CN100543440C

CN100543440C - Failure testing method of asynchronous motor bearing

Info

Publication number: CN100543440C
Application number: CNB2007100616333A
Authority: CN
Inventors: 许伯强; 孙丽玲
Original assignee: North China Electric Power University
Current assignee: North China Electric Power University
Priority date: 2007-03-28
Filing date: 2007-03-28
Publication date: 2009-09-23
Anticipated expiration: 2027-03-28
Also published as: CN101034038A

Abstract

A kind of failure testing method of asynchronous motor bearing belongs to the detection technique field, is used to solve the problem of online detection asynchronous motor bearing incipient failure.Its technical scheme is: it is by the stator current momentary signal i to gathering _sDo continuous refinement Fourier transform, obtaining its first-harmonic is reference signal u _S, again according to reference signal u _SAnd frequency f ₁To stator current momentary signal i _sDo auto adapted filtering, then to the filtering output signal e _TDo continuous refinement Fourier transform, determine current | f ₁± mf _v| side frequency component and the ratio of fundametal compoment amplitude and it as fault signature, determine fault index according to detection threshold at last, judge whether to exist bearing fault according to fault index.The present invention can high sensitivity, the first property the sent out bearing fault of the various asynchronous motors of the high reliability online detection in ground.

Description

Asynchronous motor bearing fault detection method

Technical Field

The invention relates to a method and a device capable of detecting initial faults of an asynchronous motor bearing on line, belonging to the technical field of detection.

Background

Rolling bearings are widely used in asynchronous motors with overwhelming advantages. The rolling bearing is composed of an inner raceway, an outer raceway, and rolling bodies rotating therebetween. Under normal operating conditions of balanced load, good centering, fatigue failure starts from tiny cracks located under the raceway and rolling element surfaces and gradually propagates, causing material fragments to fall off, resulting in bearing failure with a probability of about 40% of that of asynchronous motor failure.

At present, vibration signal frequency spectrum analysis is a more accurate and reliable asynchronous motor bearing fault detection method. The method collects a bearing time domain vibration signal and transforms the bearing time domain vibration signal to a frequency domain, and then compares the frequency domain vibration signal with the inherent frequency domain vibration characteristic of the bearing to judge whether the bearing fails or not. The method has the disadvantages that a vibration sensor needs to be installed, and the vibration sensor is high in manufacturing cost and easy to damage, so that the further popularization and application of the method are limited.

A method for analyzing the frequency spectrum of the stator current signal has been proposed. Different from each otherAfter a fault in the stepper motor bearing occurs, | f will appear in the stator current₁±mf_vAdditional current component (f) of the | frequency₁For supply frequency, f_vIs the bearing vibration characteristic (natural) frequency, m is 1, 2, 3 …). The bearing failure is generally classified into an outer raceway failure, an inner raceway failure, a rolling element failure, and a cage failure, and the vibration characteristic (natural) frequency is calculated according to equations (1), (2), (3), and (4), respectively.

Outer raceway failure natural frequency:

inner raceway fault natural frequency:

rolling element failure natural frequency:

cage failure natural frequency:

wherein f is_rmThe rotor frequency of the motor is n, the number of the rolling elements is n, BD and PD are the diameters of the rolling elements and the pitch diameters of the bearings, and phi is the contact angle of the rolling elements.

From the natural frequency, the stator current characteristic frequency can be calculated according to equation (5).

f_CF＝|f₁±mf_v| m＝1，2，3 …(5)

Wherein f is₁For supply frequency, f_vIs the natural frequency of vibration shown by the formula (1-4).

In view of the fact that stator current signals are easy to collect, the method has wide development and application prospects compared with vibration signal frequency spectrum analysis. However, due to bearing failure characteristics-stator currents|f₁±mf_vMagnitude of | frequency component relative to f₁The components are very small and the method does not guarantee the sensitivity of extracting the fault features. In addition, due to the inherent asymmetry, air gap eccentricity, rotor misalignment and other factors of the motor itself, the stator current of an asynchronous motor may contain | f even in a normal operation state₁±mf_vThe | frequency component, and for different asynchronous motors the situation is complex. This is easily confused with bearing fault characteristics, resulting in erroneous judgment and affecting fault detection reliability. This method cannot account for the above factors, and reliability still remains to be improved.

Disclosure of Invention

The invention aims to provide a fault detection method and a fault detection device for an asynchronous motor bearing, which can detect the initial fault of the asynchronous motor bearing on line with high sensitivity and high reliability.

The problem is realized by the following technical scheme:

a method for detecting the failure of asynchronous motor bearing features that the instantaneous signal i of collected stator current is used_sPerforming continuous thinning Fourier transform to obtain the fundamental wave thereof, namely a reference signal u_SBased on the reference signal u_SAnd frequency f thereof₁For stator current instantaneous signal i_sAdaptive filtering and then outputting the filtered output signal e_TPerforming continuous thinning Fourier transform to determine the current | f₁±mf_vAnd finally, determining a fault index according to a detection threshold value, and judging the bearing fault according to the fault index.

According to the asynchronous motor bearing fault detection method, the fault index is calculated according to the following steps:

a. collecting a phase stator current transient signal, i_s；

b. To i_sPerforming continuous fine Fourier transform to determineFrequency f of fundamental component thereof₁Amplitude I_mAnd an initial phase angle phi to form a reference signal u_S：

For a sampling frequency of f_sTime series i (t) with the number of sampling points N_k)，

u_S(K)＝I_m sin(2πf₁kT_S+φ+π)

Wherein k is 0, 1, 2, …, N-1, I_m、f₁Phi is determined by a continuous refined fourier transform.

c. According to frequency f of fundamental component₁Reference signal u_STo i_sAdaptive filtering to cancel f₁Component, filter output result is noted as e_T；

d. To e_TPerforming continuous fine Fourier transform at e_TIn the continuously refined spectrogram_CF(f_CF＝|f₁±mf_vTaking m as 1, 2) component information to determine current f_CFComponent and f₁Ratio of component amplitudes

{ratio}_{f_{CF}} = {ratio}_{f_{1} + f_{v}} + {ratio}_{| f_{1} - f_{v} |} + {ratio}_{f_{1} + 2 f_{v}} + {ratio}_{| f_{1} - 2 f_{v} |},

Wherein,

is f₁+f_vComponent and f₁The ratio of the magnitudes of the components,

is f₁-f_vComponent and f₁The ratio of the magnitudes of the components,

is f₁+2f_vComponent and f₁The ratio of the magnitudes of the components,is f₁-2f_vComponent and f₁The ratio of component amplitudes;

e. determining a fault index:

in the case where a normal motor sample reference has not been established, the detection threshold (typically 0.1%) is set according to conventional experience,

the ratio is the fault index;

f. judging whether the fault exists or not according to the fault index:

the fault index value is less than 1, which indicates that the motor is in a healthy state, and the smaller the value is, the more definite the healthy state is; the fault index value is greater than 1, which indicates that the motor is in a fault state, and the larger the value is, the more serious the fault state is.

The asynchronous motor bearing fault detection method is used for filtering the output signal e_TIn continuously refining the spectrogramQuery f_CFThe side frequency component information should calculate the difference rate s first:

<math> <mrow> <mi>s</mi> <mo>=</mo> <mn>1</mn> <mo>-</mo> <mfrac> <mi>P</mi> <msub> <mi>Z</mi> <mi>r</mi> </msub> </mfrac> <mrow> <mo>(</mo> <mfrac> <msub> <mi>f</mi> <mi>rsh</mi> </msub> <msub> <mi>f</mi> <mn>1</mn> </msub> </mfrac> <mo>&PlusMinus;</mo> <mi>v</mi> <mo>)</mo> </mrow> <mo>,</mo> <mi>v</mi> <mo>=</mo> <mn>1,3,5</mn> <mo>,</mo> <mo>·</mo> <mo>·</mo> <mo>·</mo> </mrow></math>

wherein f is_rshIs the harmonic component frequency of the rotor tooth slot, P is the number of pole pairs of the motor, Z_rThe number of the rotor grooves is the number of the rotor grooves,

then according to the fundamental component frequency f₁Slip s, on the filtered output signal e_TIn the continuously refined spectrogram_CFSide frequency component information, determining current f_CFThe side frequency component and f₁Ratio of component amplitudes

(the motor frequency f in the natural frequency expression)_rm＝(1-s)f₁)。

In order to eliminate the influence of asymmetry, air gap eccentricity, rotor misalignment and other factors inherent in the actual motor, the method for detecting the bearing fault of the asynchronous motor needs to be carried out according to the slip ratio s and the fault characteristics under the normal condition of the motor bearing

Establishing a sample database according to the specific numerical value, and setting a detection threshold value according to the sample database:

if the current value of the slip is between the upper limit and the lower limit of the slip of the sample data, setting a detection threshold value by adopting a linear interpolation mode; otherwise, determining the sample data slip rate closest to the sample data slip rate, taking the corresponding fault characteristic value as a detection threshold value, and enabling the reliability coefficient to be not less than 1.

The invention utilizes the current transformer CT to collect the current signal of the stator winding of the asynchronous motor, the signal acquisition card transmits the signal to the computer, and the computer processes the current signal to judge whether the bearing fault exists or not, and the invention has simple structure and convenient operation. The invention employs a stator current f_CFThe side frequency component is used as a fault feature, the technologies of continuous refined Fourier transform, adaptive filtering, rotor tooth space harmonic slip rate estimation and detection threshold self-setting are organically combined, the sensitivity is improved, meanwhile, the influences of inherent asymmetry, air gap eccentricity, rotor misalignment and other factors of an actual motor on the extraction of the fault feature are eliminated, and misjudgment is effectively prevented. The invention can detect the initial bearing fault of various asynchronous motors on line with high sensitivity and high reliability.

Drawings

The present invention will be described in further detail with reference to the accompanying drawings.

FIG. 1 is an electrical schematic block diagram of the present invention;

FIG. 2 is a functional block diagram of an adaptive filtering method;

FIG. 3 is a schematic diagram of a signal acquisition circuit;

FIG. 4 is the stator a-phase current for the case of a Y100L-2 model experimental motor fully loaded and bearing with outer race fault;

FIG. 5 is f phase current frequency spectrum of stator a of Y100L1-2 type experimental motor₁+f_vFrequency bands;

FIG. 6 is f phase current frequency spectrum of stator a of Y100L1-2 type experimental motor₁-f_vFrequency bands;

FIG. 7 is f phase current frequency spectrum of stator a of Y100L1-2 type experimental motor₁+2f_vFrequency bands;

FIG. 8 is f phase current frequency spectrum of stator a of Y100L1-2 type experimental motor₁-2f_vFrequency bands.

The reference numbers in the figures are: CT, a current transformer, PT, a voltage transformer, M and a motor; r1, R2 and a resistor.

The meaning of the symbols used herein: s, slip; f. of_rmThe frequency conversion of the motor; n, the number of rolling bodies; BD. The diameter of the rolling body; PD, bearing pitch diameter; phi, contact angle of the rolling body; f. of_vNatural frequency of vibration; f. of₁Stator current fundamental frequency (supply frequency); f. of_ODOuter raceway failure natural frequency; f. of_IDInner raceway fault natural frequency; f. of_BDRolling element failure natural frequency; f. of_CDThe natural frequency of the fault of the retainer; f. of_CFStator current characteristic frequency;

f_CFcomponent and f₁The ratio of component amplitudes;

f₁+f_vcomponent and f₁The ratio of component amplitudes;

f₁-f_vcomponent and f₁The ratio of component amplitudes;

f₁+2f_vcomponent and f₁The ratio of component amplitudes;

f₁-2f_vcomponent and f₁The ratio of component amplitudes; i.e. i_sA stator current signal; s_TSelf-adaptively filtering the signal to be extracted; n is_TA noise signal; u. of_SReference signal (adaptive filtering); e.g. of the type_TFiltering the output signal; y is_TAnd a filter response; i is_mA reference signal amplitude; phi, reference signal initial phase angle; f. of_sSampling frequency; n, sampling point number; i (t)_k) Current sampling instantaneous values; t is t_kSampling time; f. of_rshThe harmonic component frequency of the rotor tooth slot, P and the number of pole pairs of the motor; z_rThe number of rotor slots; t is_sSampling time interval; a (n), b (n), a (0) represent Fourier coefficients; Δ f, frequency resolution unit.

Detailed Description

The invention adopts the circuit shown in figure 1 for detection, the circuit consists of a current transformer CT, a signal acquisition card and a computer, the current transformer is connected on a phase line of a stator winding of an asynchronous motor, the signal output end of the current transformer is connected with an analog signal input channel 5 (input terminals 5 and 17) of the signal acquisition card, and the output port of the signal acquisition card is connected with a USB port of the computer. The signal acquisition card adopts a Ribohua RBH8321 type signal acquisition card, the model of the computer is DELL M1210, and the signal acquisition card integrates circuits such as a low-pass filter, a signal acquisition hold circuit, an analog/digital conversion circuit and the like. The instantaneous signal of the stator current is sent to a signal acquisition card which is connected with a portable computer through a USB interface. The portable computer control signal acquisition card samples the stator current instantaneous signal with proper frequency and stores the signal in the hard disk, and the computer processes the current signal to judge whether the bearing fault exists. The supporting software of the device is based on a Windows XP operating system and is compiled by adopting a Visual C + + application program development platform.

The basic idea of the adaptive filtering method is as follows: counteracting motor stator current f by adopting self-adaptive filtering method₁Component, highlighted in the spectrogram_CFSide frequency component-bearing fault characteristic, thereby greatly improving the sensitivity of bearing fault detection. Reference signal u_SApplying a continuous refined fourier transform determination.

The principle of adaptive filtering is as follows:

the following diagram is a functional block diagram of an adaptive filtering method. In the figure, i_sRepresenting the actual stator current signal, which contains the signal S to be extracted_TAnd noise n_TAnd u is_SIs a reference signal. Here, S_TI.e. f in the stator current₁±mf_v| frequency component, n_TFor f in stator current₁Frequency component, e_TThen represents the pair i_sThe resulting signal after being subjected to an adaptive filtering process. Setting the response of the adaptive filter to y_TObviously, e_T＝i_s-y_T. According to e_TIs adjusted by an adaptive algorithm, and y is changed appropriately_TCan make y_TCancelling n in the sense of minimum mean square error_TAnd e is_TWill approximate the signal S to be extracted in the sense of minimum mean square error_T。

FIG. 2 is a schematic block diagram of an adaptive filtering method, i in FIG. 2_sRepresenting the actual stator current signal, which contains the signal S to be extracted_TAnd noise n_TAnd u is_SIs a reference signal. Here, S_TI.e. f in the stator current₁±mf_v| frequency component, n_TFor f in stator current₁Frequency component, e_TThen represents the pair i_sThe resulting signal after being subjected to an adaptive filtering process. Setting the response of the adaptive filter to y_TObviously, e_T＝i_s-y_T. According to e_TIs adjusted by an adaptive algorithm, and y is changed appropriately_TCan make y_TCancelling n in the sense of minimum mean square error_TAnd e is_TWill approximate the signal S to be extracted in the sense of minimum mean square error_T。

Fig. 3 is a current signal acquisition circuit. Obviously, the resistance R₁The voltage signal at is i in FIG. 2_s。

The motor can be extracted with high sensitivity by applying the continuous refined Fourier transform and the self-adaptive filtering technologyStator current f_CFThe side frequency component can be used for solving an accurate analytical expression of a certain main frequency component in the signal to be analyzed by applying a continuous thinning Fourier transform method, namely the frequency, the amplitude and the initial phase angle.

For a sampling frequency of f_sTime series i (t) with the number of sampling points N_k) The discrete fourier series is:

wherein, t_k＝kT_s，T_s＝1/f_sAnd k is 0, 1, 2, …, N-1, a (N), b (N), a (0) represents fourier coefficients.

The fast fourier transform is a special case of the discrete transform described above, i.e. N-2^m(m is a positive integer), in this case, the fourier transform may employ a recursive fast algorithm. In the conversion, the frequency resolution unit is delta f ═ f_sand/N, inversely proportional to the number of sampling points N. Obviously, if the frequency resolution is to be improved, the number of sampling points must be increased by several times, and if the number of sampling points is constant, the frequency resolution cannot be further improved.

Time series i (t)_k) Including signals from 0 to f_sThe information in the frequency domain,/2, if the spectral curve is regarded as continuous, N in equation (6) is considered to belong to the interval [0, N/2 ]]The expression (6) can be rewritten as the expression (7) by successive real numbers of (c). At this time, the frequency resolution is no longer limited by the number of sampling points, and the value of the frequency f is continuous.

When the continuous thinning Fourier transform is applied, the thinning range and the thinning density can be carried out step by step so as to improve the calculation speed.

Reference signal u_SDetermined according to equation (8).

u_S(K)＝I_m sin(2πf₁kT_S+φ+π) (8)

After a bearing fault occurs in an actual motor, the stator current frequency spectrum of the actual motor often comprises a plurality of spectral peaks, and f_CFThe peak value of the frequency component spectrum peak is not necessarily the largest. This can be attributed to asymmetry, air gap eccentricity, rotor misalignment and other factors inherent in the actual motor itself.

Characteristic f of extreme easiness of side frequency component and bearing fault_CFThe frequency components are mixed up to cause misjudgment, and the reliability of bearing fault detection is influenced. Therefore, when detecting a bearing failure, f should be predicted₁And s, thereby to purposefully query f in the stator current spectrum_CFAnd the frequency component is quantitatively analyzed to ensure the reliability of bearing fault detection.

During the operation of the motor, the stator current contains the rotor cogging harmonic component due to the interaction of the rotor magnetomotive force cogging harmonic and the fundamental wave air gap flux. According to its frequency f_rshNumber of pole pairs P and number of rotor slots Z of motor_rThe motor slip can be determined as in equation (9):

<math> <mrow> <mi>s</mi> <mo>=</mo> <mn>1</mn> <mo>-</mo> <mfrac> <mi>P</mi> <msub> <mi>Z</mi> <mi>r</mi> </msub> </mfrac> <mrow> <mo>(</mo> <mfrac> <msub> <mi>f</mi> <mi>rsh</mi> </msub> <msub> <mi>f</mi> <mn>1</mn> </msub> </mfrac> <mo>&PlusMinus;</mo> <mi>v</mi> <mo>)</mo> </mrow> <mo>,</mo> <mi>v</mi> <mo>=</mo> <mn>1,3,5</mn> <mo>,</mo> <mo>·</mo> <mo>·</mo> <mo>·</mo> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>9</mn> <mo>)</mo> </mrow> </mrow></math>

in engineering practice, f is often chosen_CFFrequency component and f₁The ratio of the component amplitudes is taken as a fault characteristic, and bearing fault detection is realized by judging whether the value exceeds a certain threshold value.

For reasons of process, manufacture and installation, any actual motor is necessarily inherent with certain degree of asymmetry, air gap eccentricity, rotor misalignment and other factors, and even if the motor is in a normal operation state, the stator current of the motor can also contain f_CFAnd other frequency components. And the situation is complicated for different asynchronous motors.

This raises the problem of how to properly set the detection threshold to simultaneously achieve sensitivity and reliability. If the detection threshold is set too high, it is not beneficial to improve the sensitivity. On the other hand, if the detection threshold is set too low to be lower than the stator current f during normal operation of the motor_CFComponent and f₁The ratio of component amplitudes inevitably causes misjudgment of faults, and the reliability is not mentioned.

The above analysis shows that, for bearing fault detection with high sensitivity and high reliability, it is necessary to first specify the normal motor f_CFFrequency component and f₁The ratio of component amplitudes, according to which an appropriate detection threshold is set to avoidAnd the failure missing detection and the misjudgment are avoided.

In order to account for the inherent asymmetry, air gap eccentricity, rotor misalignment and other factors of the actual motor, a detection threshold self-tuning strategy based on sample learning is adopted to improve the sensitivity and reliability.

Assuming the motor bearing is normal initially, the slip ratio s and the fault characteristics are used

Because of fault characteristics, establish a sample database

Substantially only on the slip s. A sample database is established, the simplicity is emphasized, and the normal fluctuation of the slip ratio is covered as much as possible by combining the actual operation mode of the motor.

Once the normal motor sample database is established, the detection threshold value can be self-set according to the current slip rate value, which is specifically as follows: if the current value of the slip is between the upper limit and the lower limit of the slip of the sample data, setting a detection threshold value by adopting a linear interpolation mode; otherwise, determining the sample data slip rate closest to the sample data slip rate, taking the corresponding fault characteristic value as a detection threshold value, and enabling the reliability coefficient to be not less than 1.

After the asynchronous motor has a bearing fault, the stator current f_CF(f_CF＝|f₁±mf_vL, taking m-1, 2) the frequency components appear matched. F should be queried simultaneously for bearing fault detection sensitivity and reliability considerations_CFFrequency component information. Method for detecting initial bearing fault of asynchronous motor by simultaneously adopting stator current f_CF(including f)₁+f_v、|f₁-f_v|、f₁+2f_v、|f₁-2f_v|)) frequency components as fault signatures.

The bearing fault detection experiment is carried out on the Y100L-2 type experimental motor by applying the invention, the motor bearing model is 6206-2 RS, the outer diameter is 62mm, the inner diameter is 30mm, the thicknesses of the inner and outer raceways are the same, the bearing pitch diameter PD is 46mm, the number n of the bearing balls is 9, the diameter (BD) of the bearing balls is about 10mm, and the contact angle beta is 0 deg. The bearing has outer raceway faults (the surface of the outer raceway is provided with a small hole with the diameter of 1.5mm and the depth of 10 mm).

Fig. 4 to 8 show the experimental results in the case of full motor load, and the specific data are shown in table 1.

TABLE 1 Experimental results of the motor in full load and bearing with outer raceway fault

Bearing failure experimental results

And (3) performing experiments on the bearing, wherein the load conditions of the motor are respectively set to be no-load, half-load and full-load. And collecting three-phase current signals and recording the rotating speed of the motor. Respectively calculating the vibration characteristic frequency f of the bearing fault according to the formula (1) and the formula (2)_vDetermination of the supply frequency f by means of a continuously refined Fourier transform₁Then calculating the bearing fault characteristic frequency f in the stator current according to the formula (5)_CFThe calculation results are shown in Table 2.

TABLE 2 bearing failure characteristic frequency

Claims

1. The method for detecting the bearing fault of asynchronous motor is characterized by collecting the instantaneous signal i of stator current_sPerforming continuous thinning Fourier transform to obtain the fundamental wave thereof, namely a reference signal u_SBased on the reference signal u_SAnd frequency f thereof₁For stator current instantaneous signal i_sAdaptive filtering and then outputting the filtered output signal e_TPerforming continuous thinning Fourier transform to determine the current | f₁±mf_vThe ratio of the amplitude of the I side frequency component to the amplitude of the fundamental wave component is taken as a fault characteristic, and finally, the fault characteristic is determined according to a detection threshold valueJudging the fault index according to the fault index;

the fault index is calculated according to the following steps:

a. collecting a phase stator current transient signal, i_s；

b. To i_sPerforming continuous fine Fourier transform to determine the frequency f of its fundamental component₁Amplitude I_mAnd an initial phase angle phi to form a reference signal u_S：

u_S(K)＝I_m sin(2πf₁kT_S+φ+π)

Wherein k is 0, 1, 2, …, N-1, I_m、f₁Phi is determined by continuous thinning Fourier transform; t is_sIs a sampling time interval;

d. To e_TPerforming continuous fine Fourier transform at e_TIn the continuously refined spectrogram_CFComponent information, f_CFIs the characteristic frequency of stator current, f_CF＝|f₁±mf_vL, taking m as 1, 2, determining current f_CFComponent and f₁Ratio of component amplitudes

{ratio}_{f_{CF}} = {ratio}_{f_{1} + f_{v}} + {ratio}_{| f_{1} - f_{v} |} + {ratio}_{f_{1} + 2 f_{v}} + {ratio}_{| f_{1} - 2 f_{v} |},

Wherein,

is f₁+f_vComponent and f₁The ratio of the magnitudes of the components,

is f₁-f_vComponent and f₁The ratio of the magnitudes of the components,

is f₁+2f_vComponent and f₁The ratio of the magnitudes of the components,

is f₁-2f_vComponent and f₁The ratio of component amplitudes; f. of_vIs the natural frequency of vibration;

e. determining a fault index:

in the case where a normal motor sample reference file has not been established, the detection threshold is set to 0.1% according to conventional experience,

the ratio is the fault index;

f. judging whether the fault exists or not according to the fault index:

the fault index value is less than 1, which indicates that the motor is in a healthy state, and the smaller the value is, the more definite the healthy state is; the fault index value is greater than 1, which indicates that the motor is in a fault state, and the larger the value is, the more serious the fault state is;

to be able to filter the output signal e in step d_TIn continuously refined spectrograms_CFThe side frequency component information should calculate the difference rate s first:

Motor rotating frequency f in expressions of outer raceway fault natural frequency, inner raceway fault natural frequency, rolling element fault natural frequency and retainer fault natural frequency of bearing_rm＝(1-s)f₁。

2. A method for detecting bearing faults of an asynchronous motor as claimed in claim 1, characterized in that, in the normal condition of the motor bearing, the current f is determined according to the slip s_CFThe side frequency component and f₁Ratio of component amplitudes