft or three collections of bearing state signal according to the acquisition that receives.The storage space that the signal of gathering utilizes the nRF24E1 chip itself to have is stored, and adopts ShockBurst after data acquisition is finished ^TMPattern is carried out data and is transmitted, and its software flow pattern as shown in Figure 4.

The function module that software comprises mainly comprises: receive and resolve signals collecting command functions, signal A/D conversion and storage function and packet transmission function.

Integrated embedded self diagnosis algorithm in the wireless vibration microsensor module based on cyclo-stationary AR model:

Because of calculated amount is little, easy to use, be suitable for designing embedded self diagnosis algorithm based on the inference method of temporal model.Theoretical analysis shows that the bearing state signal often shows cyclo-stationary, therefore cyclo-stationary signal reason characteristic and classical temporal model method need be combined.To define cyclo-stationary AR model as follows for this reason:

$x\left(n\right)+\underset{i=1}{\overset{p}{\mathrm{\&Sigma;}}}a(n,i)x(n-i)=e\left(n\right)---\left(6\right)$

y(n)=x(n)+v(n)----------(7)

And satisfy following hypothesis:

1) e (n) is zero-mean, independent identically distributed nongausian process, and E{x (n) }=0, ${\mathrm{\&sigma;}}_x^2=E\left\{x^2\left(n\right)\right\}\&NotEqual;0,$ γ _3x＝E{x ³(n)}≠0，γ _4x＝E{x ⁴(n)}≠0；

2) a (n becomes the coefficient of AR model when i) being, a (n, 0) ≡ 1,  n, and it be in time almost the cycle change, promptly exist:

$a(n,i)=\underset{\mathrm{\&alpha;}}{\mathrm{\&Sigma;}}A(\mathrm{\&alpha;},i)e^{\mathrm{j\&alpha;n}}---\left(8\right)$

3) v (n) is a random observation noise stably, and independent with x (n).

In order to determine the order of model (6), basic ideas are the Gauss's amounts of at first constructing, and then utilize the derivation of existing information criterion.Existing document shows the poor Normal Distribution of k rank circulative accumulation amount estimated value and theoretical value, promptly remembers variable ${\mathrm{\&epsiv;}}_{\mathrm{ky}}={\hat{c}}_{k\hat{y}}^N-c_{\mathrm{ky}}^N,$ Then ${\mathrm{\&epsiv;}}_{\mathrm{ky}}\stackrel{D}{\&RightArrow;}N(0,{\mathrm{\&Sigma;}}_k/N),$ So ε _KyIt is Gauss's amount.Get the N group and postpone τ ₁..., τ _N, then the likelihood function of model (7) is

$L=\underset{i=1}{\overset{N}{\mathrm{\&Pi;}}}\left({\frac{1}{\sqrt{2\mathrm{\&pi;}}\mathrm{\&sigma;}}e}^{-\frac{{({\mathrm{\&epsiv;}}_{\mathrm{ky}}\left({\mathrm{\&tau;}}_i\right)-0)}^2}{{2\mathrm{\&sigma;}}^2}}\right)---\left(9\right)$

In the formula ${\mathrm{\&epsiv;}}_{\mathrm{ky}}\left({\mathrm{\&tau;}}_i\right)={\hat{c}}_{k\hat{y}}^N\left({\mathrm{\&tau;}}_i\right)-c_{\mathrm{ky}}^N\left({\mathrm{\&tau;}}_i\right),{\mathrm{\&sigma;}}^2={\mathrm{\&Sigma;}}_k/N.$ Formula (9) is taken from right logarithm to be obtained

$\ln L=\ln \left[{\left(\frac{1}{{2\mathrm{\&pi;\&sigma;}}^2}\right)}^{\frac{N}{2}}{e}^{-\frac{1}{{2\mathrm{\&sigma;}}^2}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{\mathrm{\&epsiv;}}_{\mathrm{ky}}^2\left({\mathrm{\&tau;}}_i\right)}\right]=\frac{\mathrm{<figure-callout\; id="2"\; label="N"\; filenames="S2007101924810E00011.png"\; state="\{\{state\}\}">N</figure-callout>}}{2}\mathrm{<figure-callout\; id="1"\; label="ln"\; filenames="S2007101924810E00011.png,S2007101924810E00021.png"\; state="\{\{state\}\}">ln</figure-callout>}\frac{1}{2\mathrm{\&pi;}{\mathrm{\&sigma;}}^2}-\frac{1}{2{\mathrm{\&sigma;}}^2}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{\mathrm{\&epsiv;}}_{\mathrm{ky}}^2\left({\mathrm{\&tau;}}_i\right)---\left(10\right)$

$=-\frac{\mathrm{<figure-callout\; id="2"\; label="N"\; filenames="S2007101924810E00011.png"\; state="\{\{state\}\}">N</figure-callout>}}{2}\mathrm{<figure-callout\; id="2"\; label="ln"\; filenames="S2007101924810E00011.png"\; state="\{\{state\}\}">ln</figure-callout>}2\mathrm{\&pi;}-\frac{\mathrm{<figure-callout\; id="2"\; label="N"\; filenames="S2007101924810E00011.png"\; state="\{\{state\}\}">N</figure-callout>}}{2}\ln {\mathrm{\&sigma;}}^2-\frac{1}{{2\mathrm{\&sigma;}}^2}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{\mathrm{\&epsiv;}}_{\mathrm{ky}}^2\left({\mathrm{\&tau;}}_i\right)$

$=-\frac{\mathrm{<figure-callout\; id="2"\; label="N"\; filenames="S2007101924810E00011.png"\; state="\{\{state\}\}">N</figure-callout>}}{2}\mathrm{<figure-callout\; id="2"\; label="ln"\; filenames="S2007101924810E00011.png"\; state="\{\{state\}\}">ln</figure-callout>}2\mathrm{\&pi;}-\frac{\mathrm{<figure-callout\; id="2"\; label="N"\; filenames="S2007101924810E00011.png"\; state="\{\{state\}\}">N</figure-callout>}}{2}\ln \frac{{\mathrm{\&Sigma;}}_k}{N}-\frac{\mathrm{<figure-callout\; id="2"\; label="N"\; filenames="S2007101924810E00011.png"\; state="\{\{state\}\}">N</figure-callout>}}{2{\mathrm{\&Sigma;}}_k}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{\mathrm{\&epsiv;}}_{\mathrm{ky}}^2\left({\mathrm{\&tau;}}_i\right)$

With the limited sample information criterion of formula (10) substitution, obtain:

Wherein $v\left(i\right)=\frac{N-i}{N(N+2)},$ Can define the limited sample information criterion of determining model order thus:

$\mathrm{CFSIC}\left(p\right)=-{\mathrm{\&epsiv;}}_{\mathrm{ky}}\left({\mathrm{\&tau;}}_i\right){\mathrm{\&Sigma;}}_k^{-1}{\mathrm{\&epsiv;}}_{\mathrm{ky}}^{\&prime;}\left({\mathrm{\&tau;}}_i\right)+\underset{i=1}{\overset{\mathrm{<figure-callout\; id="1"\; label="p"\; filenames="S2007101924810E00011.png,S2007101924810E00021.png"\; state="\{\{state\}\}">p</figure-callout>}}{\mathrm{\&Pi;}}}\frac{1+v\left(i\right)}{1-v\left(i\right)}-1---\left(12\right)$

The order of model is the p of CFSIC (p) when getting minimum value.

Then, character and the formula (6) according to semi-invariant can obtain:

$\underset{i=0}{\overset{p}{\mathrm{\&Sigma;}}}a(t,i)\mathrm{cum}\{x(t-i),x(t-{\mathrm{\&tau;}}_1),\&CenterDot;\&CenterDot;\&CenterDot;,x(t-{\mathrm{\&tau;}}_{k-1})\}---\left(13\right)$

$=\mathrm{cum}\{\underset{i=0}{\overset{p}{\mathrm{\&Sigma;}}}a(t,i)x(t-i),x(t-{\mathrm{\&tau;}}_1),\&CenterDot;\&CenterDot;\&CenterDot;,x(t-{\mathrm{\&tau;}}_{k-1})\}$

$=\mathrm{cum}\{e\left(t\right),x(t-{\mathrm{\&tau;}}_1),\&CenterDot;\&CenterDot;\&CenterDot;,x(t-{\mathrm{\&tau;}}_{k-1})\}$

By formula (6) as can be known, e (t) only with x (t), x (t+1) ..., x (t+n) is relevant, and with x (t-1) ..., x (t-n) is irrelevant, if the therefore τ of formula (13) ₁..., τ _K-1In have one at least greater than zero, then obtain being similar to the cyclo-stationary high-order Yule-Walker equation under the stationary signal:

$\underset{i=0}{\overset{p}{\mathrm{\&Sigma;}}}a(t,i)c_{\mathrm{kx}}(t-i;i-{\mathrm{\&tau;}}_1,...,i-{\mathrm{\&tau;}}_{k-1})=0---\left(14\right)$

Especially, in formula (14), make τ ₁=...=τ _K-1=τ gets the section of main diagonal angle, obtains:

$\underset{i=0}{\overset{p}{\mathrm{\&Sigma;}}}a(t,i)c_{\mathrm{kx}}(t-i;i-\mathrm{\&tau;},...,i-\mathrm{\&tau;})=0,\&ForAll;\mathrm{\&tau;}>0---\left(15\right)$

Again because v (t) and x (t) are independent, therefore by the time become semi-invariant character as can be known, when k 〉=3:

c _ky(t，τ ₁，...，τ _k-1)＝c _kx(t，τ ₁，...，τ _k-1)+c _kv(t，τ ₁，...，τ _k-1) (16)

＝c _kx(t，τ ₁，...，τ _k-1)

Formula (16) substitution formula (15) can be obtained:

$\underset{i=0}{\overset{p}{\mathrm{\&Sigma;}}}a(t,i)c_{\mathrm{ky}}(t-i;i-\mathrm{\&tau;},...,i-\mathrm{\&tau;})=0,\&ForAll;\mathrm{\&tau;}>0---\left(17\right)$

To fixing t, select τ=1 ..., p can obtain following system of equations:

$\left[\begin{array}{ccc}a(t,1)& \&CenterDot;\&CenterDot;\&CenterDot;& a(t,p)\end{array}\right]\left[\begin{array}{ccc}{c}_{\mathrm{ky}}(t-\mathrm{1,1}-1,...,1-1)& \&CenterDot;\&CenterDot;\&CenterDot;& c_{\mathrm{ky}}(t-\mathrm{1,1}-p,...,1-p)\\ \&CenterDot;& \&CenterDot;& \&CenterDot;\\ \&CenterDot;& \&CenterDot;& \&CenterDot;\\ \&CenterDot;& \&CenterDot;& \&CenterDot;\\ c_{\mathrm{ky}}(t-p,p-1,...,p-1)& \&CenterDot;\&CenterDot;\&CenterDot;& c_{\mathrm{ky}}(t-p,p-p,...,p-p)\end{array}\right]---\left(18\right)$

$=-\left[\begin{array}{ccc}{c}_{\mathrm{ky}}(t,-1,...,-1)& \&CenterDot;\&CenterDot;\&CenterDot;& c_{\mathrm{ky}}(t,-p,...,-p)\end{array}\right]$

Note as vector form is

a _t·C _t＝c _t (19)

C in the formula _tBe the square formation of p * p, if C _tNonsingularly (can guarantee C by the value of adjusting τ _tNonsingular), a then _tUnique solution is arranged

$a_t=c_t\&CenterDot;C_t^{-1}---\left(20\right)$

Thereby can find the solution cyclo-stationary AR model coefficient a (t, i).

At first utilize the bearing normal signal off-line of collection to estimate the order of model during practical application

(t i), and is placed in the storer of microsensor inside modules with parameter a.The prediction of online computation model output then:

$\hat{y}\left(t\right)=-\underset{i=0}{\overset{\hat{p}}{\mathrm{\&Sigma;}}}\hat{a}(t,i)\hat{y}(t-i)---\left(21\right)$

Obtain the residual sequence of real data and predicted data:

$e\left(t\right)=y\left(t\right)-\hat{y}\left(t\right)---\left(22\right)$

From e (n), can extract the characteristic parameter that characterizes bearing state, fault is carried out on-line monitoring and diagnosis, and compare with the level threshold value of prior setting and to determine whether reporting to the police, thus realize fault online from detect, self diagnosis.

Case study on implementation:

With deep groove ball bearing 6220 is example, and slotting position is chosen in bearing outer ring.Getting external loading is F _r=50kN, this moment, bearing was worked in elastic range.Calculate the maximum load Q that outer ring raceway bears according to formula (2) _Max=25kN; Calculate the elliptical region size a=5.525mm that outer ring raceway carries, b=0.7482mm according to formula (3); At the load-bearing capacity coefficient is 0.8 o'clock, and the width that obtains sulculus is that 6mm, length are that 34mm (equaling the thickness of bearing), the degree of depth are 3mm.The wireless vibration microsensor that will be integrated with the self diagnosis algorithm then embeds in the sulculus, adopts remote monitoring PC monitoring wireless vibration microsensor that the bearing running status is diagnosed.

Effect of the present invention can be illustrated by following contrast experiment.

Simulate the fatigue flake fault of bearing inner race raceway at width of the inner ring raceway line of bearing processing for the 0.5mm crackle, and lay the microsensor module three different measuring points: measuring point one is on the bearing seat, measuring point two be embedded in the sulculus of outer ring and the sulculus position directly over, measuring point three be embedded in the sulculus of outer ring and the sulculus position under.The z shaft vibration signal that rotating speed is gathered three measuring points during for 300r/min as shown in Figure 5.As can be seen from Figure 5, the vibration signal at measuring point two and measuring point three places has the obvious periodic pulse than measuring point one, is indicating the existence of fault.Therefore, the fault-signal signal to noise ratio (S/N ratio) height that obtains of the microsensor module of embedding.

Fig. 6 is the observation interface of reality when measuring point three is laid the microsensor module, and the kurtosis coefficient that goes out according to the self diagnosis algorithm computation has produced fault alarm greater than threshold value as can be seen.

In sum, utilize apparatus of the present invention, the user only needs startup command, embedded microsensor module is the state of online acquisition bearing regularly, and utilize integrated self diagnosis algorithm to handle, and analysis result and warning message be shown to the user, On-line Fault monitoring and diagnostic procedure do not need the user to participate in.

Claims (3)

1. bearing fault self-diagnosis method, it is characterized in that on bearing arrangement, opening sulculus, the wireless vibration microsensor module that is integrated with fault diagnosis algorithm is embedded in the sulculus on the bearing arrangement, by wireless vibration microsensor module online acquisition and processing bearing state signal, with the wireless receiving end that outputs to of analysis result.

2. bearing fault self-diagnosis method according to claim 1 is characterized in that described sulculus determining dimensions is that the length of sulculus and the degree of depth are determined according to following step with the thickness of the bearing width as sulculus:

1) at first according to a Stribeck theory and a hertz contact theory, can obtain the maximum load value Q that outer ring raceway bears _MaxFor:

$Q_{\max }=\frac{5}{Z}{F}_r$

F in the formula _rBe the bearing external loading, Z is the number of bearing ball;

2), calculate the major semi-axis a of contact ellipse of rolling body and bearing outer ring ball track and the value of minor semi-axis b by following formula according to hertz Elastic Contact theory:

$a=\mathrm{\&mu;}\sqrt[3]{\frac{3}{8}\&times;\frac{Q({\mathrm{\&theta;}}_1+{\mathrm{\&theta;}}_2)}{\mathrm{\&Sigma;\&rho;}}},b=\mathrm{\&nu;}\sqrt[3]{\frac{3}{8}\&times;\frac{Q({\mathrm{\&theta;}}_1+{\mathrm{\&theta;}}_2)}{\mathrm{\&Sigma;\&rho;}}}$

Q is a load in the formula, θ ₁And θ ₂Be respectively:

${\mathrm{\&theta;}}_1=\frac{4(1-1/{m_1}^2)}{E_1},{\mathrm{\&theta;}}_2=\frac{4(1-1/{m_2}^2)}{E_2}$

E in the formula ₁, E ₂Be the elastic modulus of material, 1/m ₁, 1/m ₂Be Poisson ratio, μ and v are the contact region sizes of two articles;

3) definition fluting rear bearing load-bearing capacity coefficient is:

$\mathrm{\&eta;}=\frac{1}{1+{\mathrm{\&alpha;}}_d}$

α in the formula _dBe back and the preceding distortion of the fluting rate of growth of slotting;

4) utilize the finite element simulation result to draw the graph of a relation of fluting size and bearing load carrying capacity coefficient, by the length and the degree of depth of figure definite sulculus under selected bearing load carrying capacity.

3. bearing fault self-diagnosis device is characterized in that comprising being embedded in wireless vibration microsensor module, the remote monitoring PC in the sulculus on the bearing arrangement; Comprise radio receiving transmitting module and be integrated with the self diagnosis algorithm in the wireless vibration microsensor module, the remote monitoring PC regularly sends the self diagnosis startup command to wireless vibration microsensor module by radio receiving transmitting module, wireless vibration microsensor module receives the state that bearing is gathered in the order back, and utilize integrated self diagnosis algorithm to analyze and online treatment, with the wireless PC that outputs to of result.

Images