CN117454300A

CN117454300A - Motor abnormal sound detection method and device, electronic equipment and storage medium

Info

Publication number: CN117454300A
Application number: CN202311767952.1A
Authority: CN
Inventors: 张宇航; 徐鹏; 付旭; 孙通
Original assignee: Midea Group Co Ltd; GD Midea Air Conditioning Equipment Co Ltd
Current assignee: Midea Group Co Ltd; GD Midea Air Conditioning Equipment Co Ltd
Priority date: 2023-12-21
Filing date: 2023-12-21
Publication date: 2024-01-26
Anticipated expiration: 2043-12-21
Also published as: CN117454300B

Abstract

The application relates to the technical field of data processing, and provides a motor abnormal sound detection method, a device, electronic equipment and a storage medium, wherein the method comprises the following steps: acquiring a target operation sound signal of a motor to be detected; carrying out wiener filtering on the target operation sound signal based on the wiener filter to obtain a wiener filtering signal; determining a motor operating frequency based on the wiener filtered signal; determining a low frequency region, a medium frequency region and a high frequency region based on the motor operating frequency; respectively extracting the characteristics of sound signals in a low-frequency area, a medium-frequency area and a high-frequency area in the target operation sound signals to obtain signal characteristic data; and detecting abnormal sound of the motor based on the signal characteristic data to obtain an abnormal sound detection result of the motor to be detected. The method and the device can rapidly and accurately determine whether the motor to be detected has the abnormal sound detection result of the abnormal sound, so that the abnormal sound detection efficiency of the motor can be improved.

Description

Motor abnormal sound detection method and device, electronic equipment and storage medium

Technical Field

The present disclosure relates to the field of data processing technologies, and in particular, to a method and apparatus for detecting abnormal noise of a motor, an electronic device, and a storage medium.

Background

In the production process of the household appliance motor, due to high yield, fast production rhythm and strict cost control, the problem of abnormal motor sounds generated by manufacturing errors in a large batch can occur. Abnormal sounds such as friction, shaking and knocking sounds are commonly caused by misalignment of upper and lower shafts, abnormal number of magnetic shoes in partial areas, abrasion of bearing parts and the like. Many of these abnormal sounds are periodic and have poor sound quality, which causes user dissatisfaction, and on the other hand, the abnormal sounds reveal problems of the motor itself, which causes functional problems in long-term use.

Because the production environment is noisy, the production environment has high-frequency sound and medium-frequency sound of human voice of machine operation and low-frequency resonance and other sounds generated by carrying equipment, general enterprises directly erect a mute room in a production line, and a manual sound hearing method is adopted in the mute room to carry out sampling inspection or full inspection on the motor. The mute room with excellent performance is high in cost, and occupies a large volume to achieve the sound insulation level. The enterprise side needs to carry out professional training to staff, and the human cost is comparatively high. The motor audio test process is tedious and boring, is easy to cause fatigue, leads to judgment errors, and the hearing standard of each person is different, so that the uniformity of production quality control cannot be ensured. Thereby resulting in low efficiency in currently performing motor abnormal sound detection.

Disclosure of Invention

The present application aims to solve at least one of the technical problems existing in the related art. Therefore, the motor abnormal sound detection method can improve the speed and accuracy of motor abnormal sound detection, and therefore the motor abnormal sound detection efficiency is improved.

The application also provides a motor abnormal sound detection device, electronic equipment and a storage medium.

According to an embodiment of the first aspect of the present application, a motor abnormal sound detection method includes:

acquiring a target operation sound signal of a motor to be detected;

carrying out wiener filtering on the target operation sound signal based on a wiener filter to obtain a wiener filtering signal;

determining a motor operating frequency based on the wiener filtered signal;

determining a low-frequency region, a medium-frequency region and a high-frequency region based on the motor operating frequency;

respectively extracting characteristics of sound signals in the low-frequency area, the medium-frequency area and the high-frequency area in the target operation sound signals to obtain signal characteristic data;

and detecting abnormal sound of the motor based on the signal characteristic data to obtain an abnormal sound detection result of the motor to be detected.

According to the motor abnormal sound detection method, the wiener filter is pre-constructed, so that the target operation sound signal of the motor to be detected can be subjected to wiener filtering through the wiener filter, noise reduction processing is conducted on the target operation sound signal, the motor operation frequency can be accurately determined based on the wiener filtered signal, the motor operation frequency is accurately divided into a low-frequency area, a medium-frequency area and a high-frequency area, finally, the sound signals located in the frequency areas in the target operation sound signal can be respectively subjected to feature extraction, and whether abnormal sound detection results of abnormal sounds exist in the motor to be detected or not can be rapidly and accurately determined according to the extracted signal feature data, so that the motor abnormal sound detection efficiency can be improved.

According to one embodiment of the present application, the determining the motor operating frequency based on the wiener filtered signal includes:

performing short-time Fourier decomposition on the wiener filtering signal to obtain a decomposed signal;

windowing is carried out on the decomposed signal to obtain a windowed signal;

and determining the motor operating frequency based on the windowing signal.

According to one embodiment of the present application, the determining the motor operating frequency based on the windowed signal includes:

median filtering is carried out on the windowing signals in the time domain direction and the frequency domain direction respectively, and a time domain median filtering signal and a frequency domain median filtering signal are obtained respectively;

and determining the motor operating frequency based on the time domain median filtering signal and the frequency domain median filtering signal.

According to one embodiment of the present application, the determining the motor operating frequency based on the time-domain median filtered signal and the frequency-domain median filtered signal includes:

respectively carrying out binarization processing on the time domain median filtering signal and the frequency domain median filtering signal to respectively obtain a time domain binarization signal and a frequency domain binarization signal;

and determining the motor operating frequency based on the time domain binarization signal and the frequency domain binarization signal.

According to one embodiment of the present application, the determining the motor operating frequency based on the time domain binarized signal and the frequency domain binarized signal includes:

determining the average energy of the fast Fourier transform in a first preset frequency range on the frequency domain binarization signal;

determining a motor operating frequency range of the frequency domain binarization signal within a second preset frequency range based on the fast Fourier transform average energy;

and determining the motor operating frequency based on the motor operating frequency range and the time domain binarization signal.

According to one embodiment of the present application, the determining the motor operating frequency based on the motor operating frequency range and the time domain binarized signal includes:

determining the corresponding skewness of each frequency in the motor operating frequency range on the time domain binarization signal;

based on each of the skewness, a motor operating frequency is determined from the motor operating frequency range.

According to an embodiment of the present application, the extracting features of the sound signals located in the low frequency region, the intermediate frequency region, and the high frequency region in the target operation sound signal to obtain signal feature data includes:

Performing autocorrelation quantity extraction on sound signals in the low-frequency region and the high-frequency region in the target operation sound signals to obtain autocorrelation quantity characteristic data;

performing kurtosis extraction on the sound signals positioned in the intermediate frequency region in the target operation sound signals to obtain kurtosis characteristic data;

signal feature data is generated based on the autocorrelation quantity feature data and the kurtosis feature data.

According to one embodiment of the present application, in performing autocorrelation amount extraction on the sound signals located in the low frequency region and the high frequency region in the target operation sound signal, the following steps are performed for each sound signal located in the low frequency region and the high frequency region, respectively:

performing Hilbert transform on the current sound signal to obtain a transformed signal;

determining an envelope spectrum based on the transformed signal and a current sound signal;

removing direct current components based on the envelope spectrum to obtain an envelope signal;

carrying out time delay on the envelope signal to obtain a time delay signal;

an autocorrelation amount of the delayed signal and a current sound signal is determined.

According to one embodiment of the application, the filter coefficient of the wiener filter is obtained by solving a wiener hoff equation based on an operation sound signal of the sample motor in a first environment and an operation sound signal of the sample motor in a second environment; the first environment is an environment with external interference, and the second environment is an environment without external interference.

According to an embodiment of the second aspect of the present application, a motor abnormal sound detection device includes:

the acquisition module is used for acquiring a target operation sound signal of the motor to be detected;

the filtering module is used for carrying out wiener filtering on the target operation sound signal based on a wiener filter to obtain a wiener filtering signal;

the first determining module is used for determining the running frequency of the motor based on the wiener filtering signal;

the second determining module is used for determining a low-frequency area, a medium-frequency area and a high-frequency area based on the motor operating frequency;

the extraction module is used for respectively carrying out feature extraction on the sound signals positioned in the low-frequency area, the intermediate-frequency area and the high-frequency area in the target operation sound signal to obtain signal feature data;

and the detection module is used for detecting abnormal sound of the motor based on the signal characteristic data to obtain an abnormal sound detection result of the motor to be detected.

An electronic device according to an embodiment of the third aspect of the present application includes a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor implements any one of the above-mentioned motor abnormal sound detection methods when executing the program.

According to a fourth aspect of the present application, the storage medium is a non-transitory computer readable storage medium, on which a computer program is stored, which when executed by a processor implements a motor abnormal sound detection method as described in any one of the above.

The above technical solutions in the embodiments of the present application have at least one of the following technical effects:

through constructing the wiener filter in advance, the wiener filter can be used for carrying out wiener filtering on a target operation sound signal of the motor to be detected, so that noise reduction processing is carried out on the target operation sound signal, and further, the motor operation frequency can be accurately determined based on the wiener filtered signal, so that the frequency is accurately divided into a low-frequency area, a medium-frequency area and a high-frequency area through the motor operation frequency, finally, the sound signals positioned in the frequency areas in the target operation sound signal can be respectively subjected to feature extraction, and whether the abnormal sound detection result of the motor to be detected exists or not can be rapidly and accurately determined according to the extracted signal feature data, so that the abnormal sound detection efficiency of the motor can be improved.

Additional aspects and advantages of the application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the application.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the related art, the drawings that are required to be used in the embodiments or the related technical descriptions will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and other drawings may be obtained according to the drawings without inventive effort for a person having ordinary skill in the art.

Fig. 1 is a flow chart of a motor abnormal sound detection method provided in an embodiment of the present application;

fig. 2 is an overall flow chart of a motor abnormal sound detection method provided in an embodiment of the present application;

fig. 3 is a schematic structural diagram of a motor abnormal sound detection device according to an embodiment of the present application;

fig. 4 is a schematic structural diagram of an electronic device provided in the present application.

Detailed Description

Embodiments of the present application are described in further detail below with reference to the accompanying drawings and examples. The following examples are illustrative of the present application but are not intended to limit the scope of the present application.

In the description of the embodiments of the present application, it should be noted that, directions or positional relationships indicated by terms such as "center", "longitudinal", "lateral", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., are based on those shown in the drawings, are merely for convenience in describing the embodiments of the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a specific orientation, be configured and operated in a specific orientation, and thus should not be construed as limiting the embodiments of the present application. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.

In the description of the embodiments of the present application, it should be noted that, unless explicitly specified and limited otherwise, the terms "connected," "connected," and "connected" are to be construed broadly, and may be, for example, fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium. The specific meaning of the terms in the embodiments of the present application will be understood by those of ordinary skill in the art in a specific context.

In the examples herein, a first feature "on" or "under" a second feature may be either the first and second features in direct contact, or the first and second features in indirect contact via an intermediary, unless expressly stated and defined otherwise. Moreover, a first feature being "above," "over" and "on" a second feature may be a first feature being directly above or obliquely above the second feature, or simply indicating that the first feature is level higher than the second feature. The first feature being "under", "below" and "beneath" the second feature may be the first feature being directly under or obliquely below the second feature, or simply indicating that the first feature is less level than the second feature.

In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the embodiments of the present application. In this specification, schematic representations of the above terms are not necessarily directed to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, the different embodiments or examples described in this specification and the features of the different embodiments or examples may be combined and combined by those skilled in the art without contradiction.

As can be appreciated, conventional methods employ sound level weighting to monitor more powerful anomaly sounds, which cannot resolve less powerful anomaly samples. Modern methods mostly adopt sound quality indexes for judgment, such as comprehensive measurement of loudness, sharpness, roughness and the like. The method is essentially to calculate the difference in statistics between the sample to be measured and the standard sample, and based on the thought, the Euclidean distance of the short-time feature group is calculated by adopting a word bag model.

However, the prior knowledge of the motor rotating structure is not fully considered in the method, and a large number of high-intensity magnetic fields such as magnetism charging and discharging and high-power equipment exist in the motor production environment, so that interference such as harmonic waves and white noise can be generated on the acquisition equipment, and if the pretreatment filtering is not fully effective, the subsequent analysis is difficult.

Fig. 1 is a flow chart of a motor abnormal sound detection method provided in an embodiment of the present application, as shown in fig. 1, the motor abnormal sound detection method includes:

step 110, a target operation sound signal of the motor to be detected is obtained.

It should be noted that, the execution body of the motor abnormal sound detection method provided in the embodiments of the present application may be a server, a computer device, such as a mobile phone, a tablet computer, a notebook computer, a palm computer, a vehicle-mounted electronic device, a wearable device, an Ultra-mobile personal computer (Ultra-Mobile Personal Computer, UMPC), a netbook, a personal digital assistant (Personal Digital Assistant, PDA), or the like.

The execution main body of the motor abnormal sound detection device can be arranged, and the motor abnormal sound detection method can be realized by controlling the motor abnormal sound detection device.

The motor to be detected in the method can be a household appliance motor needing abnormal sound detection.

It should be noted that the motor to be detected may be a complete motor that has been produced but not assembled.

Therefore, the abnormal sound detection can be performed on the motor to be detected in a general motor production environment (namely, an environment with external interference, such as an environment with interference of harmonic waves, white noise and the like generated by a collection device due to a large amount of high-intensity magnetic field and high-power equipment such as magnetism charging and discharging, which can be defined as a first environment in the application).

The method and the device can collect sound signals generated by running of each motor to be detected in the motor production environment, and determine the collected sound signals as target running sound signals of the motors to be detected.

The method does not limit the collection process and collection equipment of the sound signals of the motor operation, and can select the collection equipment and the collection method according to actual requirements and precision requirements.

And 120, carrying out wiener filtering on the target operation sound signal based on the wiener filter to obtain a wiener filtering signal.

In the application, a wiener filter can be pre-constructed, and filter coefficients can be solved through a wiener Huffman equation.

In particular, the present application can collect the operation sound signal of the sample motor in the second environment without external interference and form a data set Collecting the operation sound signal of the sample motor in the actual production environment (i.e. the first environment) and forming a data set +.>Wherein the second environment may be, for example, mute for isolating external interferenceA house.

After two data sets are acquired, the data sets acquired in the laboratory environment are acquiredAs a desired signal +.>The actual production environment data set X is used as an observation signal +.>. According to the desired signal->And observe signal->Carrying out wiener Hough equation solving to obtain filter coefficients of the wiener filter, wherein the method comprises the following specific steps:

Step1.1：

defining parameters of an optimal filter by mean square errorObserving signal->The output of the signal through the filter is +.>The mean square error is defined as:

；

wherein,representing the nth desired signal;

Step1.2：

calculating mean square error with respect to wiener filter coefficientsIs the derivative of:

；

Step1.3：

calculation ofAnd->And let the derivative of the mean square error be 0,/or%>The nth observed signal is represented by:

；

wherein,is->Is self-correlated (I)>Is->And->Is defined as:

；

wherein b is a time unit,for n to the left by b time units.

Step1.4：

Therefore, the wiener Hough equation can be constructed according to the autocorrelation information of the operation sound signal of the motor in the production environment, the cross-correlation information of the operation sound signal of the motor in the first environment and the operation sound signal in the second environment and the filter coefficient The method comprises the following steps:

；

when (when)Positive timing, the filter coefficients can pass +.>Obtaining the product.

Therefore, after the target operation sound signal is acquired, the target operation sound signal can be input into the wiener filter for wiener filtering, and the signal output by the wiener filter is determined as a wiener filtering signal.

Step 130, determining the motor operating frequency based on the wiener filtered signal.

After the wiener filter signal is obtained, the signal after noise reduction of the wiener filter can be subjected to short-time Fourier decomposition and windowing.

For the windowed signals, the signals can be respectively in the time domainAnd frequency domain->And median filtering is carried out in the direction to strengthen concentrated and continuous signals in the frequency domain and weaken discontinuous signals in the time domain, so that frequency domain signals and time domain signals are obtained.

Further, binarization processing can be performed on the time domain signal and the frequency domain signal respectively, so as to obtain a binarized signal in the time domain and a binarized signal in the frequency domain.

Further, a fast fourier transform (Fast Fourier Transform, fft) can be performed on the time-domain binarized signal to calculate fft average energy.

Further, extracting a set of possible ranges of motor operation according to the fft average energy, and calculating the frequency corresponding to the maximum deviation in the set on the frequency domain signal to obtain the motor operation frequency 。

Step 140, determining a low frequency region, a medium frequency region and a high frequency region based on the motor operating frequency.

The application relates to the determination of motor operating frequencyAfter that, [400Hz, ] -can be defined>) In the low-frequency region of the band,is of intermediate frequency region, [ ->，10000Hz]Is a high frequency region.

And step 150, respectively extracting the characteristics of the sound signals in the low-frequency region, the medium-frequency region and the high-frequency region in the target operation sound signals to obtain signal characteristic data.

After the division of the low-frequency region, the intermediate-frequency region and the high-frequency region is completed, the method and the device can respectively conduct feature extraction on sound signals located in different regions in the target operation sound signals in different modes, so that overall signal feature data are obtained.

Specifically, the present application may perform autocorrelation quantity extraction by adopting an envelope signal autoregressive manner with respect to a sound signal located in a low frequency region and a high frequency region in a target operation sound signal.

And performing kurtosis extraction on a sound signal located in the intermediate frequency region in the target operation sound signal.

Signal characteristic data consisting of characteristics of all sound signals in the target operation sound signal are thereby obtained.

Step 160, detecting abnormal sound of the motor based on the signal characteristic data to obtain an abnormal sound detection result of the motor to be detected.

After the signal characteristic data are obtained, the method and the device can respectively obtain a preset kurtosis threshold value and a preset autocorrelation quantity threshold value. In one embodiment, the kurtosis threshold may be 0.9 and the autocorrelation amount threshold may be 0.7.

Further, the respective correlation amounts in the signal feature data may be compared with the autocorrelation amount threshold, respectively, and the kurtosis in the signal feature data may be compared with the kurtosis threshold.

And if the comparison determines that any autocorrelation quantity in the signal characteristic data is greater than or equal to an autocorrelation quantity threshold value or any kurtosis is greater than or equal to a kurtosis threshold value, determining that abnormal sound exists in the motor to be detected.

And if all the autocorrelation amounts in the signal characteristic data are smaller than the autocorrelation amount threshold and all the kurtosis is smaller than the kurtosis threshold, determining that the motor to be detected does not have abnormal sound.

Based on the above embodiment, the step 130 may include:

step 131, performing short-time Fourier decomposition on the wiener filtered signal to obtain a decomposed signal;

step 132, windowing is carried out on the decomposed signal to obtain a windowed signal;

step 133, determining the motor operating frequency based on the windowed signal.

The application can carry out short-time Fourier decomposition and windowing treatment on the wiener filter signal through the following calculation formula:

；

wherein,is the angular frequency e ^-iωt Is the Euler's formula, which represents a complex number, cos (ωt) -isin (ωt). This formula can be divided into three parts: time domain signal f (t), complex exponential function e ^-iωt And an integration operation. />A hann window function referring to a center point of 0,>is a signal subjected to fourier transformation, +.>Is the delay time of the window function. The Hann window function is:

；

wherein the method comprises the steps ofFor the length of the window function, +.>Is the magnitude of the window function.

Thereby, a windowed signal can be obtained.

Further toThe application can be used for windowing signals in the time domain respectivelyAnd frequency domain->And median filtering is carried out in the direction to strengthen concentrated and continuous signals in the frequency domain and weaken discontinuous signals in the time domain, so that frequency domain signals and time domain signals are obtained.

Further, the fft average energy may be obtained by performing a fast fourier transform calculation on the time-domain binarized signal.

Further, extracting a set of possible ranges of motor operation according to the fft average energy, and calculating the frequency corresponding to the maximum deviation in the set on the frequency domain signal to obtain the motor operation frequency。

According to the embodiment, the signals subjected to noise reduction by the wiener filter are subjected to short-time Fourier decomposition and windowed by using a Hann window function, so that more accurate, smooth and visual frequency spectrum information can be provided, the motor operating frequency can be conveniently and accurately determined according to the windowed signals, the frequency is accurately divided into a low-frequency area, a medium-frequency area and a high-frequency area through the motor operating frequency, finally, the characteristics of sound signals positioned in each frequency area in a target operating sound signal can be respectively extracted, and the abnormal sound detection result of whether the motor to be detected has abnormal sound can be rapidly and accurately determined according to the extracted signal characteristic data, so that the abnormal sound detection efficiency of the motor can be improved.

Based on the above embodiment, the step 133 may include:

step 1331, median filtering is performed on the windowed signal in the time domain direction and the frequency domain direction respectively, so as to obtain a time domain median filtered signal and a frequency domain median filtered signal respectively;

step 1332, determining the motor operating frequency based on the time domain median filtered signal and the frequency domain median filtered signal.

After the windowing signals are obtained, the windowing signals can be respectively in the time domainAnd frequency domain->Filtering in the direction by using a median filter with length of 5 to strengthen concentrated and continuous signals in the frequency domain and weaken discontinuous signals in the time domain to obtain a frequency domain median filtering signal->And a time-domain median-filtered signal->。

Wherein, the median filter is:

；

wherein,is a real signal sequence->Is a signal of (a).

Wherein, the median filtering on the frequency domain is:

；

the median filtering in the time domain is:

；

wherein,the spectrum after STFT is the spectrum of the windowed signal; />And->The lengths of the frequency domain filter and the time domain filter, respectively. m and k represent the window size and pixel depth of the median filter, respectively.

Further, the frequency domain median filtered signal can beAnd a time-domain median filtered signalAnd respectively performing binarization processing to obtain a binarization signal on a time domain and a binarization signal on a frequency domain.

According to the embodiment, median filtering is respectively carried out on the windowed signals in the time domain direction and the frequency domain direction, concentrated and continuous signals in the frequency domain can be enhanced, discontinuous signals in the time domain are weakened, and the motor operation frequency can be conveniently and accurately determined according to the frequency domain median filtering signals and the time domain median filtering signals, so that the frequency is accurately divided into a low frequency region, an intermediate frequency region and a high frequency region through the motor operation frequency, finally, characteristic extraction can be carried out on sound signals positioned in each frequency region in a target operation sound signal respectively, and whether abnormal sound detection results of the motor to be detected exist or not can be rapidly and accurately determined according to the extracted signal characteristic data, so that the motor abnormal sound detection efficiency can be improved.

Based on the above embodiment, the step 1332 may include:

step 13321, respectively performing binarization processing on the time domain median filtering signal and the frequency domain median filtering signal to obtain a time domain binarization signal and a frequency domain binarization signal;

Step 13322, determining the motor operating frequency based on the time domain binarized signal and the frequency domain binarized signal.

The method can respectively carry out binarization processing on the time-domain median filtering signal and the frequency-domain median filtering signal by using a binarization filter to obtain a frequency-domain binarization signalAnd a time domain binarized signal->。

The binarization filter includes a frequency domain binarization filter and a time domain binarization filter.

Wherein, the frequency domain binarization filter is:

；

the time domain binarization filter in the step is as follows:

；

the frequency domain binarization filtering is as follows:

；

the time domain binarization filtering is as follows:

；

wherein,refers to the spectrum after STFT by short-time Fourier transform, < >>Representing a point-wise multiplication of the matrix. m and k represent different thresholds.

According to the embodiment, the binarization filter is used for respectively carrying out binarization processing on the time domain median filtering signal and the frequency domain median filtering signal, so that noise interference can be reduced, the motor operating frequency can be conveniently and accurately determined according to the time domain binarization signal and the frequency domain binarization signal, the frequency is accurately divided into a low frequency area, an intermediate frequency area and a high frequency area through the motor operating frequency, finally, characteristic extraction can be carried out on sound signals positioned in each frequency area in a target operating sound signal respectively, and whether abnormal sound detection results of the motor to be detected exist or not can be rapidly and accurately determined according to the extracted signal characteristic data, so that the motor abnormal sound detection efficiency can be improved.

Based on the above embodiment, the step 13322 may include:

step 133221, determining the mean energy of the fast fourier transform within a first preset frequency range on the frequency domain binarized signal;

step 133222, determining a motor operating frequency range of the frequency domain binarized signal within a second preset frequency range based on the fast fourier transform average energy;

step 133223, determining the motor operating frequency based on the motor operating frequency range and the time domain binarized signal.

According to the method and the device, the fft average energy in the first preset frequency range can be calculated on a spectrogram of the frequency domain binarized signal, and the first preset frequency range in the method and the device can be 400Hz to 8000Hz.

Specifically, the fft average energy is:

；

；/>

wherein,representing the length of the spectral time, +.>Representing average energy +.>In the motor detection scene, select +.>；/>Represents the mean value of the fft average energy, +.>Representing the fft average energy, +.>Indicating the operating frequency range of the motor, < >>Representing the ith frequency in the second preset frequency range.

Further, a set of possible ranges of motor operation can be extracted by using a local extremum method within a second preset frequency range according to the average energy of the fast fourier transform, and a motor operation frequency range is obtained. The second preset frequency range in this application may be 400Hz to 2000Hz. The process of how to extract the set of possible ranges of motor operation by the local extremum method is not specifically limited in the application, and the use process of the local extremum method in the prior art can be applied to the scene of the application.

Further, the frequency corresponding to the maximum deviation in the motor operating frequency range can be calculated on the frequency domain signal to obtain the motor operating frequency。

According to the method, the average energy of the fast Fourier transform is calculated on the frequency domain binarization signal, then a possible range set of motor operation is extracted by using a local extremum method according to the average energy of the fast Fourier transform, and as the local extremum method can help to eliminate influence of interference signals outside an interested frequency range on a result, the accuracy of motor operation frequency extraction can be improved, so that frequencies are accurately divided into a low-frequency region, a medium-frequency region and a high-frequency region through motor operation frequencies, finally, characteristic extraction can be carried out on sound signals located in each frequency region in a target operation sound signal respectively, and whether abnormal sound detection results of a motor to be detected exist or not can be rapidly and accurately determined according to the extracted signal characteristic data, so that the motor abnormal sound detection efficiency can be improved.

Based on the above embodiment, the step 133223 may include:

step 1332231, determining the corresponding skewness of each frequency in the motor operating frequency range on the time domain binarized signal;

step 1332232, a motor operating frequency is determined from the motor operating frequency range based on the respective skewness.

The bias corresponding to each frequency in the motor operation frequency range is determined on the time domain binarization signal, and the bias can be calculated specifically through the following calculation formula:

；

wherein,the degree of deviation is indicated by the degree of deviation,/>representation->Is>Representing the mean value of the signal data corresponding to each frequency in the time domain binarized signal,/for each frequency>Represents the mean threshold value->Representing the variance of the signal data corresponding to each frequency in the time domain binarized signal.

Further, the magnitude of each deflection can be compared, and the frequency corresponding to the deflection with the largest value can be determined as the motor running frequency. Specifically, there are:

。

wherein,indicating the frequency corresponding to the deviation with the largest value,/->Representing a function for taking the maximum value. />

According to the embodiment, the deviation corresponding to each frequency in the motor operation frequency range can be determined on the time domain binarization signal, the frequency corresponding to the maximum deviation of the numerical value is determined to be the motor operation frequency, so that the motor operation frequency can be accurately obtained, the frequency can be accurately divided into a low frequency area, an intermediate frequency area and a high frequency area through the motor operation frequency, finally, the characteristic extraction can be respectively carried out on sound signals positioned in each frequency area in the target operation sound signals, and whether the abnormal sound detection result of the motor to be detected exists or not can be rapidly and accurately determined according to the extracted signal characteristic data, so that the abnormal sound detection efficiency of the motor can be improved.

Based on the above embodiment, the step 150 may include:

step 151, extracting autocorrelation quantity of the sound signals in the low frequency region and the high frequency region in the target operation sound signal to obtain autocorrelation quantity feature data;

step 152, performing kurtosis extraction on the sound signal in the intermediate frequency region of the target operation sound signal to obtain kurtosis characteristic data;

step 153 generates signal characteristic data based on the autocorrelation quantity characteristic data and the kurtosis characteristic data.

In the method, the autocorrelation quantity of the sound signals in the low-frequency region and the high-frequency region in the target operation sound signals can be extracted by using an envelope signal autoregressive method, so that the autocorrelation quantity characteristic data are obtained.

And performing kurtosis extraction on the sound signal in the intermediate frequency region in the target operation sound signal to obtain kurtosis characteristic data. The kurtosis can be determined by the following calculation formula:

；

wherein,data (energy value) for the ith sound signal in the mid-frequency region, and>is the average value of the data of all sound signals located in the mid-frequency region,/->Is the number of sound signals in the mid-frequency region.

Further, signal characteristic data may be formed from the autocorrelation quantity characteristic data and the kurtosis characteristic data.

According to the embodiment, the sound signals in different frequency areas are subjected to feature extraction according to different feature extraction methods, so that accurate signal feature data can be obtained, and whether the motor to be detected has an abnormal sound detection result of the abnormal sound or not can be rapidly and accurately determined according to the extracted signal feature data, so that the abnormal sound detection efficiency of the motor can be improved.

Further, in the step 151, the following steps may be performed for each of the sound signals located in the low frequency region and the high frequency region:

step 1511, performing hilbert transform on the current sound signal to obtain a transformed signal;

a step 1512 of determining an envelope spectrum based on the transformed signal and the current sound signal;

step 1513, removing direct current components based on the envelope spectrum to obtain an envelope signal;

step 1514, performing time delay on the envelope signal to obtain a time delay signal;

at step 1515, an autocorrelation amount of the delayed signal and the current sound signal is determined.

The present application can determine the autocorrelation amount of each sound signal located in the low frequency region and the high frequency region in a simultaneous or one-by-one processing manner. For the currently processed sound signal, it may be transformed by the following hilbert transform formula to obtain a transformed signal:

；/>

Wherein,for the original input signal, +.>90 DEG phase shift of the original input signal, < >>Representing the time delay of the signal function.

Further, the envelope spectrum may be determined according to the following envelope spectrum calculation formula:

；

further, the envelope spectrum may be subjected to dc component removal, resulting in the following envelope signal:

；

wherein,is the average of the envelope signal.

Note that the calculation of the autocorrelation is:

；

wherein,represents the complex conjugate number, when the function is a real-valued function +.>，/>Is defined as，/>Representing the convolution of the function. />Representing the time delay of the signal function.

Assuming that the signal is a real-valued signal, the autocorrelation of the envelope signal is:

；

further, the envelope signal is delayed, and the autocorrelation quantity of the delayed signal and the original signal is calculated:

；

wherein the maximum autocorrelation amountI.e. the degree of periodicity, the time delay of the signal can be represented>Representing the period of the signal.

In motor manufacturing, periodic friction with the stator during rotation is caused by rotor protrusion or foreign matter adsorption, and local high-energy sounds are generated, and the sounds show periodicity positively related to the rotation speed. Assuming a rotational speed of 3000rpm, 50 periodically arranged abnormal signals can be generated per second. After the calculation is completed Then, the motor speed s (rpm) is compared with the motor speed s (rpm), if->Then an abnormal sample may be identified.

According to the embodiment, the characteristic extraction can be carried out on each sound signal in the low-frequency area and the high-frequency area through the envelope signal autoregressive method, so that accurate autocorrelation characteristic data are obtained, whether the motor to be detected has an abnormal sound detection result of the abnormal sound or not can be conveniently and rapidly determined according to the extracted signal characteristic data, and therefore the abnormal sound detection efficiency of the motor can be improved.

Fig. 2 is a schematic overall flow chart of a motor abnormal sound detection method provided in an embodiment of the present application, as shown in fig. 2, in one embodiment, wiener filter parameters may be solved by using a laboratory data set and a production environment data set without interference, and a motor operation signal collected by a production environment may be filtered by using a wiener filter after the parameters are solved, so as to obtain a noise reduction signal.

Further, the noise reduction signal is STFT and windowed using Hann window.

Further, median filtering can be performed on the windowed signals in the time domain and frequency domain directions, respectively, so as to strengthen concentrated and continuous signals in the frequency domain and weaken discontinuous signals in the time domain.

Further, the signal may be binarized and filtered in the frequency domain direction and the time domain direction.

And calculating the fft energy within 400-4000 Hz for the frequency domain direction binarized and filtered signal, and extracting a local extremum point set according to the fft energy.

And calculating the maximum bias corresponding frequency of the signal subjected to binarization filtering in the time domain direction according to the local extreme point set, and dividing the low-middle-high frequency band according to the frequency.

Further, motor fault signatures are calculated using envelope autoregressions and kurtosis. The motor fault characteristics of the sound signals of the low frequency band and the high frequency band can be calculated by using envelope autoregressive, and the motor fault characteristics of the medium frequency band can be calculated by using kurtosis.

Fig. 3 is a schematic structural diagram of a motor abnormal sound detection device according to an embodiment of the present application, as shown in fig. 3, the motor abnormal sound detection device includes:

an acquisition module 310, configured to acquire a target operation sound signal of a motor to be detected;

the filtering module 320 is configured to perform wiener filtering on the target operation sound signal based on a wiener filter, so as to obtain a wiener filtered signal;

a first determination module 330 for determining a motor operating frequency based on the wiener filtered signal;

a second determining module 340, configured to determine a low frequency region, a medium frequency region, and a high frequency region based on the motor operating frequency;

The extracting module 350 is configured to perform feature extraction on the sound signals located in the low-frequency region, the intermediate-frequency region, and the high-frequency region in the target operation sound signal, so as to obtain signal feature data;

and the detection module 360 is configured to perform abnormal sound detection on the motor based on the signal characteristic data, so as to obtain an abnormal sound detection result of the motor to be detected.

According to the motor abnormal sound detection device, the wiener filter is pre-constructed, so that the target operation sound signal of the motor to be detected can be subjected to wiener filtering through the wiener filter, noise reduction processing is conducted on the target operation sound signal, the motor operation frequency can be accurately determined based on the wiener filtered signal, the frequency is accurately divided into a low-frequency area, a medium-frequency area and a high-frequency area through the motor operation frequency, finally, the sound signals located in the frequency areas in the target operation sound signal can be respectively subjected to feature extraction, and whether abnormal sound detection results of abnormal sounds exist in the motor to be detected can be rapidly and accurately determined according to the extracted signal feature data, so that the motor abnormal sound detection efficiency can be improved.

Based on any of the above embodiments, the first determining module 330 is specifically configured to:

windowing is carried out on the decomposed signal to obtain a windowed signal;

and determining the motor operating frequency based on the windowing signal.

Based on any of the above embodiments, the first determining module 330 includes a first determining unit configured to:

Based on any of the above embodiments, the first determining module 330 includes a second determining unit for:

Based on any of the above embodiments, the first determining module 330 includes a third determining unit for:

Based on any of the above embodiments, the first determining module 330 includes a fourth determining unit for:

Based on any of the above embodiments, the extraction module 350 is specifically configured to:

Based on any of the above embodiments, the extraction module 350 comprises an extraction unit for:

carrying out time delay on the envelope signal to obtain a time delay signal;

Fig. 4 illustrates a physical schematic diagram of an electronic device, as shown in fig. 4, which may include: processor 410, communication interface (Communications Interface) 420, memory 430 and communication bus 440, wherein processor 410, communication interface 420 and memory 430 communicate with each other via communication bus 440. The processor 410 may call logic instructions in the memory 430 to perform the following method: acquiring a target operation sound signal of a motor to be detected;

determining a motor operating frequency based on the wiener filtered signal;

Further, the logic instructions in the memory 430 described above may be implemented in the form of software functional units and may be stored in a computer-readable storage medium when sold or used as a stand-alone product. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the related art or in a part of the technical solution, in the form of a software product stored in a storage medium, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.

In yet another aspect, embodiments of the present application further provide a non-transitory computer readable storage medium having stored thereon a computer program that, when executed by a processor, is implemented to perform the method provided by the above embodiments, for example, comprising: acquiring a target operation sound signal of a motor to be detected;

determining a motor operating frequency based on the wiener filtered signal;

The apparatus embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.

From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on such understanding, the foregoing technical solution may be embodied essentially or in a part contributing to the related art in the form of a software product, which may be stored in a computer readable storage medium, such as ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform the method described in the respective embodiments or some parts of the embodiments.

Finally, it should be noted that the above embodiments are only for illustrating the present application, and are not limiting of the present application. While the present application has been described in detail with reference to the embodiments, those skilled in the art will understand that various combinations, modifications, or equivalents of the technical solutions of the present application may be made without departing from the spirit and scope of the technical solutions of the present application, and all such modifications are intended to be covered by the claims of the present application.

Claims

1. A motor abnormal sound detection method, characterized by comprising:

Acquiring a target operation sound signal of a motor to be detected;

determining a motor operating frequency based on the wiener filtered signal;

2. The motor abnormal sound detection method according to claim 1, wherein the determining the motor operation frequency based on the wiener filtered signal includes:

windowing is carried out on the decomposed signal to obtain a windowed signal;

and determining the motor operating frequency based on the windowing signal.

3. The motor abnormal sound detection method according to claim 2, wherein the determining the motor operation frequency based on the windowed signal includes:

4. A motor abnormal sound detection method according to claim 3, wherein said determining a motor operating frequency based on said time-domain median filtered signal and said frequency-domain median filtered signal comprises:

5. The motor abnormal sound detection method according to claim 4, wherein the determining the motor operating frequency based on the time domain binarized signal and the frequency domain binarized signal includes:

6. The motor abnormal sound detection method according to claim 5, wherein the determining the motor operating frequency based on the motor operating frequency range and the time domain binarized signal includes:

7. The method for detecting abnormal sounds of a motor according to claim 1, wherein the feature extracting the sound signals in the low frequency region, the intermediate frequency region and the high frequency region in the target operation sound signal respectively to obtain signal feature data includes:

8. The motor abnormal sound detection method according to claim 7, wherein, in performing autocorrelation amount extraction on the sound signals located in the low frequency region and the high frequency region in the target operation sound signal, the following steps are performed for each sound signal located in the low frequency region and the high frequency region, respectively:

carrying out time delay on the envelope signal to obtain a time delay signal;

9. The motor abnormal sound detection method according to any one of claims 1 to 8, wherein the filter coefficient of the wiener filter is obtained by performing wiener hough equation solution based on an operation sound signal of the sample motor in a first environment and an operation sound signal in a second environment; the first environment is an environment with external interference, and the second environment is an environment without external interference.

10. A motor abnormal sound detection device, characterized by comprising:

11. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the motor abnormal sound detection method according to any one of claims 1 to 9 when executing the program.

12. A storage medium being a non-transitory computer readable storage medium having stored thereon a computer program, characterized in that the computer program, when executed by a processor, implements the motor abnormal sound detection method according to any one of claims 1 to 9.