Method and apparatus for processing noise data, and electronic device and vehicle

Abstract

The present application relates to the technical field of data processing, and particularly relates to a method and apparatus for processing noise data, and an electronic device and a vehicle. The method comprises: firstly acquiring a target frequency corresponding to a spectrum component to be determined; determining a Nyquist frequency corresponding to the target frequency, wherein the target frequency is lower than a preset frequency threshold value; then, according to the Nyquist frequency, determining a time-domain sampling number R required by the target frequency; and selecting, from among a displacement sequence, a target sequence corresponding to the first R sampling points, and performing butterfly recursive computation of a fast Fourier transform on the target sequence, wherein an obtained result is used for determining sound pressure level information corresponding to the spectrum component, and the displacement sequence is obtained by means of performing butterfly odd-even displacement of a fast Fourier transform according to a time-domain sequence corresponding to the spectrum component. By means of applying the technical solution of the present application, for the analysis and computation of low-frequency noise, the computation redundancy can be reduced, and the computing power is saved on while ensuring that a precision error is within an acceptable range.

Description

Noise data processing method, device, electronic equipment and vehicle

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority to the Chinese patent application filed with the China Patent Office on June 20, 2023, with application number 202310731400.9 and application name “Method, device, electronic device and vehicle for processing low-frequency noise data”, the entire contents of which are incorporated by reference into this application. Technical Field

[0003] The present application relates to the field of data processing technology, and relates to, but is not limited to, a method, device, electronic equipment, and vehicle for processing noise data. Background Art

[0004] The noise data generated by the operation of equipment such as motors or engines in vehicles can be used to analyze whether the equipment has operating abnormalities.

[0005] Currently, when measuring both high-frequency and low-frequency noise data from a device, the most accurate algorithm is required. This means that the Nyquist sampling theorem for the highest frequency of these frequencies must be met. For example, the same sampling points (derived from the Nyquist sampling theorem for the highest frequency) are used for analysis and calculation of both high-frequency and low-frequency noise data.

[0006] However, this method does not require so many sampling points for the analysis and calculation of low-frequency noise data, which will cause computational redundancy and waste computing power.

[0007] Summary of the Invention

[0008] In view of this, the present application provides a noise data processing method, device, electronic device and vehicle, the main purpose of which is to improve the technical problem of the analysis and calculation of low-frequency noise data in the current related technology, which causes computational redundancy and wastes computing power.

[0009] In a first aspect, the present application provides a method for processing noise data, comprising:

[0010] Obtaining a target frequency corresponding to a spectral component to be determined; the frequency of the spectral component is lower than a preset frequency threshold;

[0011] Determine the Nyquist frequency corresponding to the target frequency;

[0012] According to the Nyquist frequency, determine the number of time domain samples R required for the target frequency;

[0013] A target sequence corresponding to the first R sampling points is selected from the permutation sequence, and a butterfly recursive calculation of a fast Fourier transform is performed on the target sequence. The result is used to determine the sound pressure level information corresponding to the spectral component, wherein the permutation sequence is obtained by performing a butterfly parity permutation of a fast Fourier transform on the time domain sequence corresponding to the spectral component.

[0014] In a second aspect, the present application provides a noise data processing device, comprising:

[0015] An acquisition part is configured to acquire a target frequency corresponding to the spectral component to be determined; the target frequency is lower than a preset frequency threshold;

[0016] a determining portion configured to determine a Nyquist frequency corresponding to a target frequency;

[0017] The determining part is further configured to determine the number R of time domain samples required for the target frequency based on the Nyquist frequency;

[0018] The calculation part is configured to select a target sequence corresponding to the first R sampling points in the permutation sequence, and perform butterfly recursive calculation of fast Fourier transform on the target sequence, and the obtained result is used to determine the sound pressure level information corresponding to the spectral component, wherein the permutation sequence is obtained by performing butterfly parity permutation of fast Fourier transform on the time domain sequence corresponding to the spectral component.

[0019] In a third aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the noise data processing method described in the first aspect.

[0020] In a fourth aspect, the present application provides an electronic device comprising a memory for storing a computer program; and a processor for executing the computer program stored in the memory to implement the noise data processing method described in the first aspect.

[0021] In a fifth aspect, the present application provides a vehicle comprising the device as described in the second aspect, or the electronic device as described in the fourth aspect.

[0022] In a sixth aspect, the present application provides a computer program product, comprising computer executable instructions or a computer program, which, when executed by a processor, implements the noise data processing method described in the first aspect.

[0023] By means of the above technical solution, the present application provides a method, device, electronic device and vehicle for processing noise data. Compared with the current related technologies, the present application can save computing power for the analysis and calculation of low-frequency noise. First, the target frequency corresponding to the spectral component to be determined is obtained; the Nyquist frequency corresponding to the target frequency is determined; then, based on the Nyquist frequency, the number of time domain samples R required for the target frequency is determined; the target sequence corresponding to the first R sampling points in the permutation sequence is selected to perform butterfly recursive calculation of fast Fourier transform, and the result obtained is used to determine the sound pressure level information corresponding to the spectral component, wherein the permutation sequence is obtained by performing butterfly parity permutation of fast Fourier transform on the time domain sequence corresponding to the spectral component. This solution is equivalent to downsampling the frequency domain data according to the number of time domain samples R that meet the Nyquist frequency, and then obtaining a sound pressure level calculation result that can meet the calculation accuracy requirements through butterfly recursive calculation of fast Fourier transform. By applying the technical solution of the present application, for the analysis and calculation of low-frequency noise, while ensuring that the accuracy error is within an acceptable range, the calculation redundancy is reduced and computing power is saved.

[0024] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0026] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0027] FIG1 is a schematic diagram showing a flow chart of a method for processing noise data provided in an embodiment of the present application;

[0028] FIG2 shows a schematic diagram 1 of an example of a butterfly parity shift provided by an embodiment of the present application;

[0029] FIG3 shows a second schematic diagram of an example of a butterfly parity shift provided by an embodiment of the present application;

[0030] FIG4 is a flow chart showing another method for processing noise data provided by an embodiment of the present application;

[0031] FIG5A is a schematic diagram showing sound pressure level information corresponding to the butterfly recursive calculation method of the fast Fourier transform provided in an embodiment of the present application and the current butterfly recursive calculation method of the fast Fourier transform;

[0032] FIG5B is a schematic diagram showing the difference in sound pressure levels between two types of sound pressure level information in FIG5A ;

[0033] FIG6 is a schematic diagram showing an exemplary display effect provided by an embodiment of the present application;

[0034] FIG7 shows a schematic structural diagram of a noise data processing device provided in an embodiment of the present application;

[0035] FIG8 shows a schematic structural diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0036] In order to more clearly understand the above-mentioned objectives, features and advantages of the present application, the scheme of the present application will be further described below. It should be noted that, in the absence of conflict, the embodiments of the present application and the features therein can be combined with each other.

[0037] In order to improve the current related technologies for analyzing and calculating low-frequency noise data, which will cause computational redundancy and waste of computing power. The present application can be applied to noise measurement scenarios. For example, it can be applied to application scenarios where high-frequency noise and low-frequency noise are measured simultaneously. This embodiment provides a method for processing noise data, as shown in Figure 1, the method includes:

[0038] Step 101: Obtain a target frequency corresponding to a spectral component to be determined.

[0039] A spectrum is a representation of a time-domain signal in the frequency domain. It can be obtained by performing a Fourier transform on the signal. The resulting graphs are plotted with amplitude and phase on the vertical axis and frequency on the horizontal axis. However, phase information is sometimes omitted, leaving only the amplitude data corresponding to different frequencies. The "amplitude spectrum" is sometimes used to show how amplitude varies with frequency, while the "phase spectrum" shows how phase varies with frequency. Simply put, a spectrum shows which frequencies of sinusoidal waves make up a signal, and also reveals information such as the magnitude and phase of each frequency.

[0040] Sound is a mechanical vibration that can travel through matter in all its states. These substances that can propagate sound are called media. Sound cannot propagate in a vacuum. The sound we hear is also a sound wave with a certain frequency. The human ear's frequency range is approximately 20 to 20,000 Hz; frequencies outside this range are imperceptible to the human ear. Sounds below 20 Hz are considered infrasound, while sounds above 20 kHz are considered ultrasonic. The higher the frequency, the higher the pitch, while the lower the frequency, the lower the pitch.

[0041] In the embodiment of the present application, the target frequency refers to the frequency of the noise data corresponding to the spectral component to be determined. The target frequency is lower than the preset frequency threshold. In some embodiments, the spectral component to be determined

[0042] In some embodiments, the spectrum containing the spectral components to be determined may also include other spectral components that are higher than a preset frequency threshold. That is to say, in the same spectrum, for spectral components that are higher than the preset frequency threshold, such as spectral components of high-frequency noise, an algorithm that meets high precision can be used, such as the Nyquist sampling theorem that meets the highest frequency in the spectral components that are higher than the preset frequency threshold, for analysis and calculation. For spectral components in the spectrum that are lower than the preset frequency threshold, that is, spectral components to be determined, such as spectral components of low-frequency noise, the noise data processing method in the embodiment of the present application can be used for analysis and calculation. In this way, non-uniform FFT calculations can be used in the same spectrum to meet spectrum requirements while reducing computing power requirements and improving algorithm complexity under the same computing power.

[0043] Step 102: Determine the Nyquist frequency corresponding to the target frequency.

[0044] In the embodiment of the present application, the Nyquist frequency is the minimum sampling frequency that needs to be defined to prevent signal aliasing. Aliasing refers to the phenomenon that when the discrete Fourier transform is used to perform frequency domain sampling on the signal domain, the number of sampling points is less than the time domain sequence length, causing the time domain periodic extension sequences to overlap with each other.

[0045] The sampling frequency is set to avoid overlap. If the sampling frequency is too low, the sampled result will differ from the original sample. If a sample's spectrum is band-limited (zeros outside a certain frequency), the sampling frequency must be greater than twice the current frequency to prevent overlap and distortion. This requirement is known as the Nyquist criterion, or the Nyquist sampling theorem.

[0046] Downsampling is the process of reducing the sampling rate of a specific signal, typically to reduce data transmission rate or data size. The downsampling factor (often represented by the symbol M) is generally an integer or rational number greater than 1. This factor expresses how many times longer the sampling period has become, or equivalently, how much the sampling rate has been reduced to a fraction of the original value. Because downsampling reduces the sampling rate, it is necessary to ensure that the Nyquist sampling theorem still holds at the new lower sampling rate.

[0047] In the embodiment of the present application, the Nyquist frequency corresponding to the target frequency is determined, that is, the minimum sampling frequency needs to be defined to prevent signal aliasing at the target frequency.

[0048] By determining the minimum sampling frequency, the minimum number of samples in the sample set required for downsampling at the current target frequency, that is, the minimum sampling number, can be determined.

[0049] Step 103: Determine the number R of time domain samples required for the target frequency based on the Nyquist frequency.

[0050] In an embodiment of the present application, the number of time domain samples is determined by determining the number of time domain samples required to calculate the frequency corresponding to the current spectral component based on the Nyquist frequency, wherein the number of time domain signals corresponds to the number of frequency domain signals.

[0051] The time domain is the real world, the only domain that actually exists. When evaluating the performance of digital products, analysis is usually performed in the time domain because that is where the product's performance is ultimately measured.

[0052] The frequency domain is particularly common in radio frequency and communication systems, and is also encountered in high-speed digital applications. Its most important property is that it is not real, but a mathematical construct. While the time domain is the only objective domain, the frequency domain is a mathematical domain that follows specific rules. Some scholars have even referred to the frequency domain as the "God's perspective."

[0053] The time domain represents dynamic signal relationships using time as the coordinate, while the frequency domain represents signals using frequency as the coordinate. Generally speaking, the time domain is more vivid and intuitive, while the frequency domain is more concise, allowing for deeper and more convenient analysis. Currently, signal analysis in the time and frequency domains is interconnected, each indispensable and complementary.

[0054] Step 104: Select a target sequence corresponding to the first R sampling points in the displacement sequence, and perform butterfly recursive calculation of fast Fourier transform on the target sequence. The obtained result is used to determine the sound pressure level information corresponding to the spectral component.

[0055] The permutation sequence is obtained by performing butterfly parity permutation of fast Fourier transform on the time domain sequence corresponding to the spectral components.

[0056] Fast Fourier Transform (FFT), also known as Fast Fourier Transform, is a fast algorithm for Discrete Fourier Transform (DFT). It is obtained by improving the DFT algorithm based on the odd, even, imaginary, and real characteristics of DFT.

[0057] Fast Fourier transform is based on DFT, using point-valued polynomial representation and the unit root property to perform recursive calculation of polynomial parity decomposition to accelerate the algorithm. The premise of recursion is to rearrange the data of the original sequence according to the butterfly parity distribution.

[0058] In an embodiment of the present application, the time domain sequence corresponding to the spectral component contains 16 data, such as data 0-15, the number of time domain samples R required for the target frequency is 8, and the minimum number of samples in the sample set required for downsampling at the current target frequency is 8. As shown in Figure 2, the time domain sequence [0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15] corresponding to the spectral components of the 16 data is subjected to butterfly parity permutation to obtain the permutation sequence [0 8 4 12][2 10 6 14][1 9 5 13][3 11 7 15]. If the permutation sequence [0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15] is downsampled to 8, the new sequence is [0 2 4 6 8 10 12 14]. Performing butterfly parity permutation on the new sequence [0 2 4 6 8 10 12 14], as shown in Figure 3, yields the new permuted sequence [0 8 4 12][2 10 6 14]. It can be seen that the first R (R = 8) numbers after performing butterfly parity permutation on the original sequence (the time-domain sequence corresponding to the spectral components) are the permuted sequence (new permutation sequence) that results from downsampling the original sequence of N (N = 16) numbers to a sequence of K numbers (new sequence). This is determined by the properties of butterfly parity permutation.

[0059] In other words, after permuting the original sequence, the first R numbers are recursively computed using the Fast Fourier Transform (FFT) butterfly algorithm to obtain a sequence that matches the minimum number of samples in the sample set. This yields the spectrum corresponding to the downsampled sequence. Furthermore, the sequence obtained by first padding with zeros and then permuting remains consistent with the sequence obtained by first downsampling, then padding with zeros, and then permuting. This property of the butterfly parity permutation ensures that only a single permutation calculation is required for the non-uniform FFT. Furthermore, the spectrum corresponding to the downsampled sequence can be used to determine the sound pressure level information corresponding to the spectral components.

[0060] Compared with the current related technologies, this embodiment can be used for the analysis and calculation of low-frequency noise, which can save computing power. This embodiment first obtains the target frequency corresponding to the spectral component to be determined; determines the Nyquist frequency corresponding to the target frequency; then, based on the Nyquist frequency, obtains the number of time domain samples R required for the target frequency; selects the target sequence corresponding to the first R sampling points in the permutation sequence for butterfly recursive calculation of the fast Fourier transform, and the result is used to determine the sound pressure level information corresponding to the spectral component, wherein the permutation sequence is obtained by performing butterfly parity permutation of the fast Fourier transform on the time domain sequence corresponding to the spectral component. Thus, this embodiment can accurately analyze and calculate low-frequency noise data with a small amount of sampling point data. This embodiment is equivalent to downsampling the frequency domain data according to the number of time domain samples R that meets the Nyquist frequency, and then performing butterfly recursive calculation of the fast Fourier transform to obtain a sound pressure level calculation result that meets the calculation accuracy requirements. By applying the technical solution of this embodiment, for the analysis and calculation of low-frequency noise, while ensuring that the accuracy error is within an acceptable range, computational redundancy is reduced and computing power is saved.

[0061] Furthermore, as a refinement and extension of the above embodiment, this embodiment provides a method as shown in FIG4 , which includes:

[0062] Step 201: Obtain a target frequency corresponding to a spectral component to be determined.

[0063] Optionally, the target frequency corresponding to the spectrum component to be determined may be any frequency, taking the obtained target frequency as 16 Hz as an example.

[0064] Step 202: Determine the Nyquist frequency corresponding to the target frequency.

[0065] In some embodiments of the present application, the calculation formula for obtaining the Nyquist frequency may be as shown in Formula 1 below: NQST =2f s (Formula 1)

[0066] In formula (1), f s Indicates the target frequency, f NQST Indicates the Nyquist frequency corresponding to the target frequency. For example, based on step 201, if the target frequency is 8 Hz, the target frequency is multiplied by 2, and the resulting Nyquist frequency is 16 Hz. If the target frequency is 5 Hz, the resulting Nyquist frequency is 10 Hz. Other cases can be deduced by analogy, and no specific examples are given here.

[0067] Step 203: Obtain the number R of time domain samples required for the target frequency according to the Nyquist frequency.

[0068] Optionally, step 203 may include: determining, according to the Nyquist frequency, a minimum number of samples for downsampling the spectral components as the number R of time-domain samples required for the target frequency.

[0069] Exemplarily, step 203 further includes: using a first formula to calculate the number of time domain samples R. In some embodiments, the first formula is as shown in the following formula 2:

[0070] Among them, f NQST represents the Nyquist frequency, Indicates rounding up.

[0071] The rounding function refers to the largest integer that does not exceed the real number x, which is called the integer part of x and is denoted as [x] or INT(x).

[0072] In the embodiments of this application, Indicates rounding up, that is, the result is the smallest integer greater than the current real number. For example, when When the content in the brackets is 1.6, take the smallest integer greater than 1.6, which is 2. When the content in the brackets is 1.1, take the smallest integer greater than 1.1, which is 2. The same applies to other cases, and no specific examples are given here.

[0073] In some embodiments of the present application, the number of time domain samples R must be a positive integer, and the corresponding Nyquist frequency must be an integer power of 2. If the Nyquist frequency is not an integer power of 2, it is necessary to take the logarithm of the Nyquist frequency and round it up.

[0074] For example, if the Nyquist frequency is 200 Hz, which is greater than 2 to the 7th power and less than 2 to the 8th power, it needs to be rounded up to get R as 8. It can also be understood that the Nyquist frequency needs to be rounded up to an integer power of 2. If the Nyquist frequency is 200 Hz, it needs to be rounded up to 256 Hz, and R can be obtained as 8.

[0075] For example, based on step 202, when the obtained Nyquist frequency is 16 Hz, the obtained number of time domain samples R is 4. Similarly, when the obtained Nyquist frequency is 10 Hz, the obtained number of time domain samples R is 4. Other situations can be deduced by analogy and no specific examples are given here.

[0076] Step 204: Obtain the number of recursive layers corresponding to the number R of time domain samples.

[0077] In the embodiment of the present application, the logarithm with base 2 is taken for the number of time domain samples R to obtain the number of recursive layers, which is log2(R).

[0078] Exemplarily, based on step 203 , the number of time domain samples R obtained is 4, and the corresponding number of recursive layers is 2.

[0079] Step 205: Perform butterfly recursive calculation of fast Fourier transform on the target sequence corresponding to the first R sampling points according to the number of recursive layers, and use the obtained results to determine the sound pressure level information corresponding to the spectral components.

[0080] Optionally, step 205 may include: based on the polynomial of the layer where the target sequence is located, successively deriving and calculating the upper layer polynomial according to the number of recursive layers, and finally obtaining the polynomial as the butterfly recursive calculation result of the fast Fourier transform.

[0081] Optionally, step 205 further includes: starting from the layer where the target sequence is located, performing butterfly recursive calculation of the fast Fourier transform using the second formula according to the number of recursive layers. In some embodiments, the second formula is shown in the following formula 3:

[0082] Wherein, k represents the spectral component, Y k,m+1 i represents the i-th polynomial of the m+1-th layer, Y k,m 2i-1 represents the 2i-1th polynomial of the mth layer, represents the unit root extracted at the mth level in the recursive calculation, Y k,m 2i Represents the 2i-th polynomial of the m-th layer, where m is in the range [0, log2(R)-1].

[0083] Optionally, step 205 further includes: processing the butterfly recursive calculation result by the time domain sampling number R to obtain sound pressure level information corresponding to the spectral component.

[0084] In some embodiments, step 205 further includes: obtaining a calculation result obtained by butterfly recursive calculation of fast Fourier transform; and calculating the calculation result using a third formula to obtain sound pressure level information corresponding to the spectral component. The third formula is shown in the following formula 4:

[0085] Among them, Y k Represents the sound pressure level information corresponding to the spectral component, Y k,log2(R) 1 represents the calculation result obtained by butterfly recursive calculation of fast Fourier transform.

[0086] Optionally, according to Parsval's theorem, after a sequence of N samples is subjected to a fast Fourier butterfly parity transform, in order to satisfy energy conservation, the result needs to be divided by 2 / N. However, the fast Fourier butterfly parity transform is not performed uniformly, and different spectra use different sequence sampling numbers. After obtaining the spectrum, it is not possible to uniformly divide it by 2 / N to obtain the same energy amplitude. Therefore, when calculating each frequency component, it is divided separately by the sampled sequence length, that is, R.

[0087] Exemplarily, based on step 204, when the number of recursive layers is 2, the number of time domain samples R is 4, and a butterfly recursive calculation of the fast Fourier transform is performed on the target sequence corresponding to the first 4 sampling points. Since the number of recursive layers is 2, it is necessary to perform two recursive calculations on the target sequence to obtain a polynomial of the target sequence two layers above. The obtained polynomial is then multiplied by the number of recursive layers and divided by the length of the sampling sequence. The result is used to determine the sound pressure level information corresponding to the spectral component.

[0088] The sound pressure level is defined as the common logarithm of the ratio of the measured sound pressure p to the reference sound pressure p(ref), multiplied by 20, that is: 20lg p / p(ref), and its unit is decibel.

[0089] In the embodiment of the present application, each spectral component has corresponding sound pressure level information, and the sound pressure level information of multiple spectral components can be aggregated to obtain spectral information corresponding to the sound pressure level information of noise data.

[0090] In some embodiments, the spectral components to be determined include: spectral components corresponding to the noise of the target device; after determining the sound pressure level information corresponding to the spectral components, the following further includes:

[0091] An operating status of the target device is determined according to the sound pressure level information.

[0092] Exemplarily, the target device may include a motor or engine in a vehicle. By collecting noise data generated by the motor or engine in the vehicle during operation and analyzing the spectral components in the noise data using the noise data processing method of this embodiment, sound pressure level information can be obtained, thereby determining the operating status of the motor or engine in the vehicle, such as whether there is any operating abnormality.

[0093] Furthermore, taking the extraction of second-order energy from engine noise within 10 seconds as an example, FIG5A shows the sound pressure level information corresponding to the butterfly recursive calculation method (non-uniform FFT) of the fast Fourier transform provided in the embodiment of the present application, and the current butterfly recursive calculation method of the fast Fourier transform. As shown in FIG5B, the total sound pressure level difference between the two butterfly recursive calculation methods of the fast Fourier transform is within plus or minus 1dB, and the root mean square error is 0.116dB. It can be seen that the non-uniform FFT algorithm of the embodiment of the present application can meet the accuracy requirements.

[0094] Computing power saving explanation: The amount of computation required for the butterfly recursive calculation of the fast Fourier transform of a conventional sequence of length N = 2M is shown in the following formula 5: P0 = N log2(N) = NM (Formula 5)

[0095] Among them, M is the number of recursive layers and N is the sequence length.

[0096] It can be seen that the butterfly recursive calculation of the conventional fast Fourier transform of a sequence of length N=2M requires Nlog2(N) multiplications. The computing power P required by the butterfly recursive calculation of the non-uniform fast Fourier transform is shown in the following formula 6:

[0097] The amount of computation that can be saved is calculated using the following formula 7:

[0098] As the length of N increases from 2 13 to 2 20 The computing power that can be saved by changing is shown in Figure 6.

[0099] By applying the solution of this embodiment, the calculation accuracy error is within 1dB, and the sequence length is within 2 13 to 2 20 Changes, saving 5%-7.5% of computing power.

[0100] In summary, compared with the current related technologies, this embodiment can save computing power while ensuring that the accuracy error is within an acceptable range for the analysis and calculation of low-frequency noise.

[0101] Furthermore, this embodiment provides the following examples, including:

[0102] First, the sequence is subjected to a fast Fourier butterfly parity shift. The Nyquist frequency that satisfies the sampling theorem is calculated based on the data to be extracted. The number of time domain samples required to calculate the corresponding frequency of the current spectral component is determined based on the Nyquist frequency. In the shifted sequence, the first R sampling points are selected, and the selected sequence is subjected to a butterfly recursive calculation. The number of recursive layers is log2(R). The two parity polynomials pointed to by the arrows in the next layer are used to calculate the polynomial at the root of the arrows in the previous layer: the final calculation result of the k-th component is obtained. According to Pasval's theorem, after the fast Fourier butterfly parity transform of the N-sample sequence, in order to meet the energy conservation requirement, the result needs to be divided by 2 / N. However, the number of sequence samples used for different spectra of the non-uniform fast Fourier butterfly parity shift is different. After obtaining the spectrum, it is not possible to uniformly divide it by 2 / N to obtain the same energy amplitude. Therefore, when calculating each frequency component, it is divided by the sampled sequence length to obtain the final k-th component of the spectrum.

[0103] Compared to existing related technologies, this embodiment downsamples the frequency domain data based on the number R of time domain samples required to meet the Nyquist frequency, and then uses a butterfly-like recursive fast Fourier transform to obtain a sound pressure level calculation result that meets the requirements for computational accuracy. This embodiment can save computing power while ensuring acceptable accuracy errors in the analysis and calculation of low-frequency noise.

[0104] Furthermore, this embodiment provides a noise data processing device, as shown in FIG7 , which includes: an acquisition part 31 , a determination part 32 , and a calculation part 33 .

[0105] The acquisition part 31 is configured to acquire a target frequency corresponding to the spectral component to be determined; the target frequency is lower than a preset frequency threshold;

[0106] a determining portion 32 configured to determine a Nyquist frequency corresponding to a target frequency;

[0107] The determining part 32 is further configured to determine the number R of time domain samples required for the target frequency based on the Nyquist frequency;

[0108] The calculation part 33 is configured to select a target sequence corresponding to the first R sampling points in the permutation sequence, and perform butterfly recursive calculation of fast Fourier transform on the target sequence, and the obtained result is used to determine the sound pressure level information corresponding to the spectral component, wherein the permutation sequence is obtained by performing butterfly parity permutation of fast Fourier transform on the time domain sequence corresponding to the spectral component.

[0109] In some application scenarios, the determining part 32 is further configured to determine, based on the Nyquist frequency, a minimum number of samples for downsampling the spectral components as the number R of time-domain samples required for the target frequency.

[0110] In some application scenarios, the determining part 32 is further configured to calculate the number R of time domain samples using a first formula, where the first formula includes: Among them, f NQST represents the Nyquist frequency, Indicates rounding up.

[0111] In some application scenarios, the calculation part 33 is further configured to obtain the number of recursive layers corresponding to the number R of time domain samples; and perform butterfly recursive calculation of fast Fourier transform on the target sequence according to the number of recursive layers.

[0112] In some application scenarios, the calculation part 33 is also configured to derive and calculate the upper-layer polynomial based on the polynomial of the layer where the target sequence is located according to the number of recursive layers, and the final polynomial is used as the butterfly recursive calculation result of the fast Fourier transform.

[0113] In some application scenarios, the calculation part 33 is further configured to start from the layer where the target sequence is located and perform butterfly recursive calculation of the fast Fourier transform according to the number of recursive layers using a second formula, where the second formula includes: Wherein, k represents the spectral component, Y k,m+1 i represents the i-th polynomial of the m+1-th layer, Y k,m 2i-1 represents the 2i-1th polynomial of the mth layer, represents the unit root extracted at the mth level in the recursive calculation, Y k,m 2i Represents the 2i-th polynomial of the m-th layer, where m is in the range [0, log2(R)-1].

[0114] In some application scenarios, the calculation part 33 is further configured to process the butterfly recursive calculation result through the time domain sampling number R to obtain the sound pressure level information corresponding to the spectral component.

[0115] In some application scenarios, the calculation part 33 is further configured to obtain a calculation result obtained by butterfly recursive calculation of fast Fourier transform; and calculate the calculation result through a third formula to obtain sound pressure level information corresponding to the spectral component, wherein the third formula includes: Among them, Y k Represents the sound pressure level information corresponding to the spectral component, Y k,log2(R) 1 represents the calculation result obtained by butterfly recursive calculation of fast Fourier transform.

[0116] It should be noted that, for other corresponding descriptions of the functional units involved in the noise data processing device provided in this embodiment, reference may be made to the exemplary corresponding descriptions in FIG. 1 to FIG. 4 , which will not be repeated here.

[0117] Based on the above-mentioned method shown in Figures 1 to 4, accordingly, this embodiment also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the noise data processing method provided in this application.

[0118] Based on this understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, USB flash drive, mobile hard disk, etc.), and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the noise data processing method of each implementation scenario of the present application.

[0119] Based on the noise data processing method provided in this application and the noise data processing device embodiment provided in this application, in order to achieve the above-mentioned purpose, the embodiment of the present application also provides an electronic device that can be configured on the vehicle side (such as a new energy vehicle) or the server side, etc. As shown in Figure 8, the electronic device 3 includes a memory 32 and a processor 33; wherein the memory 32 and the processor 33 are connected via a communication bus 34; the memory 32 is used to store computer programs; the processor 33 is used to execute the computer program stored in the memory to implement any one of the low-frequency noise data processing methods provided in the embodiments of the present application.

[0120] Optionally, the physical device may further include a user interface, a network interface, a camera, a radio frequency (RF) circuit, a sensor, an audio circuit, a Wi-Fi module, and the like. The user interface may include a display, an input unit such as a keyboard, and the like. Optional user interfaces may also include a USB interface, a card reader interface, and the like. The network interface may optionally include a standard wired interface, a wireless interface (such as a Wi-Fi interface), and the like.

[0121] Those skilled in the art will understand that the above-mentioned physical device structure provided in this embodiment does not constitute a limitation on the physical device, and may include more or fewer components, or a combination of certain components, or different component arrangements.

[0122] The memory may also include an operating system and a network communication component. The operating system is a program that manages the hardware and software resources of the physical device, supporting the execution of information processing programs and other software and / or programs. The network communication component is used to enable communication between components within the memory, as well as with other hardware and software in the physical information processing device.

[0123] Furthermore, based on the embodiment of the noise data processing method provided in this application, the embodiment of the noise data processing device provided in this application, and the above-mentioned electronic device embodiment, this embodiment also provides a vehicle, which may include the above-mentioned electronic device, and further may execute the noise data processing method provided in this application, exemplarily, as shown in Figures 1 to 4.

[0124] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present application can be implemented by means of software plus the necessary general hardware platform, or by hardware. By applying the solution of this embodiment, compared with the current related art, this embodiment can save computing power for the analysis and calculation of low-frequency noise. First, the target frequency corresponding to the spectral component to be determined is obtained; the Nyquist frequency corresponding to the target frequency is determined; then, based on the Nyquist frequency, the number of time domain samples R required for the target frequency is obtained; the target sequence corresponding to the first R sampling points in the permutation sequence is selected for butterfly recursive calculation of fast Fourier transform, and the result obtained is used to determine the sound pressure level information corresponding to the spectral component, wherein the permutation sequence is obtained by performing butterfly parity permutation of fast Fourier transform on the time domain sequence corresponding to the spectral component, and thus this embodiment can accurately perform analysis and calculation of low-frequency noise data with a small amount of sampling point data. This embodiment is equivalent to downsampling the frequency domain data according to the number of time domain samples R that meets the Nyquist frequency, and then performing butterfly recursive calculation of fast Fourier transform to obtain a sound pressure level calculation result that can meet the calculation accuracy requirements. By applying the technical solution of this embodiment, for the analysis and calculation of low-frequency noise, it is possible to reduce calculation redundancy and save computing power while ensuring that the accuracy error is within an acceptable range.

[0125] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprises" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a..." do not exclude the presence of other identical elements in the process, method, article or device that includes the elements.

[0126] The foregoing is merely a list of specific embodiments of the present application, intended to enable those skilled in the art to understand and implement the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application is not limited to the embodiments described herein, but is intended to conform to the broadest scope consistent with the principles and novel features of the present application. Industrial Applicability

[0127] The present application provides a method, device, electronic device and vehicle for processing noise data, which can save computing power for the analysis and calculation of low-frequency noise. First, the target frequency corresponding to the spectral component to be determined is obtained; the Nyquist frequency corresponding to the target frequency is determined; the target frequency is lower than the preset frequency threshold; then, based on the Nyquist frequency, the number of time domain samples R required for the target frequency is determined; the target sequence corresponding to the first R sampling points in the displacement sequence is selected to perform butterfly recursive calculation of fast Fourier transform, and the result obtained is used to determine the sound pressure level information corresponding to the spectral component, wherein the displacement sequence is obtained by performing butterfly parity displacement of fast Fourier transform on the time domain sequence corresponding to the spectral component. The present application is equivalent to downsampling the frequency domain data according to the number of time domain samples R that meet the Nyquist frequency, and then obtaining a sound pressure level calculation result that can meet the calculation accuracy requirements through butterfly recursive calculation of fast Fourier transform. By applying the technical solution of the present application, for the analysis and calculation of low-frequency noise, the accuracy error can be guaranteed to be within an acceptable range, and the calculation redundancy is reduced, saving computing power.

Claims

1. A method for processing noise data, comprising: Obtaining a target frequency corresponding to a spectral component to be determined; The target frequency is lower than a preset frequency threshold; determining a Nyquist frequency corresponding to the target frequency; Determine the number of time domain samples R required for the target frequency according to the Nyquist frequency; A target sequence corresponding to the first R sampling points is selected in the displacement sequence, and a butterfly recursive calculation of a fast Fourier transform is performed on the target sequence, and the obtained result is used to determine the sound pressure level information corresponding to the spectral component, wherein the displacement sequence is obtained by performing a butterfly parity displacement of a fast Fourier transform on a time domain sequence corresponding to the spectral component.

2. The method according to claim 1, wherein: Determining the number of time domain samples R required for the target frequency according to the Nyquist frequency includes: According to the Nyquist frequency, the minimum number of samples for downsampling the spectrum components is determined as the number R of time domain samples required for the target frequency.

3. The method according to claim 2, wherein: The step of determining, based on the Nyquist frequency, a minimum number of samples for downsampling the spectral components as the number of time domain samples R required for the target frequency includes: The time domain sampling number R is calculated using a first formula, wherein the first formula includes: Among them, f NQST represents the Nyquist frequency, Indicates rounding up.

4. The method according to claim 1, wherein: The target sequence corresponding to the first R sampling points is selected in the displacement sequence to perform butterfly recursive calculation of fast Fourier transform, including: Obtaining the number of recursive layers corresponding to the number R of time domain samples; According to the number of recursive layers, a butterfly recursive calculation of fast Fourier transform is performed on the target sequence.

5. The method according to claim 4, wherein: The butterfly recursive calculation of fast Fourier transform of the target sequence according to the number of recursive layers includes: Based on the polynomial of the layer where the target sequence is located, the upper layer polynomial is derived and calculated successively according to the number of recursive layers, and the polynomial finally obtained is used as the butterfly recursive calculation result of the fast Fourier transform.

6. The method according to claim 5, wherein: The polynomial based on the layer where the target sequence is located, deriving and calculating the upper layer polynomial one by one according to the number of recursive layers, includes: Starting from the layer where the target sequence is located, according to the number of recursive layers, a butterfly recursive calculation of fast Fourier transform is performed using a second formula, wherein the second formula includes: Wherein, k represents the spectral component, Y k,m+1 i represents the i-th polynomial of the m+1-th layer, Y k,m 2i-1 Indicates the mth The 2i-1th polynomial of the layer, represents the unit root extracted at the mth level in the recursive calculation, Y k,m 2i represents the 2i-th polynomial of the m-th layer, where m is in the range of [0, log2(R)-1].

7. The method according to claim 5, wherein: The method further comprises: The butterfly recursive calculation result is processed by the time domain sampling number R to obtain the sound pressure level information corresponding to the spectrum component.

8. The method according to claim 7, wherein: The step of processing the butterfly recursive calculation result by the time domain sampling number R to obtain the sound pressure level information corresponding to the spectral component includes: The butterfly recursive calculation result is calculated by a third formula to obtain the sound pressure level information corresponding to the spectral component, and the third formula includes: Among them, Y k represents the sound pressure level information corresponding to the spectral component, Y k,log2(R) 1 represents the butterfly recursive calculation result.

9. A noise data processing device, comprising: An acquisition part, configured to acquire a target frequency corresponding to a spectral component to be determined; The target frequency is lower than a preset frequency threshold; a determining part configured to determine a Nyquist frequency corresponding to the target frequency; The determination part is further configured to determine the number of time domain samples R required for the target frequency according to the Nyquist frequency; The calculation part is configured to select a target sequence corresponding to the first R sampling points in the displacement sequence, and perform butterfly recursive calculation of fast Fourier transform on the target sequence, and the result obtained is used to determine the sound pressure level information corresponding to the spectral component, wherein the displacement sequence is obtained by performing butterfly parity displacement of fast Fourier transform on the time domain sequence corresponding to the spectral component.

10. A computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the method according to any one of claims 1 to 8 when executed by a processor.

11. An electronic device, comprising: Memory for storing computer programs; A processor, configured to implement the method according to any one of claims 1 to 8 when executing the computer program stored in the memory.

12. A vehicle comprising: The device as claimed in claim 9, or the electronic device as claimed in claim 11.

13. A computer program product, comprising computer executable instructions or a computer program, wherein the computer executable instructions or the computer program, when executed by a processor, implements the noise data processing method according to any one of claims 1 to 8.

Images