Household appliance abnormal sound detection method, device and system

Abstract

The application relates to the technical field of household appliance manufacturing, and provides a household appliance abnormal sound detection method, device and system, the method comprising the following steps: acquiring at least one to-be-detected vibration signal of a to-be-detected household appliance during operation; generating a to-be-detected spectrogram corresponding to each to-be-detected vibration signal; inputting each to-be-detected spectrogram into an abnormal sound detection model respectively to obtain an abnormal sound detection result output by the abnormal sound detection model; in the application, the abnormal sound detection model is trained based on a sample spectrogram and an abnormal sound type label corresponding to the sample spectrogram, the sample spectrogram is generated based on a sample vibration signal, therefore, the abnormal sound detection model can accurately detect various abnormal sounds of the to-be-detected household appliance through image recognition on each to-be-detected spectrogram, can accurately determine the abnormal sound type, and the detection result has better consistency compared with artificial sound detection.

Description

Technical Field

[0001] This invention relates to the field of home appliance manufacturing technology, and in particular to a method, device and system for detecting abnormal sounds in home appliances. Background Technology

[0002] Common household appliances include refrigerators, air conditioners, and microwave ovens. These appliances may produce abnormal noises during use, which could be warning signs of equipment malfunction.

[0003] For example, a microwave oven works by using microwave radiation to heat water molecules in food, causing them to vibrate rapidly and generate heat, thus achieving rapid heating. Normally, using a microwave oven simply involves setting the heating time and power to start the heating program. Inside, a motor-driven turntable or stirrer ensures even heating of the food. However, during use, microwave ovens may exhibit abnormal noises such as magnetron aging, motor malfunction, loose door seals and / or screws, and electromagnetic failures. These abnormal noises may be warning signs of equipment failure and can also affect the user experience.

[0004] Currently, most detection of abnormal sounds in home appliances is done manually by listening to the sounds. This means that technicians listen to the sounds to determine if there are any abnormal noises when the appliances are working. However, listening for extended periods can cause fatigue, affecting the quality of the detection and leading to inaccurate results. Moreover, manual listening is highly subjective, and there are significant differences in human experience. Different people may obtain different results for the same device, resulting in poor consistency and a lack of standardized testing criteria. Summary of the Invention

[0005] This invention provides a method, device, and system for detecting abnormal sounds from home appliances, which solves the technical problems of inaccurate detection and poor consistency of detection results caused by manual listening in the prior art.

[0006] This invention provides a method for detecting abnormal sounds from household appliances, comprising the following steps:

[0007] Acquire at least one vibration signal to be tested during the operation of the appliance under test;

[0008] Generate the spectrum diagrams of each vibration signal to be measured;

[0009] Each of the spectrograms to be tested is input into the abnormal sound detection model to obtain the abnormal sound detection result output by the abnormal sound detection model;

[0010] The abnormal sound detection model is trained based on the sample spectrogram and the abnormal sound type label corresponding to the sample spectrogram, and the sample spectrogram is generated based on the sample vibration signal.

[0011] According to the present invention, a method for detecting abnormal sounds in household appliances generates a spectrogram corresponding to each of the vibration signals to be tested, including:

[0012] Short-time Fourier transform is performed on each of the vibration signals to be measured to convert each vibration signal to be measured from a time domain signal to a frequency domain signal;

[0013] Determine the amplitude spectrum of each time window in the corresponding short-time Fourier transform of each frequency domain signal;

[0014] For each frequency domain signal, the amplitude spectra of all time windows are arranged in chronological order to generate the spectrograms of each signal to be tested.

[0015] According to a method for detecting abnormal noises in household appliances provided by the present invention, short-time Fourier transforms are performed on each of the vibration signals to be tested to convert each of the vibration signals to be tested from a time-domain signal to a frequency-domain signal, including:

[0016] Each of the vibration signals to be measured is divided into multiple time windows, and windowing is applied between two adjacent time windows.

[0017] Each of the measured vibration signals after windowing is subjected to a Fast Fourier Transform to obtain the frequency components within each time window, thereby converting each measured vibration signal from a time-domain signal to a frequency-domain signal.

[0018] According to the present invention, a method for detecting abnormal sounds in household appliances, after arranging the amplitude spectra of all time windows in chronological order to generate the spectrograms to be tested for each frequency domain signal, further includes:

[0019] Logarithms are taken on the amplitude spectra of each of the proposed spectrograms to be tested to obtain logarithmically transformed spectrograms. The proposed spectrograms to be tested for inputting the abnormal sound detection model are logarithmically transformed spectrograms. During training of the abnormal sound detection model, the sample spectrograms are logarithmically transformed sample spectrograms.

[0020] According to the present invention, a method for detecting abnormal sounds in home appliances is provided, and the training process of the abnormal sound detection model is as follows:

[0021] The sample spectrogram is input into the initial convolutional neural network model to obtain the abnormal sound prediction result output by the initial convolutional neural network model;

[0022] Substitute the abnormal sound prediction results and the abnormal sound type labels corresponding to the sample spectrograms into the loss function. If the loss function converges, the training is completed, and the abnormal sound detection model is obtained.

[0023] According to the present invention, a method for detecting abnormal noises in household appliances, after acquiring at least one vibration signal to be tested during the operation of the household appliance under test, further includes:

[0024] Based on the vibration characteristics of the motor under test vibration signal in each rotation cycle of the motor in the home appliance, abnormal noise of the motor is detected. The vibration signal of the motor under test is obtained based on the vibration sensor located in the corresponding area of ​​the motor.

[0025] According to the present invention, a method for detecting abnormal noises in household appliances is provided, which detects abnormal noises based on the vibration characteristics of the vibration signal of the motor under test in each rotation cycle of the motor in the household appliance, including:

[0026] Spectral analysis is performed on the vibration signal of the motor under test within each rotation cycle to determine the target frequency band of the vibration energy abrupt change.

[0027] The number of vibration energy exceeding the energy threshold in the target frequency band is counted as a first target number of cycles. When the ratio of the first target number of cycles to the total number of cycles reaches a first preset ratio threshold, the abnormal noise of the motor is determined to be an abnormal noise caused by friction between the motor rotor and foreign objects inside the motor. The total number of cycles is determined based on the rotation cycle of the motor and the total duration of the vibration signal of the motor under test.

[0028] According to the present invention, a method for detecting abnormal noises in household appliances is provided, which detects abnormal noises based on the vibration characteristics of the vibration signal of the motor under test in each rotation cycle of the motor in the household appliance, including:

[0029] Kujic analysis is performed on the vibration signal of the motor under test in each rotation cycle to obtain the kujicic value in each rotation cycle;

[0030] The number of second target cycles in which the kurtosis value exceeds the kurtosis threshold is counted. When the ratio of the second target cycle number to the total number of cycles reaches a second preset ratio threshold, the abnormal noise of the motor is determined to be abnormal noise of the motor bearing. The total number of cycles is determined based on the rotation cycle of the motor and the total duration of the vibration signal of the motor under test.

[0031] The present invention also provides a device for detecting abnormal sounds in household appliances, comprising the following modules:

[0032] The vibration signal acquisition module is used to acquire at least one vibration signal to be tested when the appliance is running.

[0033] The spectrogram generation module is used to generate spectrograms corresponding to each of the vibration signals to be tested.

[0034] The model execution module is used to input each of the spectrograms to be tested into the abnormal sound detection model, and obtain the abnormal sound detection result output by the abnormal sound detection model.

[0035] The abnormal sound detection model is trained based on the sample spectrogram and the abnormal sound type label corresponding to the sample spectrogram, and the sample spectrogram is generated based on the sample vibration signal.

[0036] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the program to implement the household appliance abnormal sound detection method as described above.

[0037] The present invention also provides a home appliance abnormal sound detection system, comprising: an assembly line, a lifting mechanism, a vibration detection mechanism, and the aforementioned electronic equipment;

[0038] The production line is equipped with a tooling plate for carrying the household appliances to be tested;

[0039] The lifting mechanism is located below the production line without contact with it, and is used to lift the tooling plate;

[0040] The vibration detection mechanism is located above the lifting mechanism and is used to transmit at least one vibration signal from the household appliance under test to the electronic device.

[0041] According to the present invention, a home appliance abnormal sound detection system further includes a mounting frame, the mounting frame comprising: a column and a horizontal support plate, the horizontal support plate being movably mounted on the column along the axial direction of the column, the column being mounted on the lifting mechanism, and the vibration detection mechanism being adjustablely mounted on the horizontal support plate.

[0042] According to the present invention, a home appliance abnormal sound detection system includes a vibration detection mechanism comprising: a vibration sensor, a mounting base, a slider, and a cylinder. The mounting base is provided with a slide rail, the slider is mounted on the slide rail, the vibration sensor is adjustablely mounted on the slider, the cylinder is mounted on the mounting base, the piston rod of the cylinder is connected to the slider, and the mounting base is adjustablely mounted on the transverse support plate.

[0043] According to the present invention, in a home appliance abnormal sound detection system, the piston rod of the cylinder is connected to the slider through an elastic component.

[0044] According to the present invention, a home appliance abnormal sound detection system includes a horizontal support plate comprising a first support sub-plate, a second support sub-plate, and a third support sub-plate. The first support sub-plate is movably mounted on the column along the column axis. The second support sub-plate is movably connected to one end of the first support sub-plate, and the third support sub-plate is movably connected to the other end of the first support sub-plate.

[0045] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the appliance abnormal sound detection method as described above.

[0046] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the home appliance abnormal sound detection method as described above.

[0047] The present invention provides a method, apparatus, and system for detecting abnormal sounds in household appliances. This involves acquiring at least one vibration signal from the appliance under test during operation; generating a spectrogram corresponding to each vibration signal; and inputting each spectrogram into an abnormal sound detection model to obtain the abnormal sound detection result output by the model. Since the abnormal sound detection model is trained based on sample spectrograms and corresponding abnormal sound type labels, and these sample spectrograms are generated from sample vibration signals, the model can accurately detect various abnormal sounds from the appliance under test and accurately determine the type of abnormal sound by performing image recognition on each spectrogram. Furthermore, the detection results show good consistency with manual listening detection. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of the present invention or related technologies, the drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1 This is a flowchart illustrating the method for detecting abnormal noises in home appliances provided by this invention.

[0050] Figure 2 This is a schematic diagram of the spectrogram in the abnormal sound detection method for home appliances provided by the present invention.

[0051] Figure 3 This is a schematic diagram of the structure of the appliance abnormal sound detection device provided by the present invention.

[0052] Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention.

[0053] Figure 5 This is one of the structural schematic diagrams of the home appliance abnormal sound detection system provided by the present invention.

[0054] Figure 6 This is the second schematic diagram of the home appliance abnormal sound detection system provided by the present invention.

[0055] Figure 7 This is the third schematic diagram of the structural design of the home appliance abnormal sound detection system provided by the present invention.

[0056] Figure 8 This is the fourth schematic diagram of the home appliance abnormal sound detection system provided by the present invention. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0058] It should be noted that in the description of this invention, the terms "comprising," "including," or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element. The terms "upper," "lower," etc., indicating orientation or positional relationships based on the orientation or positional relationships shown in the accompanying drawings, are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Unless otherwise expressly specified and limited, the terms "installed," "connected," and "linked" should be interpreted broadly, for example, as a fixed connection, a detachable connection, or an integral connection; a mechanical connection or an electrical connection; a direct connection or an indirect connection through an intermediate medium; or a connection within two elements. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0059] The terms "first," "second," etc., used in this invention are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more.

[0060] In related technologies, abnormal sound detection of home appliances mainly relies on manual listening. However, prolonged listening can easily cause fatigue, affecting the quality of the detection and leading to inaccurate results. In addition, manual listening is highly subjective, and human experience varies greatly. Different people may obtain different test results for the same home appliance, resulting in poor consistency and the inability to standardize the testing.

[0061] To address the aforementioned technical problems in existing related technologies, embodiments of the present invention provide a method for detecting abnormal noises from household appliances, such as... Figure 1 As shown, the procedure includes steps S110 to S130.

[0062] Step S110: Acquire at least one vibration signal to be tested during the operation of the appliance under test, wherein the appliance under test can be a common household appliance such as a refrigerator, washing machine, air conditioner, and microwave oven. During the operation of the appliance under test, a vibration sensor (e.g., an accelerometer) is used to contact the appliance to collect its vibration signal. In this step, at least one vibration signal to be tested is obtained by acquiring the vibration signal collected by the vibration sensor during the operation of the appliance under test.

[0063] For any appliance under test, there are various reasons that can cause abnormal noises. For example, in a microwave oven, issues such as an aging magnetron, motor malfunction (or failure), loose door seals and / or screws, and electromagnetic faults can all produce abnormal noises. Therefore, when there are at least one or more vibration signals to be tested, each vibration signal is collected by a vibration sensor located in the corresponding area of ​​the component causing the abnormal noise in the appliance under test. For example, a vibration sensor can be installed on the outer wall of the corresponding areas of the magnetron, motor, and door seals and / or screws in the microwave oven. Each vibration sensor simultaneously collects the vibration signals of each component, thus allowing all abnormal noises from the appliance under test to be detected at once by a subsequent abnormal noise detection model.

[0064] It is understandable that, since the test involves detecting the vibration of the appliance during operation, the appliance needs to be placed on a stationary platform to avoid interference from external vibrations and ensure that the vibration signals do not contain components of external vibrations.

[0065] It should be noted that the vibration signal to be tested can be the vibration signal of the appliance under test under different operating conditions. For example, for a microwave oven, at least one vibration signal to be tested can be collected under the three operating conditions of the microwave oven: high power, medium power and low power, so as to achieve comprehensive detection of abnormal sounds of the appliance under test under different operating conditions.

[0066] Step S120: Generate the spectrum diagrams corresponding to each of the vibration signals to be measured. For example... Figure 2As shown, the horizontal axis of the spectrogram represents time, the vertical axis represents the frequency of the vibration signal, and the value at each coordinate point is the energy of the vibration signal. Since a two-dimensional plane is used to represent three-dimensional information, the energy value is represented by the brightness of the color; the greater the brightness of the color, the stronger the vibration energy at that coordinate point. In this step, by generating the spectrogram corresponding to each vibration signal to be tested, each vibration signal to be tested is converted into an RGB color image. This facilitates the subsequent recognition of the spectrogram by an image detection-based artificial intelligence model to detect abnormal sounds from the appliance under test.

[0067] Step S130: Input each of the spectrograms to be tested into the abnormal sound detection model to obtain the abnormal sound detection result output by the abnormal sound detection model. The abnormal sound detection model is trained based on sample spectrograms and their corresponding abnormal sound type labels. The sample spectrograms are generated based on sample vibration signals. The abnormal sound detection model can be an artificial intelligence model based on Convolutional Neural Networks (CNNs). This CNN extracts features from the sample spectrograms and, leveraging the powerful capabilities of deep learning, can convert these signal features into image features for deep feature representation. After training with a large number of sample spectrograms and corresponding abnormal sound type labels, the CNN gradually learns how to map the input sample spectrograms to the corresponding abnormal sound type labels. After training, the obtained abnormal sound detection model can infer based on the spectrograms corresponding to each vibration signal to be tested, thereby detecting the abnormal sound type of the appliance under test and determining the cause of the abnormal sound. Since each vibration signal to be tested is collected based on the vibration sensor located in the corresponding area of ​​each component of the appliance to be tested that causes abnormal noise, after any component fails, the corresponding vibration signal to be tested has highly similar signal characteristics to the sample vibration signal when that component fails, and their corresponding spectrograms have highly similar image characteristics. Therefore, when the abnormal noise detection model is reasoning, it will give the detection result of the abnormal noise caused by that component.

[0068] Specifically, taking microwave ovens as an example, abnormal sounds can be produced by various factors, including magnetron aging, motor malfunction (or failure), loose door seals and / or screws, and electromagnetic faults. Before training the abnormal sound detection model, sample data and corresponding labels are created. Five types of microwave oven samples are selected: normal microwave ovens (no abnormal sounds), microwave ovens with aging magnetrons, microwave ovens with malfunctioning motors, microwave ovens with loose door seals and / or screws, and microwave ovens with electromagnetic faults. Several microwave ovens of each type are included, and the data for each type of microwave oven sample should ideally be equal. Vibration signals were collected for each type of microwave oven sample, and corresponding spectrograms were generated. Abnormal sound type labels were assigned to the five types of spectrograms. For example: a normal microwave oven's spectrogram was labeled 0; a microwave oven with an aging magnetron was labeled 1; a microwave oven with a motor malfunction was labeled 2; a microwave oven with a loose door seal and / or screws was labeled 3; and a microwave oven with an electromagnetic fault was labeled 4. A convolutional neural network was trained using these five types of spectrograms and their corresponding abnormal sound type labels. During training, different types of spectrograms were mapped to their corresponding abnormal sound type labels. After training, an abnormal sound detection model was obtained. During inference, the trained abnormal sound detection model outputs the abnormal sound type of each spectrogram as the detection result.

[0069] The appliance abnormal sound detection method of this invention acquires at least one vibration signal of the appliance under test during operation, generates a spectrogram corresponding to each vibration signal, and inputs each spectrogram into an abnormal sound detection model to obtain the abnormal sound detection result output by the model. Since the abnormal sound detection model is trained based on sample spectrograms and corresponding abnormal sound type labels, and the sample spectrograms are generated based on sample vibration signals, the model can accurately detect various abnormal sounds from the appliance under test and accurately determine the abnormal sound type by performing image recognition on each spectrogram. Furthermore, the detection results show good consistency with manual listening detection.

[0070] In some embodiments, step S120 specifically includes steps S121 and S123.

[0071] Step S121: Perform a Short-time Fourier Transform (STFT) on each of the vibration signals to be measured to convert them from time-domain signals to frequency-domain signals. Preferably, perform wavelet analysis on each frequency-domain signal to remove clutter and spikes in the corresponding waveform, thereby obtaining a more accurate frequency-domain signal.

[0072] Step S122: Determine the amplitude spectrum of each time window in the frequency domain signal corresponding to the short-time Fourier transform, that is, the vibration intensity of the frequency component corresponding to each time window in the frequency domain signal.

[0073] Step S123: For each frequency domain signal, arrange the amplitude spectra of all time windows in chronological order to generate the spectrograms to be tested. These spectrograms can be plotted using visualization tools, such as the matplotlib library in Python. For example, the plotted spectrograms are shown below. Figure 2 As shown.

[0074] In this embodiment, each vibration signal to be tested is subjected to a short-time Fourier transform to generate the corresponding spectrogram of the test signal quickly and accurately.

[0075] In some embodiments, the step of performing short-time Fourier transform on each of the vibration signals to be measured to convert each vibration signal to be measured from a time-domain signal to a frequency-domain signal specifically includes:

[0076] Each of the vibration signals to be measured is divided into multiple time windows, and windowing is applied between adjacent time windows. Specifically, firstly, each vibration signal to be measured is divided into multiple time windows (short time windows). Since the vibration signals to be measured are continuous time-domain signals, they can be segmented on the time axis according to the duration of the time windows, thus dividing each vibration signal to be measured into multiple time windows. Secondly, windowing is applied to the vibration signals between adjacent time windows, for example, using a Hanning window.

[0077] A Fast Fourier Transform (FFT) is performed on each of the windowed vibration signals to obtain the frequency components within each time window, thereby converting each vibration signal from a time-domain signal to a frequency-domain signal. Specifically, a Fast Fourier Transform is performed on the vibration signal of each windowed time window to obtain the frequency components within each time window, thus obtaining the frequency-domain signal of each vibration signal to be measured.

[0078] In this embodiment, during the short-time Fourier transform process, the vibration signal within each time window is windowed to reduce the boundary effect between frames and avoid the gap between two adjacent time windows that would cause some key signal features to be missed, thus making the final generated spectrograms more accurate.

[0079] In some embodiments, for each frequency domain signal, after arranging the amplitude spectra of all time windows in chronological order to generate each of the proposed spectrograms, the method further includes: taking the logarithm of the amplitude spectrum of each proposed spectrogram to obtain each proposed spectrogram after logarithmic coordinate transformation, wherein the logarithmic transformation can be performed with base 10 or base e. The proposed spectrogram for the input abnormal sound detection model is the proposed spectrogram after logarithmic coordinate transformation, and the sample spectrogram during abnormal sound detection model training is the sample spectrogram after logarithmic coordinate transformation.

[0080] In this embodiment, because logarithmic coordinate transformation can compress the dynamic range of data, the data details are clearer after logarithmic coordinate transformation for the spectrogram under test and various types of sample spectrograms, thus better displaying the characteristics of the vibration signal. During the training phase of the abnormal sound detection model, features can be extracted more quickly and accurately from the logarithmically transformed sample spectrograms for training, resulting in a more accurate abnormal sound detection result for the tested household appliance.

[0081] In some embodiments, the training process of the abnormal sound detection model is as follows: inputting the sample spectrogram into an initial convolutional neural network model to obtain the abnormal sound prediction result output by the initial convolutional neural network model; substituting the abnormal sound prediction result and the abnormal sound type label corresponding to the sample spectrogram into a loss function, such as the cross-entropy loss function; and training is completed when the loss function converges, thus obtaining the abnormal sound detection model.

[0082] The above training process is an iterative process. Specifically, the training process includes iteratively executing the following steps one and two.

[0083] Step 1: Input the current sample spectrogram into the initial convolutional neural network model. The initial convolutional neural network model outputs the abnormal sound prediction result of the current sample spectrogram. Substitute the abnormal sound prediction result and the abnormal sound type label corresponding to the sample spectrogram into the loss function.

[0084] Step 2: If the loss function converges, training is complete and the abnormal sound detection model is obtained; otherwise, backpropagation is performed to update the model parameters of the initial convolutional neural network model, and the next sample spectrogram is determined as the current sample spectrogram. Then, proceed to Step 1.

[0085] For some household appliances containing motors, such as washing machines and microwave ovens, motor malfunctions can manifest in various ways, producing a variety of abnormal sounds. For example, foreign objects inside the motor will contact the rotor with these objects during each rotation cycle, generating abnormal noise. Another example is improper fit between the ball bearings and the inner and outer rings of the bearings; the collision of the balls with the inner and outer rings during motor operation will also produce abnormal noise. These various abnormal motor sounds may not be clearly represented in the spectrogram, making it impossible to detect some abnormal motor sounds using the spectrogram.

[0086] To more accurately detect abnormal noises caused by the motor, in some embodiments, after step S110, the method further includes: detecting abnormal motor noises based on the vibration characteristics of the motor vibration signal under test during each rotation cycle of the motor in the household appliance. The motor vibration signal under test is obtained from a vibration sensor located in the corresponding area of ​​the motor. Specifically, each vibration sensor has a unique device ID. When acquiring the vibration signal under test from the vibration sensor in step S110, the ID of the vibration sensor is used to determine whether the vibration signal under test is the vibration signal of the motor under test.

[0087] In this embodiment, the abnormal sound detection results of the appliance under test are determined by combining the abnormal sound detection results output by the abnormal sound detection model and the abnormal sound detection results of the motor. Taking a microwave oven as an example, the abnormal sound detection model outputs the detection results of abnormal sounds caused by magnetron aging and abnormal sounds caused by electromagnetic faults. The abnormal sound of the motor is obtained by feature analysis of the vibration signal of the motor under test. Therefore, the detected abnormal sounds are: abnormal sounds caused by magnetron aging, abnormal sounds caused by electromagnetic faults, and abnormal sounds of the motor, thus making the abnormal sound detection of the appliance under test more comprehensive.

[0088] In some embodiments, the step of detecting abnormal motor noise based on the vibration characteristics of the motor vibration signal under test in each rotation cycle of the motor in the household appliance includes:

[0089] Spectral analysis is performed on the vibration signal of the motor under test within each rotation cycle to determine the target frequency band for the vibration energy abrupt change, i.e., the vibration characteristic is vibration energy. Specifically, a fast Fourier transform is performed on the vibration signal of the motor under test to obtain the frequency domain signal, and the target frequency band for the vibration energy abrupt change is determined based on the frequency domain signal. The rotation cycle T can be calculated from the rated speed of the motor using the formula T = 60 / RPM, where RPM is the rated speed of the motor, measured in revolutions per minute.

[0090] The number of vibration energy exceeding the energy threshold in the target frequency band is counted as a first target number of cycles. When the ratio of the first target number of cycles to the total number of cycles reaches a first preset ratio threshold, the abnormal noise of the motor is determined to be an abnormal noise caused by friction between the motor rotor and foreign objects inside the motor. The total number of cycles is determined based on the rotation cycle of the motor and the total duration of the vibration signal of the motor under test.

[0091] It is understandable that the total duration of the vibration signal of the motor under test is the duration of one abnormal sound detection. If there is a foreign object inside the motor, the rotor will rub against the foreign object to varying degrees with each rotation. When the friction is large, the vibration energy of the motor under test is large, which will produce abnormal sound. Therefore, in this embodiment, the number of first target cycles in which the vibration energy of the target frequency band exceeds the energy threshold is counted. The energy threshold can be set according to the actual situation, for example, 0.3G-0.6G, where G is the unit of gravitational acceleration. When the ratio of the first target cycle number to the total number of cycles reaches a first preset proportion threshold, the abnormal sound of the motor is determined to be the abnormal sound of friction between the motor rotor and the foreign object inside the motor. The first preset proportion threshold can be set according to the actual situation, for example, 15%~20%.

[0092] For example, if the motor speed is 1000 RPM and the total duration of the vibration signal of the motor under test is 2 seconds, then within these 2 seconds, the total number of motor rotation cycles is approximately 33, that is, the motor rotates 33 times. If the first preset proportional threshold is set to 18%, when the vibration energy exceeds the target number of energy threshold cycles by 6, then the abnormal sound of the motor is determined to be the abnormal sound of friction between the motor rotor and foreign objects inside the motor.

[0093] In this embodiment, the vibration energy of the target frequency band within each rotation cycle of the motor is used as the vibration characteristic. The vibration characteristic is analyzed to determine the abnormal sound caused by friction between the motor rotor and foreign objects inside the motor.

[0094] In some embodiments, abnormal motor noise is detected based on the vibration characteristics of the motor vibration signal under test in each rotation cycle of the motor in the household appliance, including:

[0095] Kujicic analysis is performed on the vibration signal of the motor under test within each rotation cycle to obtain the kujicic value within each rotation cycle, i.e., the vibration characteristic is kujicic. In this embodiment, due to improper fit between the balls inside the bearings and the inner and outer rings in the motor, the collision between the balls and the inner and outer rings during motor operation generates a large instantaneous energy. This large instantaneous energy is reflected in the vibration signal of the motor under test in the form of kujicic.

[0096] The number of second target cycles in which the kurtosis value exceeds the kurtosis threshold is counted. When the ratio of the second target cycle number to the total number of cycles reaches a second preset ratio threshold, the abnormal noise of the motor is determined to be abnormal noise of the motor bearing. The total number of cycles is determined based on the rotation cycle of the motor and the total duration of the vibration signal of the motor under test.

[0097] Both the kurtosis threshold and the second preset ratio threshold can be set according to actual conditions. The kurtosis threshold can be 8, and the second preset ratio threshold can be 10%~15%. For example, if the motor speed is 1000 RPM and the total duration of the vibration signal of the motor under test is 2 seconds, then the total number of motor rotation cycles within these 2 seconds is approximately 33, that is, the motor rotates 33 times. If the second preset ratio threshold is set to 10%, the number of second target cycles with a kurtosis value exceeding 8 is 4, that is, the ratio of the second target cycle number to the total number of cycles is 12%, which is greater than 10%. Therefore, abnormal noise from the motor bearing is determined.

[0098] In this embodiment, the kurtosis of the vibration signal within each rotation cycle of the motor is used as the vibration characteristic. The kurtosis is analyzed to determine the abnormal noise of the motor bearing, that is, the abnormal noise caused by improper fit between the balls and the inner and outer rings in the motor bearing.

[0099] The appliance abnormal sound detection device provided by the present invention is described below. The appliance abnormal sound detection device described below can be referred to in correspondence with the appliance abnormal sound detection method described above.

[0100] The appliance abnormal sound detection device of this invention, such as Figure 3 As shown, it includes the following functional modules 310 to 330.

[0101] The vibration signal acquisition module 310 is used to acquire at least one vibration signal to be tested when the household appliance is running.

[0102] The spectrogram generation module 320 is used to generate spectrograms corresponding to each of the vibration signals to be tested.

[0103] The model execution module 330 is used to input each of the spectrograms to be tested into the abnormal sound detection model to obtain the abnormal sound detection result output by the abnormal sound detection model.

[0104] The abnormal sound detection model is trained based on the sample spectrogram and the abnormal sound type label corresponding to the sample spectrogram, and the sample spectrogram is generated based on the sample vibration signal.

[0105] In this embodiment of the appliance abnormal sound detection device, at least one vibration signal to be tested is acquired during the operation of the appliance under test; a spectrogram to be tested corresponding to each vibration signal is generated; and each spectrogram to be tested is input into the abnormal sound detection model to obtain the abnormal sound detection result output by the abnormal sound detection model. Since the abnormal sound detection model is trained based on sample spectrograms and the abnormal sound type labels corresponding to the sample spectrograms, and the sample spectrograms are generated based on sample vibration signals, the abnormal sound detection model can accurately detect various abnormal sounds of the appliance under test and accurately determine the abnormal sound type by performing image recognition on each spectrogram to be tested. Furthermore, the detection results show good consistency with manual listening detection.

[0106] In some embodiments, the spectrogram generation module 320 includes the following modules.

[0107] The signal conversion module is used to perform short-time Fourier transform on each of the vibration signals to be measured, so as to convert each of the vibration signals to be measured from a time domain signal to a frequency domain signal.

[0108] The amplitude spectrum determination module is used to determine the amplitude spectrum of each time window corresponding to the short-time Fourier transform of each frequency domain signal.

[0109] The amplitude spectrum arrangement module is used to arrange the amplitude spectra of all time windows in chronological order for each frequency domain signal to generate the spectrum diagram of each signal to be tested.

[0110] In some embodiments, the signal conversion module is specifically used to divide each of the vibration signals to be tested into multiple time windows, and to perform windowing processing between two adjacent time windows; to perform fast Fourier transform on each of the vibration signals to be tested after windowing processing to obtain the frequency components within each time window, so as to convert each of the vibration signals to be tested from time domain signals to frequency domain signals.

[0111] In some embodiments, the appliance abnormal sound detection device further includes: a logarithmic coordinate transformation module, used to, for each frequency domain signal, arrange the amplitude spectra of all time windows in chronological order to generate each of the test spectrograms, and then take the logarithm of the amplitude spectrum of each of the test spectrograms to obtain each test spectrogram after logarithmic coordinate transformation. The test spectrogram input to the abnormal sound detection model is the test spectrogram after logarithmic coordinate transformation, and the sample spectrogram during abnormal sound detection model training is the sample spectrogram after logarithmic coordinate transformation.

[0112] In some embodiments, the training process of the abnormal sound detection model is as follows: inputting the sample spectrogram into an initial convolutional neural network model to obtain the abnormal sound prediction result output by the initial convolutional neural network model; substituting the abnormal sound prediction result and the abnormal sound type label corresponding to the sample spectrogram into the loss function; and when the loss function converges, the training is completed, and the abnormal sound detection model is obtained.

[0113] In some embodiments, the appliance abnormal noise detection device further includes: a motor abnormal noise detection module, used to detect motor abnormal noise based on the vibration characteristics of the motor vibration signal of the motor under test in each rotation cycle after acquiring at least one vibration signal under test during the operation of the appliance under test, wherein the vibration signal of the motor under test is acquired based on a vibration sensor located in the corresponding area of ​​the motor.

[0114] In some embodiments, the motor abnormal noise detection module is specifically used to perform spectrum analysis on the vibration signal of the motor under test within each rotation cycle to determine the target frequency band of the vibration energy change; count the number of first target cycles in which the vibration energy of the target frequency band exceeds the energy threshold; and determine that the abnormal motor noise is an abnormal noise caused by friction between the motor rotor and foreign objects inside the motor when the ratio of the first target cycle to the total number of cycles reaches a first preset ratio threshold. The total number of cycles is determined based on the rotation cycle of the motor and the total duration of the vibration signal of the motor under test.

[0115] In some embodiments, the motor abnormal noise detection module is specifically used to perform kurtosis analysis on the vibration signal of the motor under test in each rotation cycle to obtain the kurtosis value in each rotation cycle; count the number of second target cycles in which the kurtosis value exceeds the kurtosis threshold; and determine the motor abnormal noise as motor bearing abnormal noise when the ratio of the second target cycle to the total number of cycles reaches a second preset ratio threshold. The total number of cycles is determined based on the rotation cycle of the motor and the total duration of the vibration signal of the motor under test.

[0116] Figure 4 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 4 As shown, the electronic device may include a processor 410, a communications interface 420, a memory 430, and a communication bus 440. The processor 410, communications interface 420, and memory 430 communicate with each other via the communication bus 440. The processor 410 can call logical instructions from the memory 430 to execute a method for detecting abnormal sounds from household appliances. This method includes the following steps.

[0117] Acquire at least one vibration signal to be tested when the appliance is in operation.

[0118] Generate the spectrum diagrams corresponding to each of the vibration signals to be measured.

[0119] Each of the spectrograms to be tested is input into the abnormal sound detection model to obtain the abnormal sound detection result output by the abnormal sound detection model.

[0120] The abnormal sound detection model is trained based on the sample spectrogram and the abnormal sound type label corresponding to the sample spectrogram, and the sample spectrogram is generated based on the sample vibration signal.

[0121] Furthermore, the logical instructions in the aforementioned memory 430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0122] This invention also provides a home appliance abnormal sound detection system, such as... Figure 5 , Figure 6 and Figure 7 As shown, it includes: production line body 1, lifting mechanism 2, vibration detection mechanism 3, and the aforementioned electronic equipment 4.

[0123] The production line 1 is equipped with a fixture plate 5 for carrying the appliance to be tested (e.g., microwave oven 6). During testing, the appliance to be tested is placed on the fixture plate 5, and the production line 1 is used to transport the fixture plate 5. Specifically, the production line 1 can be (but is not limited to) the following structure: the production line 1 includes a first support 24 and a second support 25, which are arranged opposite each other. The top of each of the first support 24 and the second support 25 is provided with a conveyor belt or conveyor roller (not shown in the figure) with the same conveying speed. The two ends of the fixture plate 5 are placed on the conveyor belt or conveyor roller of the first support 24 and the second support 25 to realize the transport of the fixture plate 5. Moreover, there is a gap between the two supports so that the lifting mechanism 2 can lift the fixture plate 5 from between the two supports.

[0124] The lifting mechanism 2 is located below the production line 1 without contact with it, and is used to lift the tooling plate 5, raising the tooling plate 5 containing the appliance to be tested until it is detached from the production line 1. Specifically, the lifting mechanism 2 can be (but is not limited to) the following structure: the lifting mechanism 2 includes a support frame 7 and a lifting component mounted on the support frame 7, the lifting component being used to lift the tooling plate 5. The lifting component can be a cylinder lifting structure or a motor lifting structure. For the cylinder lifting structure, it includes a cylinder 8 and a lifting plate 9 connected to the piston rod end of the cylinder 8. The lifting plate 9 lifts the tooling plate 5 by extending and retracting the piston rod of the cylinder 8. Furthermore, in order to prevent the vibration of the production line 1 from being transmitted from the ground to the support frame 7, and then from the support frame 7 upwards to the appliance to be tested, a vibration damping foot cup 10 is provided at the bottom of the support frame 7 to further avoid the influence of external vibration on the vibration signal to be tested.

[0125] The vibration detection mechanism 3 is located above the lifting mechanism 2 and is used to transmit at least one vibration signal of the household appliance to be tested to the electronic device 4. After receiving at least one vibration signal, the electronic device 4 executes the household appliance abnormal sound detection method of the above embodiment to detect abnormal sounds of the household appliance to be tested.

[0126] It should be noted that in this embodiment, the electronic device 4 can be an industrial control computer, which can not only execute the abnormal sound detection method for home appliances described in the above embodiment, but also integrate the control programs for the production line 1, the lifting mechanism 2, and the vibration detection mechanism 3 into the electronic device 4. For example, the electronic device 4 can control the start and stop of the production line 1. When the tooling plate 5 is conveyed to the top of the lifting mechanism 2, the electronic device 4 controls the lifting mechanism 2 to lift the tooling plate 5 upward. After the tooling plate 5 is lifted into place, the electronic device 4 controls the vibration detection mechanism 3 to collect at least one vibration signal of the home appliance to be tested on the tooling plate 5 and acquires at least one vibration signal to be tested collected by the vibration detection mechanism 3. After the vibration signal is collected, the electronic device 4 controls the lifting mechanism 2 to fall back, and the tooling plate 5 returns to the production line 1 and is conveyed to the next process.

[0127] Of course, the assembly line 1 is also equipped with a socket 13 to provide power to the appliance under test. The appliance under test is connected to the socket 13 by a flexible wire. The vibration of the assembly line 1 will hardly be transmitted to the appliance under test through the flexible wire. Preferably, the socket 13 can be set on the tooling plate 5.

[0128] In the appliance abnormal sound detection system of this embodiment, since the lifting mechanism 2 itself does not contact the production line 1, and the tooling plate 5 is lifted and detached from the production line 1 to form a static bearing platform, the vibration of the production line 1 will not be transmitted to the appliance under test, thus avoiding interference from the vibration of the production line 1 with the vibration signal under test, making the acquisition of at least one vibration signal under test more accurate, and the detection of abnormal sounds more accurate.

[0129] The appliance abnormal sound detection system of this embodiment can be used in both appliance manufacturing and repair scenarios. In the appliance manufacturing scenario, assembly line 1 can be connected to the end of the production line. The fixture plate 5 can be shared by assembly line 1 and the production line. Each appliance coming off the production line is directly moved onto assembly line 1 via fixture plate 5, thus enabling abnormal sound detection of all newly manufactured appliances and preventing defective products from entering the market. In the repair scenario, appliances to be repaired are placed sequentially on fixture plate 5 of assembly line 1 for sequential testing. Compared to traditional manual sound detection, this improves the efficiency of repair testing.

[0130] In some embodiments, the appliance abnormal sound detection system further includes a mounting bracket comprising a column 11 and a horizontal support plate 12. The horizontal support plate 12 is axially movable (i.e., vertically movable) on the column 11, which is mounted on the lifting mechanism 2. The vibration detection mechanism 3 is adjustablely mounted on the horizontal support plate 12, such that the vibration detection mechanism 3 contacts the area corresponding to the component in the appliance under test that causes the abnormal sound. The horizontal support plate 12 moves vertically along the column 11 to adjust the height of the vibration detection mechanism 3 according to the height of the appliance under test after lifting. The vibration detection mechanism 3 is adjustablely mounted on the horizontal support plate 12, allowing the installation position of the vibration detection mechanism 3 to be adjusted according to the size of the appliance under test or according to the corresponding areas of different components in different appliances under test that cause abnormal sounds. Specifically, the horizontal support plate 12 is provided with multiple mounting holes 23, which can be randomly distributed so that the distance between each mounting hole 23 and the appliance under test and the projection position on the appliance under test are not the same. By installing the vibration detection mechanism 3 in different mounting holes 23, the installation position of the vibration detection mechanism 3 can be adjusted.

[0131] The column 11 is installed on the lifting mechanism 2. The lifting mechanism 2 does not contact the production line 1. Therefore, the column 11, the horizontal support plate 12 and the vibration detection mechanism 3 do not contact the production line 1, thus avoiding interference from the vibration of the production line 1 to the vibration detection mechanism 3, and making the collected vibration signal more accurate.

[0132] In some embodiments, such as Figure 8As shown, the vibration detection mechanism 3 includes: a vibration sensor 14, a mounting base 15, a slider 16, and a cylinder 17. The mounting base 15 is provided with a slide rail 18, and the slider 16 is mounted on the slide rail 18, meaning the slider 16 can slide on the slide rail 18. The vibration sensor 14 is adjustablely mounted on the slider 16. The cylinder 17 is mounted on the mounting base 15, and the piston rod of the cylinder 17 is connected to the slider 16. The mounting base 15 is adjustablely mounted on the transverse support plate 12. Specifically, the bottom of the mounting base 15 also has multiple countersunk holes that mate with the mounting holes 23. The vibration sensor 14 is adjustablely mounted on the slider 16, further adjusting the position of the vibration sensor 14 relative to the appliance being tested, such as... Figure 8 The installation height of the vibration sensor 14 on the slider 16 can be adjusted, as can its longitudinal position. This, combined with the adjustment of the lateral or horizontal position of the vibration detection mechanism 3 on the transverse support plate 12, allows the vibration sensor 14 to more accurately locate the corresponding areas of the components in the appliance under test that cause abnormal noise. The cylinder 17 pushes the slider 16 to move, fine-tuning the distance between the vibration sensor 14 and the appliance under test to maintain contact between them.

[0133] In some embodiments, the piston rod of the cylinder 17 is connected to the slider 16 via an elastic component, such as a spring 19, to prevent the vibration sensor 14 from pressing too tightly against the appliance under test, which could damage the outer surface of the appliance. Furthermore, if the appliance is pressed too tightly, significant vibrations could cause deformation of the outer surface, and the stress from this deformation would also act on the vibration sensor 14, leading to the erroneous acquisition of this stress as part of the vibration signal. Moreover, deformation could cause the vibration sensor 14 to lose contact with the outer surface of the appliance, preventing the acquisition of a vibration signal. Therefore, in this embodiment, the piston rod of the cylinder 17 is connected to the slider 16 via an elastic component, which buffers the vibration, preventing the aforementioned erroneous or unacceptable acquisition of vibration signals.

[0134] In some embodiments, the transverse support plate 12 includes a first support sub-plate 20, a second support sub-plate 21, and a third support sub-plate 22. The first support sub-plate 20 is movably mounted on the column 11 along the column axis. The second support sub-plate 21 is movably connected to one end of the first support sub-plate 20, and the third support sub-plate 22 is movably connected to the other end of the first support sub-plate 20. The movable connection can be a sliding connection or a rotating connection, which can be adjusted according to the shape and size of the appliance under test. For example, if the movable connection is a rotating connection, for a cuboid-shaped appliance under test, the second support sub-plate 21 and the third support sub-plate 22 can be adjusted to be parallel and perpendicular to the first support sub-plate 20, allowing vibration signals to be collected from at least three sides. For smaller cylindrical appliances under test, such as blenders, the angles between the second support sub-plate 21 and the third support sub-plate 22 and the first support sub-plate 20 can be adjusted to be less than 90 degrees, forming an enclosing structure for the blender, allowing vibration signals to be collected from multiple positions on the sides.

[0135] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the home appliance abnormal sound detection method provided by the above methods, which includes the following steps.

[0136] Acquire at least one vibration signal to be tested when the appliance is in operation.

[0137] Generate the spectrum diagrams corresponding to each of the vibration signals to be measured.

[0138] Each of the spectrograms to be tested is input into the abnormal sound detection model to obtain the abnormal sound detection result output by the abnormal sound detection model.

[0139] The abnormal sound detection model is trained based on the sample spectrogram and the abnormal sound type label corresponding to the sample spectrogram, and the sample spectrogram is generated based on the sample vibration signal.

[0140] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to perform the appliance abnormal sound detection method provided by the above methods, the method comprising the following steps.

[0141] Acquire at least one vibration signal to be tested when the appliance is in operation.

[0142] Generate the spectrum diagrams corresponding to each of the vibration signals to be measured.

[0143] Each of the spectrograms to be tested is input into the abnormal sound detection model to obtain the abnormal sound detection result output by the abnormal sound detection model.

[0144] The abnormal sound detection model is trained based on the sample spectrogram and the abnormal sound type label corresponding to the sample spectrogram, and the sample spectrogram is generated based on the sample vibration signal.

[0145] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0146] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0147] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method of detecting an abnormal sound of a home appliance, characterized by, The method comprises: obtaining at least one to-be-detected vibration signal of a household appliance in operation; generating a to-be-detected spectrogram corresponding to each to-be-detected vibration signal, and taking a logarithm of an amplitude spectrum of each to-be-detected spectrogram to obtain each to-be-detected spectrogram after logarithmic coordinate transformation; inputting each to-be-detected spectrogram after logarithmic coordinate transformation into an abnormal sound detection model to obtain an abnormal sound detection result output by the abnormal sound detection model; wherein the abnormal sound detection model is trained based on a sample spectrogram and an abnormal sound type label corresponding to the sample spectrogram, and the sample spectrogram is generated based on a sample vibration signal; during training of the abnormal sound detection model, the sample spectrogram is a sample spectrogram after logarithmic coordinate transformation; after obtaining the at least one to-be-detected vibration signal of the household appliance in operation, the method further comprises: detecting motor abnormal sound based on vibration characteristics of a to-be-detected motor vibration signal of a motor in the household appliance in each rotation period, wherein the to-be-detected motor vibration signal is collected based on a vibration sensor located in a region corresponding to the motor; wherein detecting motor abnormal sound based on vibration characteristics of a to-be-detected motor vibration signal of a motor in the household appliance in each rotation period comprises: performing frequency spectrum analysis on the to-be-detected motor vibration signal in each rotation period to determine a target frequency band with sudden change in vibration energy; counting a first target period number in which vibration energy of the target frequency band exceeds an energy threshold value, and determining that the motor abnormal sound is abnormal sound generated by friction between a motor rotor and foreign matter in the motor when a proportion of the first target period number to a total period number reaches a first preset proportion threshold value, wherein the total period number is determined based on a rotation period of the motor and a total time length of the to-be-detected motor vibration signal; or, detecting motor abnormal sound based on vibration characteristics of a to-be-detected motor vibration signal of a motor in the household appliance in each rotation period comprises: performing kurtosis analysis on the to-be-detected motor vibration signal in each rotation period to obtain a kurtosis value in each rotation period; counting a second target period number in which the kurtosis value exceeds a kurtosis threshold value, and determining that the motor abnormal sound is motor bearing abnormal sound when a proportion of the second target period number to a total period number reaches a second preset proportion threshold value, wherein the total period number is determined based on a rotation period of the motor and a total time length of the to-be-detected motor vibration signal.

2. The abnormal sound detection method of the home appliance according to claim 1, characterized in that, The method comprises: dividing each to-be-detected vibration signal into a plurality of time windows, and performing windowing processing between adjacent two time windows; ​ ​ 3. The abnormal sound detection method of claim 2, wherein, ​ ​ The windowed vibration signals are respectively subjected to fast Fourier transform to obtain frequency components in each time window, so as to convert the vibration signals into frequency domain signals.

4. The abnormal sound detection method of the home appliance according to claim 1, characterized in that, The abnormal sound detection model training process is as follows: The sample spectrogram is input into an initial convolutional neural network model to obtain an abnormal sound prediction result output by the initial convolutional neural network model. The abnormal sound prediction result and an abnormal sound type label corresponding to the sample spectrogram are substituted into a loss function, and the training is completed under the condition that the loss function converges, so as to obtain the abnormal sound detection model.

5. An abnormal sound detection device for a home appliance, characterized by comprising: Comprise: a vibration signal acquisition module configured to acquire at least one vibration signal to be measured when a household appliance is running; a spectrogram generation module configured to generate a spectrogram corresponding to each of the vibration signals to be measured, and to obtain each spectrogram after logarithmic coordinate transformation by taking the logarithm of the amplitude spectrum of each spectrogram; a model execution module configured to input each spectrogram after logarithmic coordinate transformation into an abnormal sound detection model to obtain an abnormal sound detection result output by the abnormal sound detection model; The abnormal sound detection model is trained based on a sample spectrogram and an abnormal sound type label corresponding to the sample spectrogram, and the sample spectrogram is generated based on a sample vibration signal; during the training of the abnormal sound detection model, the sample spectrogram is a sample spectrogram after logarithmic coordinate transformation; Further comprising: a motor abnormal sound detection module configured to detect motor abnormal sound based on the vibration characteristics of a motor vibration signal to be measured of a motor in each rotation period of the household appliance after acquiring at least one vibration signal to be measured when the household appliance is running, wherein the motor vibration signal to be measured is acquired by a vibration sensor located in a region corresponding to the motor; The motor abnormal sound detection module is specifically configured to perform frequency spectrum analysis on the motor vibration signal to be measured in each rotation period to determine a target frequency band with sudden changes in vibration energy; count a first target period number in which the vibration energy of the target frequency band exceeds an energy threshold; and determine that the motor abnormal sound is an abnormal sound caused by friction between a motor rotor and foreign matter in the motor when the proportion of the first target period number to a total period number reaches a first preset proportion threshold, wherein the total period number is determined based on the rotation period of the motor and the total time length of the motor vibration signal to be measured. Alternatively, the motor abnormal sound detection module is specifically configured to perform kurtosis analysis on the motor vibration signal to be measured in each rotation period to obtain a kurtosis value in each rotation period; count a second target period number in which the kurtosis value exceeds a kurtosis threshold; and determine that the motor abnormal sound is a motor bearing abnormal sound when the proportion of the second target period number to a total period number reaches a second preset proportion threshold, wherein the total period number is determined based on the rotation period of the motor and the total time length of the motor vibration signal to be measured.

6. An electronic device comprising a memory, a processor, and a computer program stored on the memory and running on the processor, characterized in that, The processor executes the computer program to implement the household appliance abnormal sound detection method of any one of claims 1 to 4.

7. An abnormal sound detection system for a home appliance, characterized by comprising: Comprise: a pipeline body, a jacking mechanism, a vibration detection mechanism, and the electronic device of claim 6; The pipeline body is provided with a tool plate for carrying the household appliance to be tested; The lifting mechanism is located below the pipeline body in a non-contact manner, and is used for lifting the tool plate; The vibration detection mechanism is located above the lifting mechanism, and is used for generating at least one vibration signal of the household appliance to be tested to the electronic device.

8. The abnormal sound detection system of claim 7, wherein Further comprising a mounting frame, the mounting frame comprises: a column and a transverse support plate, the transverse support plate is axially movably mounted on the column, the column is mounted on the lifting mechanism, and the vibration detection mechanism is positionally adjustably mounted on the transverse support plate.

9. The abnormal sound detection system of claim 8, wherein The vibration detection mechanism comprises: a vibration sensor, a mounting seat, a slider and a cylinder, the mounting seat is provided with a sliding rail, the slider is mounted on the sliding rail, the vibration sensor is positionally adjustably mounted on the slider, the cylinder is mounted on the mounting seat, the piston rod of the cylinder is connected with the slider, and the mounting seat is positionally adjustably mounted on the transverse support plate.

10. The abnormal sound detection system of claim 9, wherein The piston rod of the cylinder is connected with the slider through an elastic component.

11. The abnormal sound detection system of claim 9, wherein The transverse support plate comprises: a first support subplate, a second support subplate and a third support subplate, the first support subplate is axially movably mounted on the column, the second support subplate is movably connected with one end of the first support subplate, and the third support subplate is movably connected with the other end of the first support subplate.

12. A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to realize the household appliance abnormal sound detection method according to any one of claims 1 to 4.

13. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to realize the household appliance abnormal sound detection method according to any one of claims 1 to 4.

Images