An online noise monitoring method and system for a multi-blade centrifugal fan

Abstract

The present invention discloses a method and system for on-line monitoring of the noise of a multi-blade centrifugal fan, which relates to the technical field of noise signal data processing. The method includes: obtaining the noise time-domain signal of the centrifugal fan; inputting the noise time-domain signal into a noise separation model to obtain a separated noise time-domain signal; calculating the first sound pressure level respectively according to the separated noise time-domain signal; performing discrete Fourier transform on the noise time-domain signal to obtain a noise frequency-domain signal; calculating the second sound pressure level according to the noise frequency-domain signal; and adjusting the centrifugal fan according to the first sound pressure level and the second sound pressure level. By obtaining the time-domain signal of the centrifugal fan noise, separating the time-domain signals of the volute and mechanical noises from the time-domain signal to calculate their first sound pressure levels respectively, converting the time-domain signal into a frequency-domain signal to calculate the second sound pressure level for real-time noise monitoring, and adjusting the centrifugal fan according to the monitoring results, the present invention reduces the noise of the centrifugal fan and improves the working efficiency of the centrifugal fan.

Description

Technical Field

[0001] The present invention relates to the technical field of noise signal data processing, and particularly relates to a method and system for on-line monitoring of the noise of a multi-vane centrifugal fan. Background Art

[0002] A centrifugal fan is a widely used mechanical device that converts kinetic energy into static pressure energy by rotating an impeller to generate an air flow for transporting gas. Centrifugal fans are widely used in various commercial and residential environments due to their high efficiency, reliability, and versatility.

[0003] In scenarios that require a quiet environment, such as hospitals, laboratories, or bedrooms, systems equipped with centrifugal fans, such as ventilation systems or air conditioning systems, need to pay attention to noise problems. Therefore, it is necessary to conduct low-noise design during the design of centrifugal fans and monitor the noise of centrifugal fans during use. Existing noise monitoring of centrifugal fans starts from the perspective of protecting safety, by obtaining the magnitude of the sound emitted by the centrifugal fan and comparing it with a reference value for hearing damage. When the noise exceeds the reference value, the centrifugal fan is inspected and maintained. This monitoring method ignores the sensitivity of humans to noise, resulting in high noise of the centrifugal fan and the inability to detect problems with the centrifugal fan in a timely manner, leading to low working efficiency of the centrifugal fan. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and system for on-line monitoring of the noise of a multi-vane centrifugal fan. The present invention obtains the time-domain signal of the centrifugal fan noise, separates the time-domain signals of the volute noise and mechanical noise from the time-domain signal, calculates their first sound pressure levels respectively, converts the time-domain signal into a frequency-domain signal to calculate the second sound pressure level of the total noise for real-time monitoring of the noise, and adjusts the centrifugal fan according to the first and second sound pressure levels obtained by the monitoring calculation, reducing the noise of the centrifugal fan and improving the working efficiency of the centrifugal fan.

[0005] The purpose of the present invention is achieved by the following technical means:

[0006] In the first aspect, the present invention provides a method for on-line monitoring of the noise of a multi-vane centrifugal fan, including:

[0007] Obtain the time-domain signal of the centrifugal fan noise;

[0008] Input the time-domain signal of the noise into a noise separation model to obtain a separated noise time-domain signal;

[0009] Calculate the first sound pressure level respectively according to the separated noise time-domain signal;

[0010] Perform a discrete Fourier transform on the time-domain signal of the noise to obtain a noise frequency-domain signal;

[0011] Calculate the second sound pressure level according to the noise frequency domain signal;

[0012] Adjust the centrifugal fan according to the first sound pressure level and the second sound pressure level;

[0013] Among them, the separated noise time domain signal includes a volute noise time domain signal and a mechanical noise time domain signal.

[0014] Preferably, before inputting the noise time domain signal into the noise separation model to obtain the separated noise time domain signal, it further includes:

[0015] Obtain a plurality of simulated volute noises and simulated mechanical noises;

[0016] Randomly mix the simulated volute noise and the simulated mechanical noise to obtain an original data set;

[0017] Randomly sample the original data set with a fixed time window, perform gain summation on the sampled noise data to form new noise data, and obtain a training sample set;

[0018] Train the noise separation model through the training sample set.

[0019] Preferably, the training of the noise separation model through the training sample set includes:

[0020] Perform short-time Fourier transform on the noise data of the training sample set to obtain a characteristic signal;

[0021] Perform block mask calculation on the characteristic signal to obtain a separated signal spectrum;

[0022] Perform inverse short-time Fourier transform on the separated signal spectrum to obtain the separated noise time domain signal;

[0023] Evaluate the separated noise time domain signal and correct the parameter weights of the noise separation model.

[0024] Preferably, the evaluation of the separated noise time domain signal and the correction of the parameter weights of the noise separation model include:

[0025] Perform short-time Fourier transform on the simulated volute noise and the simulated mechanical noise in the training sample set to obtain a true signal spectrum;

[0026] Construct a loss function according to the separated signal spectrum and the true signal spectrum to evaluate the separated noise time domain signal;

[0027] The calculation formula of the loss function is as follows:

[0028] ;

[0029] Among them, is the loss function, is the true signal spectrum, is the separated signal spectrum.

[0030] Preferably, calculating the first sound pressure level according to the separated noise time-domain signal includes:

[0031] Calculating the effective sound pressure of the separated noise time-domain according to the separated noise time-domain signal;

[0032] Calculating the first sound pressure level according to the effective sound pressure of the separated noise time-domain;

[0033] The calculation formula of the effective sound pressure of the separated noise time-domain is as follows:

[0034] ;

[0035] Among them, is the effective sound pressure of the separated noise time-domain, is the number of samples of the noise time-domain signal, is the th sound pressure at the sampling moment;

[0036] The calculation formula of the first sound pressure level is as follows:

[0037] ;

[0038] Among them, is the first sound pressure level, is the effective sound pressure of the separated noise time-domain, is the atmospheric reference sound pressure.

[0039] Preferably, calculating the second sound pressure level according to the noise frequency-domain signal includes:

[0040] Calculating the noise frequency-domain amplitude according to the noise frequency-domain signal;

[0041] Calculating the effective sound pressure of the noise frequency-domain according to the noise frequency-domain amplitude;

[0042] Calculating the second sound pressure level according to the effective sound pressure of the noise frequency-domain;

[0043] The calculation formula of the noise frequency-domain amplitude is as follows:

[0044] ;

[0045] Among them, is the noise frequency-domain amplitude, is the noise frequency-domain signal, is the weighting operation symbol, is the frequency point serial number, is the sampling rate of the noise time-domain signal, is the number of samples of the noise time-domain signal;

[0046] The calculation formula of the effective sound pressure in the noise frequency domain is as follows:

[0047] ;

[0048] where, is the effective sound pressure in the noise frequency domain, is the number of samples of the noise time-domain signal, is the amplitude in the noise frequency domain;

[0049] The calculation formula of the second sound pressure level is as follows:

[0050] ;

[0051] where, is the second sound pressure level, is the effective sound pressure in the noise frequency domain, is the atmospheric reference sound pressure.

[0052] Preferably, adjusting the centrifugal fan according to the first sound pressure level and the second sound pressure level includes:

[0053] Comparing the first sound pressure levels to obtain the maximum first sound pressure level;

[0054] Comparing the maximum first sound pressure level with the working threshold of the noise time-domain signal corresponding thereto. When the maximum first sound pressure level is greater than the working threshold, the centrifugal fan is shut down and an abnormal alarm is given;

[0055] When the maximum first sound pressure level is less than the working threshold, comparing the second sound pressure level with the sensitivity threshold. When the second sound pressure level is greater than the sensitivity threshold, the rotational speed of the centrifugal fan is reduced.

[0056] In a second aspect, the present invention provides a noise on-line monitoring system for a multi-blade centrifugal fan, which is applied to the above-mentioned noise on-line monitoring method for a multi-blade centrifugal fan, and includes:

[0057] A time-domain signal acquisition module for acquiring the noise time-domain signal of the centrifugal fan;

[0058] A time-domain signal separation module for inputting the noise time-domain signal into a noise separation model to obtain a separated noise time-domain signal;

[0059] A first sound pressure level calculation module for respectively calculating the first sound pressure level according to the separated noise time-domain signal

[0060] A frequency-domain signal acquisition module, configured to perform a discrete Fourier transform on the noise time-domain signal to obtain a noise frequency-domain signal;

[0061] A second sound pressure level calculation module, configured to calculate a second sound pressure level according to the noise frequency-domain signal;

[0062] A centrifugal fan adjustment module, configured to adjust the centrifugal fan according to the first sound pressure level and the second sound pressure level;

[0063] Wherein, the separated noise time-domain signal includes a volute noise time-domain signal and a mechanical noise time-domain signal.

[0064] In a third aspect, the present invention provides an electronic device, including a processor and a memory, the memory is configured to store computer program code, the computer program code includes computer instructions, when the processor executes the computer instructions, the electronic device executes the above-mentioned online noise monitoring method for a multi-blade centrifugal fan.

[0065] In a fourth aspect, the present invention provides a computer-readable storage medium, in which a computer program is stored, the computer program includes program instructions, when the program instructions are executed by a processor of an electronic device, the processor is caused to execute the above-mentioned online noise monitoring method for a multi-blade centrifugal fan.

[0066] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0067] The present invention obtains the time-domain signal of the centrifugal fan noise, separates the time-domain signals of the volute noise and the mechanical noise from the time-domain signal, calculates their first sound pressure levels respectively, converts the time-domain signal into a frequency-domain signal to calculate the second sound pressure level of the total noise for real-time noise monitoring, and adjusts the centrifugal fan according to the first sound pressure level and the second sound pressure level obtained by the monitoring calculation, reducing the noise of the centrifugal fan and improving the working efficiency of the centrifugal fan;

[0068] The present invention mixes and randomly samples multiple simulated volute noises and multiple simulated mechanical noises, performs gain summation on the sampled noise data to form a training sample set, and trains a noise separation model through the training sample set, improving the noise separation performance of the noise separation model;

[0069] The present invention constructs a loss function to measure the similarity between the separated signal spectrum and the real signal spectrum, and thereby corrects the parameter weights of the noise separation model, improving the noise separation performance of the noise separation model;

[0070] The present invention calculates the sound pressure levels of two kinds of noises based on the time-domain signals of the volute noise and the mechanical noise, providing a data basis for the subsequent adjustment of the centrifugal fan;

[0071] The present invention performs weighted correction on the noise frequency-domain signals, removes the signal frequencies that humans are not sensitive to, and calculates the sound pressure level of the total noise of the centrifugal fan, providing a data basis for the subsequent adjustment of the centrifugal fan;

[0072] The present invention adjusts the working state of the centrifugal fan by combining the first sound pressure level and the second sound pressure level. The first sound pressure level is used to monitor whether there is an abnormality in the centrifugal fan. On this basis, considering the sensitivity of humans to noise signals, the centrifugal fan is adjusted, reducing the noise of the centrifugal fan and improving the working efficiency of the centrifugal fan. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] The drawings herein are incorporated into the specification and form a part of the specification, indicating the embodiments consistent with the present invention and used together with the specification to explain the principles of the present invention.

[0074] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0075] Figure 1 It is a schematic flow chart of a method for on-line monitoring of the noise of a multi-blade centrifugal fan provided in this embodiment;

[0076] Figure 2 It is a schematic flow chart of step S20 provided in this embodiment;

[0077] Figure 3 It is a schematic flow chart of step S14 provided in this embodiment;

[0078] Figure 4 It is a schematic flow chart of step S144 provided in this embodiment;

[0079] Figure 5 It is a schematic flow chart of step S30 provided in this embodiment;

[0080] Figure 6 It is a schematic flow chart of step S50 provided in this embodiment;

[0081] Figure 7 It is a schematic flow chart of step S60 provided in this embodiment;

[0082] Figure 8Schematic diagram of the structure of an on-line noise monitoring system for a multi-blade centrifugal fan provided in this embodiment;

[0083] Figure 9 Schematic diagram of the structure of an electronic device provided in this embodiment. Detailed implementation manners

[0084] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0085] It should be noted that all directional indications (such as up, down, left, right, front, back...) in the embodiments of the present invention are only used to explain the relative positional relationship and movement conditions between components in a specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indications will also change accordingly.

[0086] In addition, the descriptions involving "first", "second", etc. in the present invention are only for descriptive purposes, and cannot be understood as indicating or implying their relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one such feature. In addition, the technical solutions between various embodiments can be combined with each other, but it must be based on the fact that those of ordinary skill in the art can implement them. When the combination of technical solutions appears to be contradictory or unable to be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the protection scope required by the present invention.

[0087] The present invention provides an on-line noise monitoring method for a multi-blade centrifugal fan, as Figure 1 shown, including the following steps:

[0088] S10. Obtain the noise time-domain signal of the centrifugal fan;

[0089] S20. Input the noise time-domain signal into the noise separation model to obtain the separated noise time-domain signal;

[0090] S30. Calculate the first sound pressure level respectively according to the separated noise time-domain signal;

[0091] S40. Perform discrete Fourier transform on the noise time-domain signal to obtain the noise frequency-domain signal;

[0092] S50. Calculate the second sound pressure level according to the noise frequency-domain signal;

[0093] S60, adjust the centrifugal fan according to the first sound pressure level and the second sound pressure level;

[0094] Among them, the separated noise time-domain signal includes the volute noise time-domain signal and the mechanical noise time-domain signal.

[0095] It should be noted that the noise separation model is a deep learning model. The trained noise separation model can separate the volute noise time-domain signal and the mechanical noise time-domain signal from the noise time-domain signal of the centrifugal fan. Therefore, the first sound pressure level of the volute noise time-domain signal and the first sound pressure level of the mechanical noise signal can be calculated respectively.

[0096] In this embodiment, by acquiring the time-domain signal of the centrifugal fan noise, separating the time-domain signals of the volute noise and the mechanical noise from the time-domain signal, calculating their first sound pressure levels respectively, converting the time-domain signal into the frequency-domain signal to calculate the second sound pressure level of the total noise for real-time monitoring of the noise, and adjusting the centrifugal fan according to the first sound pressure level and the second sound pressure level calculated by the monitoring, the noise of the centrifugal fan is reduced and the working efficiency of the centrifugal fan is improved.

[0097] In some embodiments, before step S20 of inputting the noise time-domain signal into the noise separation model to obtain the separated noise time-domain signal, as Figure 2 shown, the following steps are further included:

[0098] S11, acquire a plurality of simulated volute noises and simulated mechanical noises;

[0099] S12, randomly mix the simulated volute noise and the simulated mechanical noise to obtain the original data set;

[0100] S13, randomly sample the original data set with a fixed time window, perform gain summation on the sampled noise data to form new noise data, and obtain the training sample set;

[0101] S14, train the noise separation model with the training sample set.

[0102] It should be noted that the data set is made by mixing the simulated volute noise and the simulated mechanical noise. The volute noise may be caused by various reasons such as impeller wear, volute deformation, or foreign objects inside the volute, and the mechanical noise may also be caused by various reasons such as bearing wear, poor lubrication, or fan overload. In order to improve the separation performance of the noise separation model, mixed noise data is dynamically created to train the model. The gain can be obtained by generating a random number sequence, and it is only necessary to ensure that the signal-to-noise ratio of the volute noise and the mechanical noise is within the required range. The data of the obtained training sample set has no artificial tendency, and at the same time, it can simulate the situation where various problems occur mixed, so that the noise separation model can fully learn the noise data.

[0103] In this embodiment, by mixing and randomly sampling multiple simulated volute noises and multiple simulated mechanical noises, the noise data obtained by sampling is subjected to gain summation to form a training sample set, and the noise separation model is trained by the training sample set, thereby improving the noise separation performance of the noise separation model.

[0104] In some embodiments, in step S14, the noise separation model is trained by the training sample set, as Figure 3 shown, and it includes the following steps:

[0105] S141, perform a short-time Fourier transform on the noise data of the training sample set to obtain a feature signal;

[0106] S142, perform a block masking calculation on the feature signal to obtain a separated signal spectrum;

[0107] S143, perform an inverse short-time Fourier transform on the separated signal spectrum to obtain a separated noise time-domain signal;

[0108] S144, evaluate the separated noise time-domain signal and correct the parameter weights of the noise separation model.

[0109] It should be noted that the noise separation model performs a Fourier transform on the noise signal and represents it as a feature signal. By dividing the feature signal into overlapping blocks for feature extraction and estimating the spectral mask, the spectral mask is multiplied by the feature signal to obtain the volute noise frequency-domain signal and the mechanical noise frequency-domain signal.

[0110] In this embodiment, by performing a short-time Fourier transform, blocking, and modeling calculation on the noise data, a separated signal spectrum is obtained, and an inverse transform is performed on the separated signal spectrum to obtain a separated noise time-domain signal, providing a data basis for subsequent noise monitoring.

[0111] In some embodiments, in step S144, the separated noise time-domain signal is evaluated to correct the parameter weights of the noise separation model, as Figure 4 shown, and it includes the following steps:

[0112] S1441, perform a short-time Fourier transform on the simulated volute noise and the simulated mechanical noise in the training sample set to obtain a true signal spectrum;

[0113] S1442, construct a loss function based on the separated signal spectrum and the true signal spectrum to evaluate the separated noise time-domain signal;

[0114] The calculation formula of the loss function is as follows:

[0115] ;

[0116] Among them, is the loss function, is the true signal spectrum, is the separated signal spectrum.

[0117] It should be noted that in the process of noise separation, the noise time-domain signal is converted into a noise frequency-domain signal for convolution modeling. Therefore, when constructing the loss function, the output of the noise separation model is evaluated by measuring the similarity between the predicted signal, i.e., the separated signal, and the true signal in the frequency spectrum domain.

[0118] In this embodiment, the similarity between the separated signal spectrum and the true signal spectrum is measured by constructing a loss function, thereby correcting the parameter weights of the noise separation model and improving the noise separation performance of the noise separation model.

[0119] In some embodiments, in step S30, according to the separated noise time-domain signal, the first sound pressure level is calculated respectively, as Figure 5 shown, including the following steps:

[0120] S31, calculate the effective sound pressure of the separated noise in the time domain according to the separated noise time-domain signal;

[0121] S32, calculate the first sound pressure level according to the effective sound pressure of the separated noise in the time domain;

[0122] The calculation formula for the effective sound pressure of the separated noise in the time domain is as follows:

[0123] ;

[0124] where, is the effective sound pressure of the separated noise in the time domain, is the number of samples of the noise time-domain signal, is the th sound pressure at the sampling moment;

[0125] The calculation formula for the first sound pressure level is as follows:

[0126] ;

[0127] where, is the first sound pressure level, is the effective sound pressure of the separated noise in the time domain, is the atmospheric reference sound pressure.

[0128] It should be noted that the first sound pressure level is the intensity of the volute noise and the mechanical noise, which is used to reflect the volute noise and the mechanical noise during the operation of the centrifugal fan. By obtaining the sound pressure value at each sampling moment of the time-domain signal, in the case of a large number of samples, the effective sound pressure of the noise can be calculated through the sound pressure values at each sampling moment, thereby calculating the first sound pressure level of the volute noise and the first sound pressure level of the mechanical noise.

[0129] In this embodiment, by calculating the sound pressure levels of the two noises based on the time-domain signals of the volute noise and the mechanical noise, a data basis is provided for the subsequent adjustment of the centrifugal fan.

[0130] In some embodiments, in step S50, according to the noise frequency-domain signal, the second sound pressure level is calculated, as Figure 6 shown, including the following steps:

[0131] S51, according to the noise frequency-domain signal, calculate the noise frequency-domain amplitude;

[0132] S52, according to the noise frequency-domain amplitude, calculate the effective sound pressure in the noise frequency domain;

[0133] S53, according to the effective sound pressure in the noise frequency domain, calculate the second sound pressure level;

[0134] The calculation formula for the noise frequency-domain amplitude is as follows:

[0135] ;

[0136] Wherein, is the noise frequency-domain amplitude, is the noise frequency-domain signal, is the weighting operation symbol, is the frequency point number, is the sampling rate of the noise time-domain signal, is the number of samples of the noise time-domain signal;

[0137] The calculation formula for the effective sound pressure in the noise frequency domain is as follows:

[0138] ;

[0139] Wherein, is the effective sound pressure in the noise frequency domain, is the number of samples of the noise time-domain signal, is the noise frequency-domain amplitude;

[0140] The calculation formula for the second sound pressure level is as follows:

[0141] ;

[0142] Wherein, is the second sound pressure level, is the effective sound pressure in the noise frequency domain, is the atmospheric reference sound pressure.

[0143] It should be noted that in scenarios such as when a quieter environment is required during sleep, the monitoring and control of the centrifugal fan of the air conditioner cannot be carried out through the first sound pressure level. The first sound pressure level is the noise intensity of each component of the centrifugal fan and is used to judge the working state of the centrifugal fan. In the above scenarios, a smaller noise threshold is required to control the working state of the centrifugal fan. The second sound pressure value is the intensity of the total noise of the centrifugal fan after removing the signal frequencies that humans are not sensitive to. When calculating the amplitude of the noise frequency domain, the amplitudes of each frequency point are weighted and corrected to make the noise frequency domain signal match human hearing.

[0144] In this embodiment, by weighting and correcting the noise frequency domain signal, removing the signal frequencies that humans are not sensitive to, and calculating the sound pressure level of the total noise of the centrifugal fan, a data basis is provided for the subsequent adjustment of the centrifugal fan.

[0145] In some embodiments, in step S60, the centrifugal fan is adjusted according to the first sound pressure level and the second sound pressure level, as Figure 7 shown, including the following steps:

[0146] S61, comparing the first sound pressure level with the first sound pressure level to obtain the maximum first sound pressure level;

[0147] S62, comparing the maximum first sound pressure level with the working threshold of the noise time domain signal corresponding to the maximum first sound pressure level. When the maximum first sound pressure level is greater than the working threshold, the centrifugal fan is turned off and an abnormal alarm is issued;

[0148] S63, when the maximum first sound pressure level is less than the working threshold, comparing the second sound pressure level with the sensitivity threshold. When the second sound pressure level is greater than the sensitivity threshold, the speed of the centrifugal fan is reduced.

[0149] It should be noted that by comparing the sound pressure level of the volute noise with the sound pressure level of the mechanical noise, the noise with a larger sound pressure level indicates that this noise makes the main contribution to the centrifugal fan in this working state. By comparing the normal sound pressure level threshold of the noise time domain signal during operation with the calculated first sound pressure level, it can be known whether there is a problem with the centrifugal fan. When the maximum first sound pressure level is greater than the working threshold, an abnormal alarm message is issued. When there is no problem with the centrifugal fan, the second sound pressure level is compared with the sensitivity threshold. The sensitivity threshold is the minimum value of the sound pressure level at which humans feel discomfort with the noise signal and can be set according to the actual situation. When the second sound pressure level is greater than the sensitivity threshold, the speed of the centrifugal fan is reduced.

[0150] In this embodiment, the operating state of the centrifugal fan is adjusted by combining the first sound pressure level and the second sound pressure level. The first sound pressure level is used to monitor whether there is any abnormality in the centrifugal fan. On this basis, considering the sensitivity of humans to noise signals, the centrifugal fan is adjusted, reducing the noise of the centrifugal fan and improving the working efficiency of the centrifugal fan.

[0151] The present invention provides a noise on-line monitoring system for a multi-blade centrifugal fan, which applies the above-mentioned noise on-line monitoring method for a multi-blade centrifugal fan, as Figure 8 shown, and includes:

[0152] A time-domain signal acquisition module, configured to acquire the noise time-domain signal of the centrifugal fan;

[0153] A time-domain signal separation module, configured to input the noise time-domain signal into a noise separation model to obtain a separated noise time-domain signal;

[0154] A first sound pressure level calculation module, configured to calculate the first sound pressure level respectively according to the separated noise time-domain signal;

[0155] A frequency-domain signal acquisition module, configured to perform discrete Fourier transform on the noise time-domain signal to obtain a noise frequency-domain signal;

[0156] A second sound pressure level calculation module, configured to calculate the second sound pressure level according to the noise frequency-domain signal;

[0157] A centrifugal fan adjustment module, configured to adjust the centrifugal fan according to the first sound pressure level and the second sound pressure level;

[0158] Among them, the separated noise time-domain signal includes a volute noise time-domain signal and a mechanical noise time-domain signal.

[0159] In this embodiment, by acquiring the time-domain signal of the centrifugal fan noise, separating the time-domain signals of the volute noise and the mechanical noise from the time-domain signal, calculating their first sound pressure levels respectively, converting the time-domain signal into a frequency-domain signal to calculate the second sound pressure level of the total noise for real-time monitoring of the noise, and adjusting the centrifugal fan according to the first sound pressure level and the second sound pressure level obtained by the monitoring calculation, the noise of the centrifugal fan is reduced and the working efficiency of the centrifugal fan is improved.

[0160] It should be understood that the disclosed system can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For example, the division of the above-mentioned modules is only a logical function division. In actual implementation, there can be other division methods. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, each functional module can be integrated in a processing module, or each module can exist physically alone, or two or more modules can be integrated in one module. The above-mentioned integrated module can be implemented in the form of hardware or in the form of a software functional module.

[0161] The present invention provides an electronic device 2, such as Figure 9 shown, a processor 21 and a memory 22. The memory 22 is used to store computer program code, and the computer program code includes computer instructions. When the processor 21 executes the computer instructions, the electronic device executes the above-mentioned online noise monitoring method for a multi-vane centrifugal fan.

[0162] The electronic device 2 includes a processor 21, a memory 22, an output device 23, and an input device 24. The processor 21, the memory 22, the output device 23, and the input device 24 are coupled through a connector, and the connector includes various interfaces, transmission lines, or buses, etc. The embodiments of the present invention do not limit this. It should be understood that in various embodiments of the present invention, coupling refers to mutual connection in a specific manner, including direct connection or indirect connection through other devices. For example, they can be connected through various interfaces, transmission lines, buses, etc.

[0163] The processor 21 can be one or more graphics processing units (GPUs). When the processor 21 is a single GPU, the GPU can be a single-core GPU or a multi-core GPU. Optionally, the processor 21 can be a processor group composed of multiple GPUs, and multiple processors are coupled to each other through one or more buses. Optionally, the processor 21 can also be other types of processors, etc. The embodiments of the present invention do not limit this.

[0164] The memory 22 can be used to store computer program instructions and various computer program codes including the program codes for executing the solution of the present invention. Optionally, the memory 22 includes but is not limited to a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), or a compact disc read-only memory (CD-ROM), and the memory 22 is used for relevant instructions and data.

[0165] The input device 24 is used to input data and / or signals, and the output device 23 is used to output data and / or signals. The output device 23 and the input device 24 can be independent devices or an integrated device.

[0166] The present invention provides a computer-readable storage medium, in which a computer program is stored. The computer program includes program instructions, and when the program instructions are executed by a processor of an electronic device, the processor is caused to execute the above-mentioned online noise monitoring method for a multi-blade centrifugal fan.

[0167] The above are only specific embodiments of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A method for online noise monitoring of a multi-blade centrifugal fan, characterized in that: include: Obtain the noise time domain signal of the centrifugal fan; Inputting the noise time domain signal into a noise separation model to obtain a separated noise time domain signal; Calculating first sound pressure levels respectively according to the separated noise time domain signals; Performing discrete Fourier transform on the noise time domain signal to obtain a noise frequency domain signal; Calculating a second sound pressure level according to the noise frequency domain signal; adjusting a centrifugal fan according to the first sound pressure level and the second sound pressure level; The separated noise time domain signal includes a volute noise time domain signal and a mechanical noise time domain signal. The volute noise is caused by impeller wear, volute deformation or foreign matter inside the volute, and the mechanical noise is caused by bearing wear, poor lubrication or fan overload. The calculating the first sound pressure level respectively according to the separated noise time domain signal comprises: Calculating the effective sound pressure of the separated noise in the time domain according to the separated noise time domain signal; Calculating the first sound pressure level according to the effective sound pressure of the separated noise in time domain; The calculation formula of the effective sound pressure of the separated noise in the time domain is as follows: ； In the formula, To separate the effective sound pressure of noise in time domain, is the sampling number of the noisy time domain signal, For the The sound pressure at each sampling moment; The calculation formula of the first sound pressure level is as follows: ； In the formula, is the first sound pressure level, To separate the effective sound pressure of noise in time domain, is the atmospheric reference sound pressure; Wherein, before inputting the noise time domain signal into the noise separation model to obtain the separated noise time domain signal, the method further includes: Get multiple simulated volute noises and simulated mechanical noises; Randomly mixing the simulated volute noise and the simulated mechanical noise to obtain an original data set; The original data set is randomly sampled in a fixed time window, and the noise data obtained by sampling is gain-summed to form new noise data to obtain a training sample set; Training the noise separation model using the training sample set; Wherein, adjusting the centrifugal fan according to the first sound pressure level and the second sound pressure level includes: Comparing the first sound pressure level with the first sound pressure level to obtain a maximum first sound pressure level; Comparing the maximum first sound pressure level with a working threshold of a noise time domain signal corresponding to the maximum first sound pressure level, and when the maximum first sound pressure level is greater than the working threshold, shutting down the centrifugal fan and issuing an abnormal alarm; When the maximum first sound pressure level is less than the working threshold, the second sound pressure level is compared with a sensitive threshold, and when the second sound pressure level is greater than the sensitive threshold, the rotation speed of the centrifugal fan is reduced.

2. The method for online noise monitoring of a multi-blade centrifugal fan according to claim 1, characterized in that: The step of training the noise separation model by using the training sample set includes: Performing short-time Fourier transform on the noise data of the training sample set to obtain a characteristic signal; Performing block mask calculation on the characteristic signal to obtain a separated signal spectrum; Performing an inverse short-time Fourier transform on the separated signal spectrum to obtain the separated noise time domain signal; The separated noise time domain signal is evaluated and the parameter weights of the noise separation model are modified.

3. The method for online noise monitoring of a multi-blade centrifugal fan according to claim 2, characterized in that: The step of evaluating the separated noise time domain signal and correcting the parameter weights of the noise separation model includes: Performing short-time Fourier transform on the simulated volute noise and the simulated mechanical noise in the training sample set to obtain a real signal spectrum; According to the separated signal spectrum and the true signal spectrum, a loss function is constructed to evaluate the separated noise time domain signal.

4. The method for online noise monitoring of a multi-blade centrifugal fan according to claim 3, characterized in that: The calculation formula of the loss function is as follows: ； in, is the loss function, is the real signal spectrum, To separate the signal spectrum.

5. The method for online noise monitoring of a multi-blade centrifugal fan according to claim 1, characterized in that: The step of calculating the second sound pressure level according to the noise frequency domain signal comprises: Calculating the noise frequency domain amplitude according to the noise frequency domain signal; Calculating the effective sound pressure in the noise frequency domain according to the noise frequency domain amplitude; Calculating the second sound pressure level according to the effective sound pressure in the noise frequency domain; The calculation formula of the noise frequency domain amplitude is as follows: ； in, is the noise frequency domain amplitude, is the noise frequency domain signal, is the weighting operation symbol, is the frequency point number, is the sampling rate of the noisy time domain signal, is the sampling number of the noisy time domain signal; The calculation formula of the effective sound pressure of the noise frequency domain is as follows: ； in, is the effective sound pressure in the noise frequency domain, is the sampling number of the noisy time domain signal, is the noise frequency domain amplitude; The calculation formula of the second sound pressure level is as follows: ； in, is the second sound pressure level, is the effective sound pressure in the noise frequency domain, is the atmospheric reference sound pressure.

6. A noise online monitoring system for a multi-blade centrifugal fan, using the noise online monitoring method for a multi-blade centrifugal fan as claimed in any one of claims 1 to 5, characterized in that: include: A time domain signal acquisition module is used to acquire the noise time domain signal of the centrifugal fan; A time domain signal separation module, used for inputting the noise time domain signal into a noise separation model to obtain a separated noise time domain signal; A first sound pressure level calculation module, used to calculate the first sound pressure level according to the separated noise time domain signal; A frequency domain signal acquisition module, used for performing discrete Fourier transform on the noise time domain signal to obtain a noise frequency domain signal; A second sound pressure level calculation module, used for calculating a second sound pressure level according to the noise frequency domain signal; a centrifugal fan adjustment module, configured to adjust the centrifugal fan according to the first sound pressure level and the second sound pressure level; The separated noise time domain signal includes a volute noise time domain signal and a mechanical noise time domain signal.

7. An electronic device, characterized in that: It includes a processor and a memory, the memory is used to store computer program code, the computer program code includes computer instructions, when the processor executes the computer instructions, the electronic device executes the noise online monitoring method of a multi-blade centrifugal fan as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which includes program instructions. When the program instructions are executed by a processor of an electronic device, the processor executes a method for online noise monitoring of a multi-blade centrifugal fan as described in any one of claims 1 to 5.

Images