Method, apparatus, computer device, and storage medium for noise field modeling

Abstract

The present application relates to a method, apparatus, computer device, storage medium, and computer program product for noise field modeling. The method includes: obtaining a principal component frequency sound field and a principal component frequency sound field diagram of a noise signal through principal component analysis according to the noise signal; obtaining a noise field distribution through a MUSIC algorithm according to the principal component frequency sound field; obtaining an initial three-dimensional noise field model through a fusion algorithm and Gaussian fitting according to the noise field distribution and the principal component frequency sound field diagram; and fusing the initial three-dimensional noise field model and an environmental signal to obtain a three-dimensional noise field model. The three-dimensional noise field model obtained by using this method can support obtaining an accurate noise field distribution result.

Description

Technical Field

[0001] This application relates to the technical field of noise, and particularly to a method, device, computer device, storage medium, and computer program product for noise field modeling. Background Art

[0002] With the development of the power industry, the harm caused by vibration noise has attracted more and more attention. Equipment failures will generate vibration noise, and important information about the working state is contained in the noise. Therefore, obtaining the distribution of the noise field through noise field modeling can well reflect the faults of power equipment and predict the health status of power equipment, avoiding the occurrence of equipment failures.

[0003] In traditional technologies, a data acquisition method using single-channel testing is mostly adopted, and vibration test analysis means and ideas are used to process sound signals. However, this method has the problem of inaccurate noise field distribution obtained in the case of mutual interference of multiple sound sources. Summary of the Invention

[0004] Based on this, in view of the technical problem of inaccurate noise field distribution obtained by traditional technologies, it is necessary to provide a noise field modeling method, device, computer device, computer-readable storage medium, and computer program product that can support obtaining accurate noise field distribution results.

[0005] In a first aspect, this application provides a noise field modeling method. The method includes:

[0006] Collect noise signals and environmental signals;

[0007] According to the noise signals, obtain the principal component frequency sound field and the principal component frequency sound field diagram of the noise signals through the principal component analysis method;

[0008] According to the principal component frequency sound field, obtain the noise field distribution through the MUSIC algorithm;

[0009] According to the noise field distribution and the principal component frequency sound field diagram, obtain the initial three-dimensional noise field model through the fusion algorithm and Gaussian fitting;

[0010] Fuse the initial three-dimensional noise field model and the environmental signals to obtain the three-dimensional noise field model.

[0011] In one embodiment, according to the noise signals, obtaining the principal component frequency sound field and the principal component frequency sound field diagram of the noise signals through the principal component analysis method includes:

[0012] According to the noise signals, obtain the frequency domain data of the noise signals through frequency domain analysis;

[0013] Based on the frequency-domain data, the principal component frequency sound field and the principal component frequency sound field diagram of the noise signal are obtained through principal component analysis.

[0014] In one embodiment, based on the principal component frequency sound field, the noise field distribution is obtained through the MUSIC algorithm, including:

[0015] Based on the principal component frequency sound field, the signal subspace and the noise subspace are obtained through frame processing and Fourier transform;

[0016] Based on the orthogonality of the signal subspace and the noise subspace, the spatial spectrum is obtained;

[0017] Based on the spatial spectrum, the noise field distribution is obtained.

[0018] In one embodiment, based on the noise field distribution and the principal component frequency sound field diagram, the initial three-dimensional noise field model is obtained through a fusion algorithm and Gaussian fitting, including:

[0019] Obtain the peak value of the spectral peak of the spatial spectrum;

[0020] Based on the principal component frequency sound field diagram and the peak value of the spectral peak of the spatial spectrum, the fusion value of the peak value of the spectral peak of the spatial spectrum is obtained through a fusion algorithm;

[0021] Based on the fusion value and the noise field distribution, the initial three-dimensional noise field model is obtained through Gaussian fitting.

[0022] In one embodiment, based on the fused initial three-dimensional noise field model and the environmental signal, the three-dimensional noise field model is obtained, including:

[0023] Establish a three-dimensional physical environment model through the initial three-dimensional noise field model and the environmental signal;

[0024] Based on the three-dimensional physical environment model, the lightweight processed three-dimensional physical environment model is obtained through lightweight processing;

[0025] Through the model fusion algorithm, the initial three-dimensional noise field model and the lightweight processed three-dimensional physical environment model are fused to obtain the three-dimensional noise field model.

[0026] In one embodiment, it further includes:

[0027] Collect the noise signal to be measured and the environmental signal to be measured;

[0028] Input the noise signal to be measured and the environmental signal to be measured into the three-dimensional noise field model to obtain the noise field distribution result corresponding to the noise signal to be measured and the environmental signal to be measured.

[0029] In a second aspect, the present application also provides a noise field modeling device. The device includes:

[0030] A signal acquisition module for acquiring noise signals and environmental signals;

[0031] A principal component analysis module for obtaining the principal component frequency sound field and the principal component frequency sound field diagram of the noise signal according to the noise signal through the principal component analysis method;

[0032] A noise field distribution acquisition module for obtaining the noise field distribution according to the principal component frequency sound field through the MUSIC algorithm;

[0033] A noise field model acquisition module for obtaining an initial three-dimensional noise field model through a fusion algorithm and Gaussian fitting according to the noise field distribution and the principal component frequency sound field diagram;

[0034] A model fusion module for fusing the initial three-dimensional noise field model and the environmental signal to obtain a three-dimensional noise field model.

[0035] In a third aspect, the present application also provides a computer device. The computer device includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, the following steps are implemented:

[0036] Acquire noise signals and environmental signals, obtain the principal component frequency sound field and the principal component frequency sound field diagram of the noise signal according to the noise signal through the principal component analysis method, obtain the noise field distribution according to the principal component frequency sound field through the MUSIC algorithm, obtain an initial three-dimensional noise field model through a fusion algorithm and Gaussian fitting according to the noise field distribution and the principal component frequency sound field diagram, and fuse the initial three-dimensional noise field model and the environmental signal to obtain a three-dimensional noise field model.

[0037] In a fourth aspect, the present application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the following steps are implemented:

[0038] Acquire noise signals and environmental signals, obtain the principal component frequency sound field and the principal component frequency sound field diagram of the noise signal according to the noise signal through the principal component analysis method, obtain the noise field distribution according to the principal component frequency sound field through the MUSIC algorithm, obtain an initial three-dimensional noise field model through a fusion algorithm and Gaussian fitting according to the noise field distribution and the principal component frequency sound field diagram, and fuse the initial three-dimensional noise field model and the environmental signal to obtain a three-dimensional noise field model.

[0039] In a fifth aspect, the present application also provides a computer program product. The computer program product includes a computer program. When the computer program is executed by a processor, the following steps are implemented:

[0040] Collect noise signals and environmental signals. According to the noise signals, obtain the principal component frequency sound field and the principal component frequency sound field diagram of the noise signals through the principal component analysis method. According to the principal component frequency sound field, obtain the noise field distribution through the MUSIC algorithm. According to the noise field distribution and the principal component frequency sound field diagram, obtain the initial three-dimensional noise field model through the fusion algorithm and Gaussian fitting. Integrate the initial three-dimensional noise field model and the environmental signals to obtain the three-dimensional noise field model.

[0041] The above noise field modeling method, device, computer device, storage medium and computer program product collect noise signals and environmental signals. According to the noise signals, obtain the principal component frequency sound field and the principal component frequency sound field diagram of the noise signals through the principal component analysis method. According to the principal component frequency sound field, obtain the noise field distribution through the MUSIC algorithm. According to the noise field distribution and the principal component frequency sound field diagram, obtain the initial three-dimensional noise field model through the fusion algorithm and Gaussian fitting. Integrate the initial three-dimensional noise field model and the environmental signals to obtain the three-dimensional noise field model. The above solution collects noise signals and environmental signals, obtains the initial three-dimensional noise field model through the principal component analysis method, MUSIC algorithm, fusion algorithm and Gaussian fitting, integrates the initial three-dimensional noise field model and the environmental signals to obtain an accurate three-dimensional noise field model. Based on this accurate three-dimensional noise field model, an accurate noise field distribution result can be supported. Description of the Drawings

[0042] Figure 1 It is an application environment diagram of the noise field modeling method in an embodiment;

[0043] Figure 2 It is a schematic flowchart of the noise field modeling method in an embodiment;

[0044] Figure 3 It is a schematic flowchart of the noise field modeling steps in an embodiment;

[0045] Figure 4 It is a schematic flowchart of the noise field modeling method in another embodiment;

[0046] Figure 5 It is a structural block diagram of the noise field modeling device in an embodiment;

[0047] Figure 6 It is an internal structure diagram of a computer device in an embodiment. Detailed Embodiments

[0048] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0049] The noise field modeling method provided by the embodiment of the present application can be applied to an application environment as follows Figure 1 shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104, or can be placed in the cloud or on other network servers. The terminal 102 collects noise signals and environmental signals, and according to the noise signals, through the principal component analysis method, obtains the principal component frequency sound field and the principal component frequency sound field diagram of the noise signals. According to the principal component frequency sound field, through the MUSIC (Multiple Signal Classification) algorithm, obtains the noise field distribution. According to the noise field distribution and the principal component frequency sound field diagram, through the fusion algorithm and Gaussian fitting, obtains the initial three-dimensional noise field model. Fuses the initial three-dimensional noise field model and the environmental signals to obtain the three-dimensional noise field model. The server 104 receives the three-dimensional noise field model result sent by the terminal 102. Among them, the terminal 102 can be but is not limited to various personal computers, laptop computers, smart phones, tablet computers, Internet of Things devices, and portable wearable devices. The Internet of Things devices can be smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, etc. The portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. The server 104 can be implemented by an independent server or a server cluster composed of multiple servers.

[0050] In one embodiment, as Figure 2 shown, a noise field modeling method is provided. Taking the terminal in Figure 1 as an example, the method includes the following steps:

[0051] S100, collect noise signals and environmental signals.

[0052] Among them, useful information is often carried in the signal source, and the superposition of noise and interference makes the signal difficult to identify. The equipment that generates noise in the substation includes transformers, reactors, cooling and exhaust devices, etc. As the load of the power equipment changes, the noise intensity and frequency will change significantly. Under different operating states, the noise signals exhibit different characteristics. Therefore, collecting noise signals and environmental signals is beneficial to the processing of noise signals and the analysis of the distribution law of noise signals.

[0053] Specifically, collect noise signals and environmental signals in the space. Optionally, this solution uses array measurement means to collect noise signals and environmental signals. Specifically, array microphones can be used to collect sound field signals.

[0054] S200, according to the noise signals, through the principal component analysis method, obtain the principal component frequency sound field and the principal component frequency sound field diagram of the noise signals.

[0055] Among them, the principal component analysis is a statistical method that transforms a set of variables that may be correlated through orthogonal transformation into a set of linearly uncorrelated variables. This set of transformed variables is called the principal components. Principal component analysis can extract a few comprehensive indicators from the individual information reflected by a relatively large number of original observable indicators. These indicators are mutually independent and can maximize the information reflected by the original relatively large number of indicators. Furthermore, a few comprehensive indicators are used to characterize the individual. Therefore, through principal component analysis of the noise signal, the principal components of the noise signal can be determined.

[0056] Specifically, through principal component analysis of the noise signal, the principal component frequency sound field and the principal component frequency sound field diagram of the noise signal are obtained.

[0057] S300. According to the principal component frequency sound field, the noise field distribution is obtained through the MUSIC algorithm.

[0058] Among them, the MUSIC algorithm is a class of spatial spectrum estimation algorithms that perform eigen-decomposition on the covariance matrix of the output data of any array, thereby obtaining the signal subspace corresponding to the signal components and the noise subspace orthogonal to the signal components. Then, using the orthogonality of the two subspaces, a spatial spectrum function is constructed, and through spectrum peak search, the parameters of the signal are estimated. The parameters can include the incident direction, polarization information, and signal strength. The noise field distribution can be reflected by the azimuth angle parameter of the signal. Therefore, the MUSIC algorithm is used to obtain the distribution of the noise field.

[0059] Specifically, according to the principal component frequency sound field, the noise field distribution is obtained through the MUSIC algorithm.

[0060] S400. According to the noise field distribution and the principal component frequency sound field diagram, an initial three-dimensional noise field model is obtained through the fusion algorithm and Gaussian fitting.

[0061] Among them, data fusion refers to making full use of the multi-sensor data resources in different times and spaces, using computer technology to analyze, synthesize, dominate, and use the multi-sensor observation data obtained in time series under certain criteria, obtaining a consistent interpretation and description of the measured object, and then realizing the corresponding decision-making and estimation, so that the system obtains more sufficient information than its individual components. Weighted data fusion is to perform weighted averaging on multi-source redundant information, and the result is used as the fusion value, which is a method that directly operates on the data source. Gaussian fitting is a fitting method that uses a Gaussian function in the form of Gi(x) = Ai * exp((x - Bi)^2 / Ci^2) to approximate the data point set.

[0062] Specifically, according to the noise field distribution and the principal component frequency sound field diagram, an initial three-dimensional noise field model is obtained through the fusion algorithm and Gaussian fitting.

[0063] S500 fuses the initial three-dimensional noise field model and the environmental signal to obtain the three-dimensional noise field model.

[0064] Among them, fusion means integrating multiple training models into one model according to a certain method through a model fusion algorithm. The model fusion methods include linear weighted fusion method, cross fusion method, waterfall fusion method, many and different fusion, prediction fusion method, and additive fusion.

[0065] Specifically, the initial three-dimensional noise field model and the environmental signal are fused to obtain the three-dimensional noise field model.

[0066] In the above noise field modeling method, by collecting the noise signal and the environmental signal, based on the principal component analysis method, MUSIC algorithm, fusion algorithm, and Gaussian fitting, the initial three-dimensional noise field model is obtained. The initial three-dimensional noise field model and the environmental signal are fused to obtain an accurate three-dimensional noise field model. Based on this accurate three-dimensional noise field model, an accurate noise field distribution result can be supported.

[0067] In one embodiment, S200 obtains the principal component frequency sound field and the principal component frequency sound field diagram of the noise signal according to the noise signal through the principal component analysis method, including:

[0068] S220: Obtain the frequency domain data of the noise signal through frequency domain analysis according to the noise signal.

[0069] S240: Obtain the principal component frequency sound field and the principal component frequency sound field diagram of the noise signal through principal component analysis according to the frequency domain data.

[0070] Standardization processes the data according to the columns of the feature matrix, converting the feature values of the samples to the same dimension. The methods of standardization processing include linear transformation method, zero-mean normalization, decimal scaling normalization, logarithmic Logistic mode, and fuzzy quantization mode. Optionally, this solution adopts zero-mean normalization, that is, y = (x - the average value of X) / the standard deviation of X. The obtained noise signal is converted into a frequency domain signal through discrete Fourier transform, and the obtained frequency domain signal is standardized to obtain the standardized result of the frequency domain data.

[0071] Then, the correlation coefficient matrix is composed of the correlation coefficients between the columns of the matrix. The element in the i-th row and j-th column of the correlation coefficient matrix is the correlation coefficient between the i-th column and the j-th column of the original matrix. Based on the normalization result of the frequency-domain data, the correlation coefficient matrix is calculated. Further, the characteristic equation of the correlation coefficient matrix is solved to obtain the eigenvalues and eigenvectors of the correlation coefficient matrix. The obtained eigenvalues are sorted from largest to smallest, and the principal components with eigenvalues greater than 1 are extracted. According to the principal components of the noise signal, the principal component frequency sound field of the noise signal is determined, and the principal component frequency sound field diagram corresponding to the principal component frequency sound field is drawn. The principal component frequency sound field diagram can be the contour diagram of the noise field distribution of the principal component frequency.

[0072] The solution of the above embodiment obtains the frequency-domain normalization result of the noise signal through the normalization processing and frequency-domain analysis of the noise signal. By using the principal component analysis method for the frequency-domain normalization result of the noise signal, the principal component frequency sound field and the principal component frequency sound field diagram of the noise signal are obtained, providing a prerequisite for noise field modeling.

[0073] In one of the embodiments, as Figure 3 shown, S300 obtains the noise field distribution through the MUSIC algorithm according to the principal component frequency sound field, including:

[0074] S320: According to the principal component frequency sound field, through frame segmentation processing and Fourier transform, the signal subspace and the noise subspace are obtained.

[0075] S340: According to the orthogonality of the signal subspace and the noise subspace, the spatial spectrum is obtained.

[0076] S360: According to the spatial spectrum, the noise field distribution is obtained.

[0077] Overall, the characteristics of the noise signal are a non-stationary process that changes over time. However, within a short time range, its characteristics remain relatively stable. Therefore, the noise signal can be analyzed in the short term, that is, the noise signal is divided into segments to analyze its characteristic parameters, and each segment can be called a frame. The frame segmentation processing is generally implemented by the method of overlapping segmentation and using a weighted method with a movable window of a fixed length to achieve smooth transition between frames and maintain its continuity. The time-domain analysis of non-stationary signals includes short-time Fourier transform, wavelet transform, Hilbert transform, and Hilbert-Huang transform. Specifically, this solution uses the short-time Fourier transform. According to the obtained principal component frequency sound field, frame segmentation processing is performed on the principal component frequency sound field to obtain the frame-segmented principal component frequency sound field. The frame-segmented principal component frequency sound field is subjected to short-time Fourier transform to obtain a two-dimensional function of the time domain and frequency domain of the principal component frequency sound field. Further, the covariance matrix of the time domain and frequency domain of the principal component frequency sound field is obtained, and the covariance matrix is subjected to eigenvalue decomposition to obtain the signal subspace and the noise subspace of the principal component frequency sound field.

[0078] Then, the orthogonal complement space of the signal subspace is the noise subspace. The source signal has the maximum projection in the signal subspace and a projection of zero in the orthogonal complement space of the signal subspace. The noise subspace is determined by the source signal. Based on the orthogonality of the signal subspace and the noise subspace, a spatial spectrum is obtained. Specifically, a minimum optimization search is used to achieve the orthogonality of the signal subspace and the noise subspace, and the spatial spectrum corresponding to the principal component frequency sound field is obtained. According to the search range of the azimuth angle parameter, the spectral peaks of the spatial spectrum are searched respectively, and the value of the azimuth angle signal parameter corresponding to the maximum value of the spectral peak peak of the spatial spectrum is found. The obtained value of the azimuth angle signal parameter is the noise field distribution.

[0079] The solution of the above embodiment obtains the signal subspace and the noise subspace through frame-by-frame processing and Fourier transform using the principal component frequency sound field. Based on the orthogonality of the signal subspace and the noise subspace, a spatial spectrum is obtained. According to the spatial spectrum, the noise field distribution is obtained, providing a prerequisite for noise field modeling and enabling an accurate noise field distribution result to be obtained.

[0080] In one of the embodiments, S400 obtains an initial three-dimensional noise field model through a fusion algorithm and Gaussian fitting based on the noise field distribution and the principal component frequency sound field diagram.

[0081] S420: Obtain the spectral peak peak of the spatial spectrum.

[0082] S440: Based on the principal component frequency sound field diagram and the spectral peak peak of the spatial spectrum, obtain the fusion value of the spectral peak peak of the spatial spectrum through a fusion algorithm.

[0083] S460: Obtain the initial three-dimensional noise field model through Gaussian fitting using the fusion value and the noise field distribution.

[0084] Obtain the spectral peak peak of the spatial spectrum. Based on the obtained principal component frequency sound field diagram, obtain the fusion value of the spectral peak peak of the spatial spectrum through a fusion algorithm. Specifically, according to Ns principal component frequency sound fields corresponding to Ns groups of spatial spectrum subbands, calculate the variance of the spectral peak peak of each group of subbands, and obtain the fusion value by passing the spectral peak peak of each group of subbands through a fusion algorithm. Optionally, in this solution, a weighted fusion algorithm is used to perform weighted fusion on the spectral peak peak of the first subband with the spectral peak peaks of the second subband, the third subband... the Ns-th subband to obtain the first observation value. Then, perform weighted fusion on the first observation value with the spectral peak peaks of the third subband, the fourth subband... the Ns-th subband to obtain the second observation value. Iterate according to the above steps to obtain the final fusion value of the spectral peak peak of the spatial spectrum.

[0085] Further, the fusion value of the spectral peak peak value of the obtained spatial spectrum is described by a Gaussian function, the description equation is transformed into a quadratic polynomial fitting function, the fitting function is solved, and the position parameters of the Gaussian curve are obtained. The position parameters of the Gaussian curve include the position where the peak of the fitting curve is located, the peak height of the fitting curve, and the peak width of the fitting curve. The fitted Gaussian curve is the initial three-dimensional model of the noise signal.

[0086] The solution of the above embodiment obtains the spectral peak peak value of the spatial spectrum. According to the main component frequency sound field diagram and the spectral peak peak value of the spatial spectrum, through a fusion algorithm, the fusion value of the spectral peak peak value of the spatial spectrum is obtained. Through the fusion value and the noise field distribution, through Gaussian fitting, the initial three-dimensional noise field model is obtained, which provides a prerequisite for noise field modeling and can support obtaining accurate noise field distribution results.

[0087] In one of the embodiments, fusing the initial three-dimensional noise field model and the environmental signal to obtain the three-dimensional noise field model includes:

[0088] S520: Establish a three-dimensional physical environment model through the initial three-dimensional noise field model and the environmental signal.

[0089] S540: Obtain the lightweight processed three-dimensional physical environment model according to the three-dimensional physical environment model through lightweight processing.

[0090] S560: Through the model fusion algorithm, fuse the initial three-dimensional noise field model and the lightweight processed three-dimensional physical environment model to obtain the three-dimensional noise field model.

[0091] According to the initial three-dimensional noise field model, the collected environmental signal is scaled proportionally to establish a three-dimensional environment model. The environmental signal includes physical objects in the physical environment. The obtained three-dimensional environment model is subjected to lightweight processing. Specifically, during the modeling process, simple geometric bodies are used for construction, segmented straight lines are used to replace curves, cuboids or hexagonal prisms are used to replace cylinders, and the number of segments of the geometric bodies is reduced, that is, simple graphics are used for construction. For some partial models in the scene, they can be appropriately simplified, or some details can be ignored, redundant geometric bodies and lines are deleted, the joint surfaces between geometric bodies are checked, and the invisible surfaces are deleted, that is, the lightweight processed three-dimensional environment model is obtained.

[0092] Further, the initial three-dimensional noise field model and the lightweight processed three-dimensional physical environment model are maintained in the same real-time state, and through the model fusion algorithm, the fusion of the three-dimensional sound field is completed to obtain the three-dimensional noise field model.

[0093] The solution of the above embodiment establishes a three-dimensional physical environment model through the initial three-dimensional noise field model and environmental signals. According to the three-dimensional physical environment model, through lightweight processing, a lightweight processed three-dimensional physical environment model is obtained. Through a model fusion algorithm, the initial three-dimensional noise field model and the lightweight processed three-dimensional physical environment model are fused to obtain a three-dimensional noise field model. This three-dimensional noise field model can support obtaining an accurate noise field distribution result.

[0094] In one embodiment, as Figure 4 shown, a noise field modeling method includes:

[0095] S620: Collect the noise signal to be measured and the environmental signal to be measured.

[0096] S640: Input the noise signal to be measured and the environmental signal to be measured into the three-dimensional noise field model to obtain the noise field distribution result corresponding to the noise signal to be measured and the environmental signal to be measured.

[0097] In this embodiment, by collecting the noise signal to be measured and the environmental signal to be measured, inputting the noise signal to be measured and the environmental signal to be measured into the three-dimensional noise field model, an initial three-dimensional noise field model is established through the transmitted noise signal data to be measured, and then the three-dimensional environment model is driven to keep the three-dimensional physical model and the initial three-dimensional noise field model in the physical space in the same real-time state, obtaining a fused three-dimensional noise field model. According to this three-dimensional noise field model, an accurate noise field distribution result is obtained.

[0098] The solution of the above embodiment can achieve the effect of supporting accurately obtaining the noise field distribution result by collecting the noise signal to be measured and the environmental signal to be measured and inputting the noise signal to be measured and the environmental signal to be measured into the three-dimensional noise field model to obtain the noise field distribution result corresponding to the noise signal to be measured and the environmental signal to be measured.

[0099] To illustrate the method and effect of noise field modeling in this solution in detail, the following takes a most detailed embodiment for illustration:

[0100] Collect noise signals and environmental signals. According to the noise signals, obtain the principal component frequency sound field and the principal component frequency sound field diagram of the noise signals through the principal component analysis method. According to the principal component frequency sound field, obtain the noise field distribution through the MUSIC algorithm. According to the noise field distribution and the principal component frequency sound field diagram, obtain the initial three-dimensional noise field model through the fusion algorithm and Gaussian fitting. Integrate the initial three-dimensional noise field model and the environmental signals to obtain the three-dimensional noise field model. According to the noise signals, obtain the frequency domain data of the noise signals through frequency domain analysis. According to the frequency domain data, obtain the principal component frequency sound field and the principal component frequency sound field diagram of the noise signals through principal component analysis. According to the principal component frequency sound field, obtain the signal subspace and the noise subspace through frame processing and Fourier transform. Use the minimum optimization search algorithm to make the signal subspace and the noise subspace orthogonal to obtain the spatial spectrum, where the spatial spectrum of the m-th sub-band is:

[0101]

[0102] Calculate and obtain the average spectral matrix of Ns sub-bands:

[0103]

[0104] According to the azimuth signal parameters r, Search for the spectral peaks within the search ranges of θ respectively, and find the values corresponding to the maximum extreme points of the peaks, which are the signal parameters r, θ values to obtain the noise field distribution. Obtain the peak value of the spatial spectrum. According to the principal component frequency sound field diagram and the peak value of the spatial spectrum, obtain the fusion value of the peak value of the spatial spectrum through the fusion algorithm. Describe the obtained fusion value with a Gaussian function, and the description equation is:

[0105]

[0106] Take the logarithm of both sides of the above description equation and transform it into a quadratic polynomial fitting function, and this function is:

[0107]

[0108]

[0109]

[0110]

[0111] Obtain x0, y according to the least squares principle max, from the expression of S, the position of the Gaussian curve can be obtained, and then the initial three-dimensional noise field model can be obtained. By fusing the initial three-dimensional noise field model and the environmental signal, a three-dimensional physical environment model is established. According to the three-dimensional physical environment model, through lightweight processing, the lightweight processed three-dimensional physical environment model is obtained. Through the model fusion algorithm, the initial three-dimensional noise field model and the lightweight processed three-dimensional physical environment model are fused to obtain the three-dimensional noise field model. The measured noise signal and the measured environmental signal are collected, and the measured noise signal and the measured environmental signal are input into the three-dimensional noise field model to obtain the three-dimensional noise field distribution result corresponding to the measured noise signal and the measured environmental signal.

[0112] For the solution of the above embodiment, by collecting the noise signal and the environmental signal, according to the principal component analysis method, the MUSIC algorithm, the fusion algorithm and Gaussian fitting, the initial three-dimensional noise field model is obtained. By fusing the initial three-dimensional noise field model and the environmental signal, an accurate three-dimensional noise field model is obtained. Based on this accurate three-dimensional noise field model, an accurate noise field distribution result can be supported.

[0113] It should be understood that although each step in the flowcharts involved in the above embodiments is displayed in sequence according to the indication of the arrows, these steps do not necessarily need to be executed in the order indicated by the arrows. Unless there is a clear description in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above embodiments may include multiple steps or multiple stages. These steps or stages do not necessarily need to be executed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be executed alternately or alternately with at least a part of other steps or steps or stages in other steps.

[0114] Based on the same inventive concept, the embodiment of the present application also provides a noise field modeling device for implementing the noise field modeling method involved above. The implementation solution provided by this device to solve the problem is similar to the implementation solution described in the above method. Therefore, the specific limitations in one or more embodiments of the noise field modeling device provided below can refer to the limitations on the noise field modeling method in the above text, and will not be repeated here.

[0115] In one embodiment, as Figure 5 shown, a noise field modeling device 700 is provided, including: a signal acquisition module 710, a principal component analysis module 720, a noise field distribution acquisition module 730, a noise field model acquisition module 740, and a model fusion module 750, where:

[0116] The signal acquisition module 710 is used to acquire the noise signal and the environmental signal.

[0117] The principal component analysis module 720 is configured to obtain the principal component frequency sound field and the principal component frequency sound field diagram of the noise signal according to the noise signal through the principal component analysis method.

[0118] The noise field distribution acquisition module 730 is configured to obtain the noise field distribution according to the principal component frequency sound field through the MUSIC algorithm.

[0119] The noise field model acquisition module 740 is configured to obtain an initial three-dimensional noise field model according to the noise field distribution and the principal component frequency sound field diagram through a fusion algorithm and Gaussian fitting.

[0120] The model fusion module 750 is configured to fuse the initial three-dimensional noise field model and the environmental signal to obtain a three-dimensional noise field model.

[0121] In the above noise field modeling device, by collecting the noise signal and the environmental signal, according to the noise signal, through the principal component analysis method, the principal component frequency sound field and the principal component frequency sound field diagram of the noise signal are obtained. According to the principal component frequency sound field, through the MUSIC algorithm, the noise field distribution is obtained. According to the noise field distribution and the principal component frequency sound field diagram, through the fusion algorithm and Gaussian fitting, an initial three-dimensional noise field model is obtained. By fusing the initial three-dimensional noise field model and the environmental signal, an accurate three-dimensional noise field model is obtained. Based on this accurate three-dimensional noise field model, an accurate noise field distribution result can be supported.

[0122] In one embodiment, the principal component analysis module 720 is further configured to obtain the frequency domain data of the noise signal according to the noise signal through frequency domain analysis, and obtain the principal component frequency sound field and the principal component frequency sound field diagram of the noise signal according to the frequency domain data through principal component analysis.

[0123] In one embodiment, the noise field distribution acquisition module 730 is further configured to obtain the signal subspace and the noise subspace according to the principal component frequency sound field through frame processing and Fourier transform, obtain the spatial spectrum according to the orthogonality of the signal subspace and the noise subspace, and obtain the noise field distribution according to the spatial spectrum.

[0124] In one embodiment, the noise field model acquisition module 740 is further configured to obtain the spectral peak value of the spatial spectrum, obtain the fusion value of the spectral peak value of the spatial spectrum according to the principal component frequency sound field diagram and the spectral peak value of the spatial spectrum through a fusion algorithm, and obtain an initial three-dimensional noise field model through the fusion value and the noise field distribution through Gaussian fitting.

[0125] In one embodiment, the model fusion module 750 is further configured to establish a three-dimensional physical environment model through an initial three-dimensional noise field model and environmental signals, obtain a lightweight processed three-dimensional physical environment model according to the three-dimensional physical environment model through lightweight processing, and fuse the initial three-dimensional noise field model and the lightweight processed three-dimensional physical environment model through a model fusion algorithm to obtain a three-dimensional noise field model.

[0126] In one embodiment, the noise field modeling device 700 is further configured to collect a to-be-measured noise signal and a to-be-measured environmental signal, and input the to-be-measured noise signal and the to-be-measured environmental signal into the three-dimensional noise field model to obtain a noise field distribution result corresponding to the to-be-measured noise signal and the to-be-measured environmental signal.

[0127] Each module in the above noise field modeling device can be implemented in whole or in part by software, hardware, and their combination. The above modules can be embedded in the processor of the computer device in hardware form or be independent of it, or be stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to the above respective modules.

[0128] In one embodiment, a computer device is provided. The computer device can be a server, and its internal structure diagram can be as Figure 6 shown. The computer device includes a processor, a memory, and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store noise field model data. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, it implements a noise field modeling method.

[0129] Those skilled in the art can understand that Figure 6 the structure shown in

[0130] is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0131] Collect a noise signal and an environmental signal;

[0132] Based on the noise signal, the principal component frequency sound field and the principal component frequency sound field diagram of the noise signal are obtained through the principal component analysis method;

[0133] Based on the principal component frequency sound field, the noise field distribution is obtained through the MUSIC algorithm;

[0134] Based on the noise field distribution and the principal component frequency sound field diagram, an initial three-dimensional noise field model is obtained through the fusion algorithm and Gaussian fitting;

[0135] The initial three-dimensional noise field model and the environmental signal are fused to obtain the three-dimensional noise field model.

[0136] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0137] Based on the noise signal, the frequency domain data of the noise signal is obtained through frequency domain analysis. Based on the frequency domain data, the principal component frequency sound field and the principal component frequency sound field diagram of the noise signal are obtained through principal component analysis.

[0138] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0139] Based on the principal component frequency sound field, the signal subspace and the noise subspace are obtained through frame processing and Fourier transform. Based on the orthogonality of the signal subspace and the noise subspace, the spatial spectrum is obtained. Based on the spatial spectrum, the noise field distribution is obtained.

[0140] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0141] The peak value of the spectrum peak of the spatial spectrum is obtained. Based on the principal component frequency sound field diagram and the peak value of the spectrum peak of the spatial spectrum, the fusion value of the peak value of the spectrum peak of the spatial spectrum is obtained through the fusion algorithm. Based on the fusion value and the noise field distribution, an initial three-dimensional noise field model is obtained through Gaussian fitting.

[0142] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0143] A three-dimensional physical environment model is established through the initial three-dimensional noise field model and the environmental signal. Based on the three-dimensional physical environment model, the three-dimensional physical environment model after lightweight processing is obtained through lightweight processing. Through the model fusion algorithm, the initial three-dimensional noise field model and the three-dimensional physical environment model after lightweight processing are fused to obtain the three-dimensional noise field model.

[0144] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0145] Collect the noise signal to be measured and the environmental signal to be measured, and input the noise signal to be measured and the environmental signal to be measured into the three-dimensional noise field model to obtain the noise field distribution result corresponding to the noise signal to be measured and the environmental signal to be measured.

[0146] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

[0147] Collect the noise signal and the environmental signal;

[0148] According to the noise signal, through the principal component analysis method, obtain the principal component frequency sound field and the principal component frequency sound field diagram of the noise signal;

[0149] According to the principal component frequency sound field, through the MUSIC algorithm, obtain the noise field distribution;

[0150] According to the noise field distribution and the principal component frequency sound field diagram, through the fusion algorithm and Gaussian fitting, obtain the initial three-dimensional noise field model;

[0151] Fuse the initial three-dimensional noise field model and the environmental signal to obtain the three-dimensional noise field model.

[0152] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0153] According to the noise signal, through frequency domain analysis, obtain the frequency domain data of the noise signal. According to the frequency domain data, through principal component analysis, obtain the principal component frequency sound field and the principal component frequency sound field diagram of the noise signal.

[0154] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0155] According to the principal component frequency sound field, through frame processing and Fourier transform, obtain the signal subspace and the noise subspace. According to the orthogonality of the signal subspace and the noise subspace, obtain the spatial spectrum. According to the spatial spectrum, obtain the noise field distribution.

[0156] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0157] Obtain the peak value of the spectral peak of the spatial spectrum. According to the principal component frequency sound field diagram and the peak value of the spectral peak of the spatial spectrum, through the fusion algorithm, obtain the fusion value of the peak value of the spectral peak of the spatial spectrum. Through the fusion value and the noise field distribution, through Gaussian fitting, obtain the initial three-dimensional noise field model.

[0158] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0159] Establish a three-dimensional physical environment model through the initial three-dimensional noise field model and environmental signals. According to the three-dimensional physical environment model, through lightweight processing, obtain the lightweight processed three-dimensional physical environment model. Through the model fusion algorithm, fuse the initial three-dimensional noise field model and the lightweight processed three-dimensional physical environment model to obtain the three-dimensional noise field model.

[0160] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0161] Collect the noise signal to be measured and the environmental signal to be measured, input the noise signal to be measured and the environmental signal to be measured into the three-dimensional noise field model, and obtain the noise field distribution result corresponding to the noise signal to be measured and the environmental signal to be measured.

[0162] In one embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0163] Collect the noise signal and the environmental signal;

[0164] According to the noise signal, through the principal component analysis method, obtain the principal component frequency sound field and the principal component frequency sound field diagram of the noise signal;

[0165] According to the principal component frequency sound field, through the MUSIC algorithm, obtain the noise field distribution;

[0166] According to the noise field distribution and the principal component frequency sound field diagram, through the fusion algorithm and Gaussian fitting, obtain the initial three-dimensional noise field model;

[0167] Fuse the initial three-dimensional noise field model and the environmental signal to obtain the three-dimensional noise field model.

[0168] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0169] According to the noise signal, through frequency domain analysis, obtain the frequency domain data of the noise signal, and according to the frequency domain data, through principal component analysis, obtain the principal component frequency sound field and the principal component frequency sound field diagram of the noise signal.

[0170] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0171] According to the principal component frequency sound field, through frame processing and Fourier transform, obtain the signal subspace and the noise subspace, according to the orthogonality of the signal subspace and the noise subspace, obtain the spatial spectrum, and according to the spatial spectrum, obtain the noise field distribution.

[0172] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0173] Obtain the peak value of the spatial spectrum. According to the main component frequency sound field diagram and the peak value of the spatial spectrum, through a fusion algorithm, obtain the fusion value of the peak value of the spatial spectrum. Through the fusion value and the noise field distribution, perform Gaussian fitting to obtain the initial three-dimensional noise field model.

[0174] In one embodiment, when the computer program is executed by the processor, the following steps are further implemented:

[0175] Establish a three-dimensional physical environment model through the initial three-dimensional noise field model and the environmental signal. According to the three-dimensional physical environment model, through lightweight processing, obtain the lightweight processed three-dimensional physical environment model. Through the model fusion algorithm, fuse the initial three-dimensional noise field model and the lightweight processed three-dimensional physical environment model to obtain the three-dimensional noise field model.

[0176] In one embodiment, when the computer program is executed by the processor, the following steps are further implemented:

[0177] Collect the noise signal to be measured and the environmental signal to be measured, input the noise signal to be measured and the environmental signal to be measured into the three-dimensional noise field model, and obtain the noise field distribution result corresponding to the noise signal to be measured and the environmental signal to be measured.

[0178] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0179] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memories. Non-volatile memories can include read-only memory (ROM), magnetic tapes, floppy disks, flash memories, optical memories, high-density embedded non-volatile memories, resistive random access memories (ReRAM), magnetoresistive random access memories (MRAM), ferroelectric random access memories (FRAM), phase change memories (PCM), graphene memories, etc. Volatile memories can include random access memory (RAM) or external cache memories, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The databases involved in the embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in the present application can be general-purpose processors, central processors, graphics processors, digital signal processors, programmable logics, data processing logics based on quantum computing, etc., without limitation.

[0180] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope recorded in this specification.

[0181] The above-described embodiments merely represent several implementation manners of the present application. The description is relatively specific and detailed, but it should not be construed as a limitation on the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A method for noise field modeling, characterized in that, The method includes: Collecting a noise signal and an environmental signal; the noise signal is generated by equipment that generates noise in a substation; According to the noise signal, obtaining the principal component frequency sound field and the principal component frequency sound field diagram of the noise signal through principal component analysis; According to the principal component frequency sound field, obtaining the noise field distribution through the MUSIC algorithm; According to the noise field distribution and the principal component frequency sound field diagram, obtaining an initial three-dimensional noise field model through a fusion algorithm and Gaussian fitting; Fusing the initial three-dimensional noise field model and the environmental signal to obtain a three-dimensional noise field model; The obtaining the noise field distribution according to the principal component frequency sound field through the MUSIC algorithm includes: According to the principal component frequency sound field, obtaining a signal subspace and a noise subspace through frame division processing and Fourier transform; According to the orthogonality of the signal subspace and the noise subspace, obtaining a spatial spectrum; According to the spatial spectrum, obtaining the noise field distribution; The obtaining the signal subspace and the noise subspace according to the principal component frequency sound field through frame division processing and Fourier transform includes: Performing frame division processing on the principal component frequency sound field to obtain the frame-divided principal component frequency sound field; Performing short-time Fourier transform on the frame-divided principal component frequency sound field to obtain the covariance matrix of the principal component frequency sound field in the time domain and the frequency domain; Performing eigenvalue decomposition on the covariance matrix to obtain the signal subspace and the noise subspace of the principal component frequency sound field; The obtaining the noise field distribution according to the spatial spectrum includes: Searching for the spectral peaks of the spatial spectrum respectively according to the search range of the azimuth signal parameters, finding the value of the azimuth signal parameter corresponding to the maximum value of the spectral peak peak of the spatial spectrum, and using the obtained value of the azimuth signal parameter as the noise field distribution.

2. The noise field modeling method according to claim 1, wherein The obtaining the principal component frequency sound field and the principal component frequency sound field diagram of the noise signal according to the noise signal through principal component analysis includes: According to the noise signal, obtaining the frequency domain data of the noise signal through frequency domain analysis; According to the frequency domain data, obtaining the principal component frequency sound field and the principal component frequency sound field diagram of the noise signal through principal component analysis.

3. The noise field modeling method according to claim 1, characterized in that The obtaining the initial three-dimensional noise field model according to the noise field distribution and the principal component frequency sound field diagram through a fusion algorithm and Gaussian fitting includes: Obtaining the spectral peak peak of the spatial spectrum; According to the principal component frequency sound field diagram and the spectral peak peak of the spatial spectrum, obtaining the fusion value of the spectral peak peak of the spatial spectrum through a fusion algorithm; Through the fusion value and the noise field distribution, obtaining an initial three-dimensional noise field model through Gaussian fitting.

4. The noise field modeling method according to claim 1, wherein The fusing the initial three-dimensional noise field model and the environmental signal to obtain a three-dimensional noise field model includes: Establishing a three-dimensional physical environment model through the initial three-dimensional noise field model and the environmental signal; According to the three-dimensional physical environment model, obtaining a lightweight processed three-dimensional physical environment model through lightweight processing; Through a model fusion algorithm, fusing the initial three-dimensional noise field model and the lightweight processed three-dimensional physical environment model to obtain a three-dimensional noise field model.

5. The noise field modeling method according to claim 1, wherein The method further includes: Collect the noise signal to be measured and the environmental signal to be measured; Input the noise signal to be measured and the environmental signal to be measured into the three-dimensional noise field model to obtain the noise field distribution result corresponding to the noise signal to be measured and the environmental signal to be measured.

6. A noise field modeling device, characterized in that The device includes: A signal acquisition module for collecting a noise signal and an environmental signal; the noise signal is generated by equipment that generates noise in a substation; A principal component analysis module for obtaining the principal component frequency sound field and the principal component frequency sound field diagram of the noise signal through principal component analysis according to the noise signal; A noise field distribution acquisition module for obtaining the noise field distribution through the MUSIC algorithm according to the principal component frequency sound field; A noise field model acquisition module for obtaining an initial three-dimensional noise field model through a fusion algorithm and Gaussian fitting according to the noise field distribution and the principal component frequency sound field diagram; A model fusion module for fusing the initial three-dimensional noise field model and the environmental signal to obtain a three-dimensional noise field model; The noise field distribution acquisition module is further configured to obtain a signal subspace and a noise subspace through frame processing and Fourier transform according to the principal component frequency sound field; obtain a spatial spectrum according to the orthogonality of the signal subspace and the noise subspace; and obtain the noise field distribution according to the spatial spectrum; The noise field distribution acquisition module is further configured to perform frame processing on the principal component frequency sound field to obtain the framed principal component frequency sound field; perform short-time Fourier transform on the framed principal component frequency sound field to obtain the covariance matrix of the principal component frequency sound field in the time domain and the frequency domain; and perform eigenvalue decomposition on the covariance matrix to obtain the signal subspace and the noise subspace of the principal component frequency sound field; The noise field distribution acquisition module is further configured to search for the spectral peaks of the spatial spectrum respectively according to the search range of the azimuth signal parameters, find the value of the azimuth signal parameter corresponding to the maximum value of the spectral peak of the spatial spectrum, and use the obtained value of the azimuth signal parameter as the noise field distribution.

7. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, the steps of the method according to any one of claims 1 to 5 are implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, the steps of the method according to any one of claims 1 to 5 are implemented.

9. A computer program product comprising a computer program, characterized in that, When the computer program is executed by the processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images