Simulation method for in-vehicle noise during constant-speed driving of vehicle

Through careful processing and synthesis of noise signals in the vehicle while driving at a constant speed, a signal sample that is consistent with the actual vehicle experience is generated, which solves the problem that the total sound pressure level in the prior art cannot accurately reflect the noise perception in the vehicle, and realizes the development of noise in the vehicle Quick and accurate formulation and optimization of goals.

CN118936620AActive Publication Date: 2024-11-12CHINA FAW CO LTD
View PDF 6 Cites 0 Cited by

Patent Information

Application Number
CN202411188397.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-28
Publication Date
2024-11-12
Estimated Expiration
2044-08-28

AI Technical Summary

Technical Problem

The prior art uses the total sound pressure level to evaluate the noise in a car that is driving at a constant speed, and cannot accurately reflect the sound experience of different spectrum distributions, and cannot reflect the impact of narrowband signals and continuous impact signals on the hearing, resulting in the evaluation indicators that are inconsistent with the actual feeling and are difficult to support. High-level sound quality development.

Method used

By collecting the noise signal of the real vehicle at a constant speed, performing frequency range division and FFT analysis, generating white noise signals, bandpass filtering and sound pressure calibration, synthesizing broadband, narrowband and impact noise signals, simulating the noise in the car, and forming a playback evaluation that can be used for playback evaluation signal sample.

Benefits of technology

The detailed parameterized adjustment of noise in the vehicle is achieved by driving at a constant speed, and a signal sample is generated consistent with the actual vehicle experience, which is used to formulate noise development goals and optimization plans to quickly and accurately meet the sound quality requirements.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118936620A_ABST
    Figure CN118936620A_ABST
Patent Text Reader

Abstract

The invention discloses a method for simulating noise in an automobile during constant-speed driving of the automobile. The method comprises the following steps: randomly generating a first white noise signal, and performing band-pass filtering and amplitude calibration on the first white noise signal through a plurality of continuous frequency bands to obtain a broadband noise simulation signal; randomly generating a second white noise signal, and performing narrowband filtering and narrowband amplitude calibration to obtain a narrowband noise analog signal; randomly generating a third white noise signal, filtering the third white noise signal to generate a band-pass filtering signal, generating an integral window function according to an in-vehicle noise signal sample acquired by a real vehicle, performing integral windowing on the band-pass filtering signal, and generating an impact noise analog signal through impact amplitude calibration; adding the broadband noise analog signal, the narrowband noise analog signal and the impact noise analog signal to obtain an analog signal of the noise in the vehicle running at a constant speed for playback listening evaluation; according to the invention, the simulation of the in-vehicle noise during the constant-speed driving of the vehicle is realized, the playback effect restores the in-vehicle hearing feeling during the constant-speed driving of a real vehicle, parameters can be quickly adjusted to change samples, playback evaluation can be carried out on different samples, the samples meeting the sound quality development requirements are selected, and the quick and accurate formulation of the in-vehicle noise development target during the constant-speed driving is realized.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of vehicle noise simulation, and in particular to a method for simulating the noise inside a vehicle when the vehicle is traveling at a constant speed. Background Art

[0002] The control of automobile interior noise is an important aspect in the field of automobile research and development. Among them, the constant speed driving interior noise index mainly reflects the wind noise and road noise control level of the whole vehicle, and also corresponds to the usage scenario that users perceive the most. In the past, the total sound pressure level was mainly used to evaluate the constant speed driving interior noise, but there is a significant problem that sound signals with different spectral distributions will bring completely different auditory feelings, even if they have exactly the same total sound pressure level, and the total sound pressure level cannot reflect the impact of narrowband signals and continuous impact signals on auditory feelings. Therefore, in practice, the quality of the total sound pressure level index of some vehicles often does not accurately correspond to the quality of their actual auditory feelings. With the development and improvement of the level of the automobile industry, the single indicator of total sound pressure level to evaluate the constant speed driving interior noise is not enough to support the high-level sound quality development and control of automobiles. It is necessary to start from the human ear auditory perception and formulate more accurate and detailed constant speed interior noise targets, such as the 1 / 3 octave band distribution of interior noise, the distribution of narrowband noise such as tire cavity noise, and the main frequency range of road impact noise. The above problems need to be solved urgently. Summary of the invention

[0003] The purpose of the present invention is to provide a method for simulating the interior noise of a car traveling at a constant speed, so as to solve the problems raised in the above-mentioned background technology. The frequency, amplitude, etc. of various components in the signal can be carefully parameterized and adjusted, so as to realize the rapid generation of a large number of signal samples. Automobile R&D personnel can play back and evaluate the signal samples simulated and generated by the method, which can be used to formulate development goals for the interior noise of a car traveling at a constant speed, or to determine the influence of the changes of various components in the signal samples on the auditory perception, and then determine the optimization direction.

[0004] To achieve the above object, the present invention provides the following technical solution: a method for simulating the noise inside a car when the car is traveling at a constant speed, comprising:

[0005] Collect the interior noise signal samples of the real car running at a constant speed, determine the frequency range of the simulated signal, divide the frequency range into multiple continuous frequency bands, calculate the average sound pressure of the interior noise signal samples in each frequency band, the average FFT spectrum result, and the FFT-time time-frequency analysis result;

[0006] A first white noise signal is randomly generated, and the ratio of the average sound pressure of the in-vehicle noise signal sample in each frequency band to the average sound pressure of the first white noise signal in each frequency band is calculated to obtain the sound pressure amplitude calibration coefficient of each frequency band; the first white noise signal is band-pass filtered in each frequency band, and then the sound pressure amplitude is calibrated by the corresponding sound pressure amplitude calibration coefficient, and then superimposed to obtain a broadband noise simulation signal;

[0007] According to the FFT spectrum average result of the in-vehicle noise signal, the frequency lower limit, frequency upper limit and amplitude of multiple narrowband components are determined; the narrowband amplitude calibration coefficient is set according to the amplitude of each narrowband component; a second white noise signal is randomly generated, the second white noise signal is filtered according to the frequency lower limit and frequency upper limit of each narrowband component, and the amplitude is calibrated by the narrowband amplitude calibration coefficient and then superimposed to obtain a narrowband noise simulation signal;

[0008] Based on the FFT-time frequency analysis results of the in-vehicle noise signal, the lower frequency limit, upper frequency limit, density, single impact time and impact amplitude of multiple road impact noise components are respectively determined; the impact amplitude calibration coefficient of the corresponding road impact noise component is set according to the impact amplitude;

[0009] Randomly generate a third white noise signal, and perform band-pass filtering on the third white noise signal according to the lower frequency limit and the upper frequency limit of the impulse noise component to generate a band-pass filtered signal having the same or similar frequency range as the impulse noise;

[0010] An overall window function of the impulse noise is generated according to the density of the impulse noise component and the single impulse time. After the bandpass filter signal is overall windowed by the overall window function, the amplitude is calibrated by the impulse amplitude calibration coefficient to generate an impulse noise simulation signal.

[0011] The broadband noise simulation signal, the narrowband noise simulation signal, and the impact noise simulation signal are added together to obtain a simulation signal of the noise inside the car traveling at a constant speed. The simulation signal is saved in a format that can be played by electronic equipment for playback and listening evaluation.

[0012] Furthermore, a sample of a noise signal inside a real vehicle traveling at a constant speed is collected, a frequency range of a simulated signal is determined, and the frequency range is divided into a plurality of continuous frequency bands, further comprising:

[0013] The frequency range of the analog signal is determined by the lower limit of the starting frequency band and the upper limit of the ending frequency band, or is determined by actual needs;

[0014] Divide the analog signal frequency range into contiguous frequency bands using integer or fractional octave bands or fixed bandwidth bands.

[0015] Further, a first white noise signal is randomly generated, and the ratio of the average sound pressure of the in-vehicle noise signal sample in each frequency band to the average sound pressure of the first white noise signal in each frequency band is calculated to obtain the sound pressure amplitude calibration coefficient of each frequency band; the first white noise signal is band-pass filtered in each frequency band, and then the sound pressure amplitude is calibrated by the corresponding sound pressure amplitude calibration coefficient, and then superimposed to obtain a broadband noise simulation signal, which also includes:

[0016] A one-to-one corresponding wideband pass filter is set according to multiple frequency bands, and the first white noise signal is passed through each wideband pass filter to obtain multiple wideband pass signals. Each wideband pass signal is amplitude calibrated by a corresponding sound pressure amplitude calibration coefficient and then superimposed to obtain a wideband noise simulation signal.

[0017] Further, according to the FFT spectrum average result of the in-vehicle noise signal, the frequency lower limit, frequency upper limit and amplitude of multiple narrowband components are determined; the narrowband amplitude calibration coefficient is set according to the amplitude of each narrowband component; a second white noise signal is randomly generated, the second white noise signal is filtered according to the frequency lower limit and frequency upper limit of each narrowband component, and the amplitude is calibrated by the narrowband amplitude calibration coefficient and then superimposed to obtain a narrowband noise simulation signal, and also includes:

[0018] The number of the multiple narrowband components is determined by the number of main narrowband components in the FFT spectrum average result of the in-vehicle noise signal and the simulation accuracy requirement;

[0019] Design narrow-band bandpass filters based on the lower and upper frequency limits of each narrow-band component;

[0020] The second white noise signal passes through corresponding narrowband bandpass filters respectively to obtain narrowband filtered signals of multiple narrowband components;

[0021] A plurality of narrowband filter signals are respectively calibrated with their corresponding narrowband amplitude calibration coefficients and then superimposed to obtain a narrowband noise simulation signal.

[0022] Furthermore, a third white noise signal is randomly generated, and the third white noise signal is band-pass filtered according to the lower frequency limit and the upper frequency limit of the impulse noise component to generate a band-pass filtered signal having the same or similar frequency range as the impulse noise, further comprising:

[0023] A bandpass filter is set according to the lower frequency limit and the upper frequency limit of the impulse noise component, and the third white noise is bandpass filtered to generate a bandpass filtered signal with the same or similar frequency range as the impulse noise.

[0024] Further, an overall window function of the impulse noise is generated, and after the bandpass filter signal is overall windowed by the overall window function, the amplitude is calibrated by the impulse amplitude calibration coefficient to generate an impulse noise simulation signal, which also includes:

[0025] Set the impulse unit window function for each impulse noise component;

[0026] The time length of the impact unit window function is the same as the single impact time;

[0027] Set the window function value at the start and end time of the impact unit window function to zero;

[0028] The window function is second-order differentiable in intervals other than the start and end times.

[0029] Further, including:

[0030] The impact signal is composed of a series of evenly and discretely distributed impact units;

[0031] The number of shocks in the simulation time is obtained by multiplying the shock density by the total time length of the simulation sample;

[0032] Generates a random number that is discretely distributed uniformly over the simulation duration based on the number of shocks.

[0033] Further, including:

[0034] The impact unit window function is processed by summing the impact number delay times to obtain the overall window function, where the delay time is a random number uniformly and discretely distributed within the simulation duration;

[0035] The impulse noise simulation signal is generated by multiplying the bandpass filtered signal with the overall window function and then multiplying it by the corresponding impulse amplitude calibration coefficient.

[0036] Further, including:

[0037] Multiply the bandpass filtered signal by the impact unit window function to obtain a single impact unit;

[0038] The delay time of the impact number is summed up for each impact unit to obtain the impact number impact units with uniform discrete distribution, where the delay time is a random number with uniform discrete distribution within the simulation duration;

[0039] After a single impact unit is multiplied by the corresponding impact amplitude calibration coefficient, the impact noise simulation signal is obtained by superposition and summation.

[0040] Furthermore, the broadband noise simulation signal, the narrowband noise simulation signal, and the impact noise simulation signal are added together to obtain a simulation signal of the noise inside the vehicle traveling at a constant speed, and the simulation signal is saved in a format that can be played by an electronic device for playback and listening evaluation, which also includes:

[0041] The synthesized uniform speed vehicle interior noise simulation signal has the energy of the frequency band where the added signal components are located increased due to the addition of the narrowband noise simulation signal and the impact noise simulation signal. The increased frequency band energy information is fed back to restore the energy of each frequency band of the synthesized uniform speed vehicle interior noise simulation signal to the original sample level.

[0042] Wherein, restoring the energy of each frequency band of the synthesized uniform speed vehicle interior noise simulation signal to the original sample level also includes:

[0043] By correcting the sound pressure amplitude calibration coefficient, the corresponding frequency band energy in the broadband noise simulation signal is reduced, and the enhanced energy information is fed back. Through multiple iterations, the energy of each frequency band of the synthesized uniform speed vehicle interior noise simulation signal is restored to the original sample level.

[0044] By adjusting the sound pressure amplitude calibration coefficient, the narrowband amplitude calibration coefficient and the impact amplitude calibration coefficient, the components of the simulated synthetic uniform speed vehicle interior noise are changed to generate different simulation samples;

[0045] The various samples generated after adjustment were played back for listening evaluation.

[0046] Compared with the prior art, the beneficial effects of the present invention are as follows: the signal processing method is used to construct three signal components, namely broadband noise, narrowband noise and road impact noise, respectively. The three signal components are superimposed to generate the noise inside the car when the vehicle is moving at a constant speed, and the characteristic parameters of the three signal components can be adjusted separately; the method can realize the simulation of the noise inside the car when the vehicle is moving at a constant speed, and the simulated noise signal samples inside the car when the vehicle is moving at a constant speed can be used for playback evaluation, and the playback effect restores the auditory experience inside the car when the real vehicle is moving at a constant speed. The method can quickly adjust the parameters to change the samples, and perform playback evaluation on different samples, and can quickly select samples that meet the requirements of sound quality development, so as to realize the rapid and accurate formulation of the development goals of the noise inside the car when the vehicle is moving at a constant speed; it can also analyze and judge the influence of different sound components on the auditory experience by comparing the changing trends of the auditory experience of the signal samples brought about by the changes of different parameters, and provide a reliable basis for formulating the optimization plan for the noise inside the car when the vehicle is moving at a constant speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 This is a flow chart of a method for simulating the noise inside a vehicle traveling at a constant speed according to an embodiment of the present invention;

[0048] Figure 2 Schematic diagram of the 1 / 3 octave frequency band of the noise inside the car when the car is traveling at a constant speed in an embodiment of the present invention, where (a) is unweighted and (b) is A-weighted;

[0049] Figure 3 Schematic diagram of the 1 / 3 octave result of the first white noise signal in an embodiment of the present invention, wherein (a) is unweighted and (b) is A-weighted;

[0050] Figure 4 Schematic diagram of amplitude response of a 1 / 3 octave filter in an embodiment of the present invention;

[0051] Figure 5 This is a schematic diagram of a time domain waveform of a 1 / 3 octave bandpass signal in an embodiment of the present invention;

[0052] Figure 6 This is a schematic diagram of a 1 / 3 octave comparison between a broadband noise simulation signal and a real vehicle collected signal sample in an embodiment of the present invention;

[0053] Figure 7 This is a schematic diagram of the average analysis results of FFT spectra of broadband noise simulation samples and original samples in an embodiment of the present invention;

[0054] Figure 8 Schematic diagram of a narrowband filter determined by main peak components of original samples in an embodiment of the present invention;

[0055] Fig. 9 It is a schematic diagram of superimposing a narrow-band bandpass filter signal to simulate the narrow-band component in the real vehicle noise in an embodiment of the present invention;

[0056] Fig.10 This is a schematic diagram of FFT-time time-frequency analysis of a real vehicle noise signal in an embodiment of the present invention;

[0057] Fig.11 Schematic diagram of the generation of the impact unit of the impact signal component 1 in the embodiment of the present invention, wherein (a) is the bandpass filtered signal 1, (b) is the time window 1, and (c) is the windowed bandpass filtered signal 1;

[0058] Fig.12 Schematic diagram of the generation of the impact unit of the impact signal component 2 in an embodiment of the present invention, wherein (a) is the bandpass filtered signal 2, (b) is the time window 2, and (c) is the windowed bandpass filtered signal 2;

[0059] Fig.13 A schematic diagram of each impact unit for generating a randomly distributed impact signal component 2 according to impact density in an embodiment of the present invention;

[0060] Fig.14 Schematic diagram of a superimposed impact signal 2 generated in an embodiment of the present invention;

[0061] Fig.15 A schematic diagram of each impact unit for generating a randomly distributed impact signal component 1 according to the impact density in an embodiment of the present invention;

[0062] Fig.16 A schematic diagram of an impact signal 1 generated by superposition in an embodiment of the present invention;

[0063] Fig.17 Schematic diagram of FFT spectrum average result of impulse signal in an embodiment of the present invention;

[0064] Fig.18 Schematic diagram of FFT-time time-frequency analysis results of an impulse signal in an embodiment of the present invention;

[0065] Fig.19 1 / 3 octave analysis schematic diagram of the composite signal Sig_combine in an embodiment of the present invention, wherein (a) is before amplitude correction, and (b) is after amplitude correction;

[0066] Fig. 20 1 / 3 octave analysis schematic diagram of the composite signal Sig_combine in an embodiment of the present invention, wherein (a) is before amplitude correction, and (b) is after amplitude correction;

[0067] Fig.21 Schematic diagram of FFT-time time-frequency analysis verification in an embodiment of the present invention, where (a) is the noise collected from a real vehicle, and (b) is the simulated synthetic noise signal Sig_combine;

[0068] Fig. 22 This is a schematic diagram of the FFT spectrum average result of the optimized samples in an embodiment of the present invention;

[0069] Fig.23 This is a schematic diagram of the 1 / 3 octave average results of the optimized samples in an embodiment of the present invention;

[0070] Fig.24 Schematic diagram of the comparison of FFT-time time-frequency analysis results of the simulated optimized samples in the embodiment of the present invention, where (a) is the noise collected from the real vehicle and (b) is the simulated optimized sample. DETAILED DESCRIPTION

[0071] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0072] In the description of the present invention, unless otherwise specified, "plurality" means two or more than two; the orientations or positional relationships indicated by the terms "upper", "lower", "left", "right", "inner", "outer", "front end", "rear end", "head", "tail", etc. are based on the orientations or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", "third", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0073] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "connected" and "connection" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0074] Please refer to the attached drawings of the specification, the present invention provides a technical solution: Figure 1 As shown, a method for simulating the noise inside a car when the car is moving at a constant speed comprises the following steps:

[0075] S102, collecting interior noise signal samples of a real vehicle traveling at a constant speed, determining the frequency range of the simulation signal, dividing the frequency range into a plurality of continuous frequency bands, and calculating the average sound pressure of the interior noise signal samples in each frequency band, the average result of the FFT spectrum, and the FFT-time time-frequency analysis result;

[0076] Specifically, the analog signal frequency range is determined by the lower limit of the starting frequency band and the upper limit of the ending frequency band, or is determined by actual needs; the analog signal frequency range is divided into multiple continuous frequency bands using integer or fractional octave bands or fixed bandwidth bands.

[0077] Among them, preferably, in this embodiment, the frequency range of the analog signal is determined to be 17.5Hz to 17959.4Hz; the frequency range of the analog signal is divided into 30 continuous 1 / 3 octave frequency bands; the selected 1 / 3 octave center frequency is 20Hz~16kHz, with a total of 30 1 / 3 octave frequency bands, and the frequency range of the analog signal that can be generated is 17.5Hz to 17959.4Hz, covering the normal auditory frequency range of the human ear and the main components of the noise inside the vehicle; the frequency range division method using other integer or fractional octave bands or fixed bandwidth bands, the analog signal frequency range determined by the lower limit of the starting frequency band and the upper limit of the ending frequency band or determined by actual needs, and the corresponding number of frequency bands generated by any of the above division methods within the frequency range are also within the protection scope of the present invention.

[0078] S104, randomly generating a first white noise signal, calculating the ratio of the average sound pressure of the in-vehicle noise signal sample in each frequency band to the average sound pressure of the first white noise signal in each frequency band, and obtaining the sound pressure amplitude calibration coefficient of each frequency band; performing band-pass filtering on the first white noise signal in each frequency band, and then calibrating the sound pressure amplitude by the corresponding sound pressure amplitude calibration coefficient, and then superimposing to obtain a broadband noise simulation signal;

[0079] Specifically, a one-to-one corresponding wideband pass filter is set according to multiple frequency bands, the first white noise signal is passed through each band pass filter to obtain a wideband pass signal, and each wideband pass signal is amplitude calibrated by a corresponding sound pressure amplitude calibration coefficient and then superimposed to obtain a wideband noise simulation signal.

[0080] Among them, the steady-state average sound pressure of each of the 30 1 / 3 octave frequency bands of the interior noise signal sample of the actual vehicle running at a constant speed is calculated, denoted as: p_car n , where n = 1:30; calculate the steady-state average sound pressure of each of the 1 / 3 octave frequency bands of the first white noise signal wh itenoise_130, denoted as p_w n , where n = 1:30; the 1 / 3 octave band sound pressure of the interior noise of the actual vehicle traveling at a constant speed is divided by the 1 / 3 octave band sound pressure of the first white noise to obtain the amplitude calibration coefficient, recorded as A_wb n , then the amplitude calibration coefficient calculation formula is: A_wb n =p_car n / p_w n , where n = 1:30; the conversion relationship between the sound pressure level and the sound pressure calculated by the amplitude calibration coefficient is: L = 20*log 10 (p 2 / p0 2 ) or p = 10 L / 20 *p0, where: L is the sound pressure level, p is the sound pressure, the reference sound pressure p0 = 2*10 -5Unit: Pa; Design 30 filters based on 1 / 3 octave bandpass range: Hd_wb n , where n = 1:30; the first white noise is passed through the above 30 1 / 3 octave filters to obtain 30 1 / 3 octave bandpass signals Sig_wb n , the calculation formula is: Sig_wb n =filter(Hd_wb n ,wh itenoise_1),n=1:30; After the amplitude of each frequency band signal is calibrated, the broadband noise simulation signal is superimposed and synthesized, which is recorded as: Sig_wb_total. The calculation formula of the broadband noise simulation signal is:

[0081] For example, Figure 2 and Figure 3 As shown in the figure, the interior noise of a real car driving at a constant speed is collected. The interior noise of a certain car model driving at a constant speed of 50km / h on a rough road is taken as an example, and 10 seconds of steady-state noise data is taken; the 1 / 3 octave average result is calculated, where A weighting is part of the national standard, which can reflect the auditory frequency response characteristics of the human ear and is widely used in environmental noise measurement. The unit is dB(A).

[0082] The calculation results of this example are shown in Table 1:

[0083] Table 1: Center frequency, frequency range, and sound pressure level, sound pressure and amplitude calibration coefficients of the interior noise and the first white noise of a real vehicle traveling at a constant speed for each 1 / 3 octave band

[0084]

[0085]

[0086] Design 30 filters based on 1 / 3 octave bandpass range Hd_wb n , n = 1:30, this application example uses a 46-order infinite impulse response elliptic filter for design, and the amplitude response of each 1 / 3 octave filter is as follows Figure 4 As shown, the 1 / 3 octave filter amplitude response; 30 1 / 3 octave bandpass signals Sig_wb n The time domain waveform is as follows Figure 5 As shown in the figure, in order to observe the waveform, the x-axis range in the figure only takes 0.1s; the synthesized broadband noise simulation signal is compared with the real vehicle collected signal sample by A-weighted 1 / 3 octave, and the results are as follows Figure 6 As shown in the figure, the simulation results of the sound pressure level in each 1 / 3 octave band are basically consistent with the original samples collected from the actual vehicle;

[0087] Thirty 1 / 3 octave bandpass filters are used to filter the white noise separately to synthesize a broadband noise analog signal. The frequency distribution characteristics and sound volume of the noise can be adjusted by changing the amplitude coefficient of each 1 / 3 octave noise.

[0088] S106, determining the lower frequency limit, upper frequency limit and amplitude of multiple narrowband components according to the FFT spectrum average result of the in-vehicle noise signal; setting a narrowband amplitude calibration coefficient according to the amplitude of each narrowband component; randomly generating a second white noise signal, filtering the second white noise signal according to the lower frequency limit and upper frequency limit of each narrowband component, and calibrating the amplitude by the narrowband amplitude calibration coefficient and then superimposing to obtain a narrowband noise simulation signal;

[0089] Specifically, the number of multiple narrowband components is determined by the number of main narrowband components in the FFT spectrum averaging result of the in-vehicle noise signal and the simulation accuracy requirements; a bandpass filter is designed according to the lower frequency limit and the upper frequency limit of each narrowband component; the second white noise signal passes through the bandpass filter respectively to obtain narrowband filtered signals of multiple narrowband components; the multiple narrowband filtered signals are respectively calibrated with their corresponding narrowband amplitude calibration coefficients and then superimposed to obtain a narrowband noise simulation signal.

[0090] First, determine the narrowband signal parameters, observe the FFT spectrum average results of the actual vehicle test signal, and determine the frequency lower limit of the main narrowband component: fl_nb n , upper frequency limit: fu_nb n , amplitude: A_nb n , n = 1: N, where N is determined by the number of main narrowband components in the average result of the FFT spectrum of the actual vehicle noise; secondly, set a narrowband filter and design a bandpass filter based on the lower and upper frequency limits of each peak component: Hd_nb n , where n=1:N; finally, a narrowband noise analog signal is generated, and a second white noise is randomly generated, denoted as whitenoise_2, and N narrowband signals are obtained through the above narrowband filters respectively, and the calculation formula is: Sig_nb n =filter(Hd_nb n ,wh itenoise_2), where n=1:N; the amplitude of each narrowband component in the actual vehicle noise sample is set as the narrowband amplitude calibration coefficient, recorded as A_nb(n), and multiplied and summed with Sig_nb(n) respectively, and the superposition result of the narrowband signal is the narrowband noise simulation signal, and its calculation formula is:

[0091] It is understandable that the broadband noise generated by the superposition simulation of 1 / 3 octave bandpass signals can be consistent with the original samples in each 1 / 3 octave band, but it cannot simulate the prominent frequency components in the original samples. The FFT spectrum average analysis of the broadband noise simulation samples in this example and the original samples collected from the real vehicle shows that there are several frequency peaks in the original vehicle signal, and the simulation results cannot match them, such as Figure 7 As shown; observe and identify the frequency ranges of several main peak components in the original sample, as shown in Table 2:

[0092] Table 2 Main peak components of original samples

[0093] Serial number Frequency lower limit / Hz Upper frequency limit / Hz 1 81 92 2 205 210 3 365 400 4 1080 1180 5 1783 1800

[0094] Design a bandpass filter based on the frequency range of each peak component in the above table, such as Figure 8 As shown, Figure 8 A narrowband filter is determined by the main peak component of the original sample; a second white noise is randomly generated, and 5 narrowband signals are obtained by passing through the above narrowband filter respectively. The formula is:

[0095] Sig_nb n =filter(Hd_nb n ,wh itenoise_2),n=1:N

[0096] In this example, narrowband simulation is performed on the five peaks of the original sample, so N is set to 5 in this example.

[0097] A_nb(n) is set according to the amplitude of each narrowband component in the real vehicle noise sample, and multiplied and summed with Sig_nb(n) to obtain the superposition result of the narrowband signal. The formula is:

[0098]

[0099] The narrowband synthesis results are as follows Fig. 9 As shown, the narrow-band bandpass filter signal is superimposed on the narrow-band component of the simulated real vehicle noise;

[0100] A bandpass filter with a freely set passband frequency range is used to filter white noise and synthesize a narrowband noise analog signal. The frequency characteristics of the narrowband noise can be adjusted by changing the upper and lower frequency limits of the bandpass filter and the filter order. The size of the narrowband noise can be adjusted by changing the amplitude coefficient.

[0101] S108, confirming the frequency lower limit, frequency upper limit, density, single impact time and impact amplitude of multiple road impact noise components respectively based on the FFT-time time-frequency analysis result of the vehicle interior noise signal; setting the impact amplitude calibration coefficient of the corresponding road impact noise component according to the impact amplitude;

[0102] Among them, the simulation parameters of the impact signal are determined, and the frequency lower limit fl_pu of the road impact noise component is confirmed by the FFT-time time-frequency analysis results of the real vehicle collected signal n 、Frequency upper limit fu_pu n , density dense_pu n (i.e., the frequency of impacts per unit time), single impact time pu_time n and the amplitude A_pu n , where n=1:M, and M is determined by the number of road impact noise categories that need to be simulated.

[0103] It is understandable that the FFT-time frequency analysis of the real vehicle noise signal is performed, such as Fig.10 As shown, the frequency range, density (i.e., the frequency of impacts per unit time), duration, and amplitude of the road impact noise components are confirmed. In this example, two impact signals of different intensities are generated to simulate the continuous impact noise of a real vehicle driving on a rough road, and the occasional strong impact noise. The parameters of the two impact signals are shown in Table 3;

[0104] Table 3 Impact signal simulation parameters

[0105]

[0106] Generate a randomly distributed impact signal according to the impact density. Multiply the impact density by the total sample duration to get the total number of impacts. Generate the time delay of all impact signals by generating random numbers. In this example, a sample of 10s length is to be generated. As shown in Table 3, the frequency of the impact noise component 2 is 3 times / s, which means that a total of 30 impact units need to be generated. Copy the windowed bandpass filtered signal 2 30 times, and then generate 30 random numbers uniformly and discretely distributed within 10s as the time delay of each impact unit, and obtain the distribution of all impact units, such as Fig.13 and Fig.14 Similarly, the impulse noise component 1 will generate 4 impulse units within 10s, and the 4 impulse units are delayed respectively by 4 random numbers uniformly and discretely distributed within 10, and the distribution of the impulse units of the impulse noise component 1 is as follows: Fig.15 , 16 As shown; the impact signal component 1 and the impact signal component 2 are multiplied by the corresponding amplitude coefficients and then added to synthesize the impact signal. The FFT spectrum average result of the impact signal is as follows: Fig.17 As shown; the FFT-time spectrum analysis results of the impact signal are as follows Fig.18 As shown, it can be seen that the synthesized impact signal can well simulate the road impact noise component of the samples collected from the real vehicle in terms of time-frequency characteristics.

[0107] S110, randomly generating a third white noise signal, and performing band-pass filtering on the third white noise signal according to the frequency lower limit and the frequency upper limit of the impulse noise component to generate a band-pass filtered signal having the same or similar frequency range as the impulse noise;

[0108] Specifically, a bandpass filter is set according to the lower frequency limit and the upper frequency limit of the impulse noise component, and the third white noise is bandpass filtered to generate a bandpass signal with the same frequency range as the impulse noise.

[0109] Among them, filtering generates a bandpass signal and randomly generates a third white noise whitenoise_3. The bandpass filter Hd_pu is designed based on the lower and upper frequency limits of the impulse noise component. n , and the third white noise is band-pass filtered to generate a band-pass signal with the same frequency range as the impulse noise. The calculation formula is:

[0110] Sig_bp n =filter(Hd_pu n ,wh itenoise_3), where n=1:M

[0111] It can be understood that a bandpass filter is designed by the frequency lower limit and the frequency upper limit in Table 3 above, and the randomly generated third white noise is bandpass filtered in the same way as the narrowband signal generation method mentioned above; the bandpass signal is then truncated by adding a time window, and the window function is determined by the duration in the above table. The amplitude of the window function at the start and end times should be 0, and it should rise and fall smoothly to avoid signal mutations and popping sounds caused by the synthesis of the impact signal with the previously generated broadband signal and narrowband signal; in this example, the time window and the bandpass filter signal after the window are used for the two impact signals, respectively. Fig.11 and Fig.12 As shown;

[0112] S112, generating an overall window function of the impulse noise according to the density of the impulse noise component and the single impulse time, performing overall windowing on the bandpass filter signal by the overall window function, and then performing amplitude calibration by the impulse amplitude calibration coefficient to generate an impulse noise simulation signal;

[0113] Specifically, an impact unit window function is set for each impact noise component; the time length of the impact unit window function is the same as the single impact time; the window function values ​​at the start and end times of the impact unit window function are set to zero; and the window function is second-order differentiable in intervals other than the start and end times.

[0114] Among them, the window function w_unit of the impact unit n The length of time and the single impact time pu_time nSimilarly, the window function of the impact unit must have the following characteristics to avoid the uneven time domain signal and the popping phenomenon when the impact signal is superimposed: the window function value at the start and end time is 0; the window function is second-order differentiable in other intervals except the start and end time.

[0115] Specifically, the impact signal is composed of a series of impact units with uniform discrete distribution; the impact number within the simulation time is obtained by multiplying the impact density with the total time length of the simulation sample; and a random number with uniform discrete distribution within the simulation time is generated according to the impact number.

[0116] The impact signal is composed of a series of uniformly discrete impact units. The number of impacts in the simulation time is obtained by multiplying the impact density: dense_pu(n) by the total time length of the simulation sample: T: n =dense_pu n *T,n=1:M.

[0117] Then generate a random number that is uniformly distributed over the simulation duration: R_t n,k ,k=1:N_pu n ,n=1:M;

[0118] The white noise is filtered using a bandpass filter with a freely set bandpass frequency range to generate a bandpass filtered signal, and then the signal is processed using a window function and uniformly discrete random distributed random numbers to generate uniformly discrete random distributed impact noise.

[0119] Specifically, the impact unit window function is respectively subjected to the sum of the impact number delay times to obtain the overall window function, wherein the delay time is a random number uniformly and discretely distributed within the simulation duration;

[0120] The impact noise simulation signal is generated by multiplying the bandpass filtered signal with the overall window function and then multiplying it by the corresponding impact amplitude calibration coefficient.

[0121] There are two main methods for simulating impact signals. Method 1: window the bandpass filter signal as a whole using a uniform discrete random distributed window function to generate an impact noise simulation signal; the window function of the impact unit is respectively N_pu n The delay times are summed up, and the delay times are: R_t n,k , and get the overall window function:

[0122]

[0123] Then by the bandpass signal: Sig_bp n , and the overall window function: w_pu n (t) and then multiply by the corresponding amplitude coefficient: A_pu n, superimposed to generate an impact noise simulation signal: Sig_pu_total

[0124]

[0125] This method generates a time window with the same length as a single impact time, delays and superimposes the window function through uniformly discretely distributed random numbers to form an overall time window function, and then windowing the bandpass noise signal as a whole to generate impact noise.

[0126] Specifically, the bandpass filtered signal is multiplied by the impact unit window function to obtain a single impact unit; the single impact unit is subjected to impact number times the delay time and the impact number of single impact units with uniform discrete distribution is summed up, wherein the delay time is a random number with uniform discrete distribution within the simulation duration; after the single impact unit is multiplied by the corresponding impact amplitude calibration coefficient, the results are superimposed and summed to obtain the impact noise simulation signal.

[0127] Among them, the second method of simulating impact signal is to superimpose the impact noise simulation signal by uniform discrete random distribution of impact units;

[0128] The passband signal: Sig_bp n Window function with impact unit: w_unit n Multiply to get a single impact unit:

[0129] Sig_unit n =Sig_bp n *w_unit n , n=1:M

[0130] The impact unit: Sig_unit n N_pu n The delay times are summed up, and the delay times are: R_t n,k , and obtain the uniform discrete distribution N_pu n impact unit, multiplied by the amplitude coefficient of the corresponding component, and superimposed and summed to obtain the impact noise simulation signal: Sig_pu_total

[0131]

[0132] It can be understood that the total number of impacts is obtained by multiplying the impact density by the total sample duration, and the time delay of all impact signals is generated by generating random numbers. In this example, a sample of 10s length is to be generated. As can be seen from Table 3, the frequency of the impact noise component 2 is 3 times / s, that is, a total of 30 impact units need to be generated. Make 30 copies of the windowed bandpass filtered signal 2 in 3.2, and then generate 30 random numbers uniformly and discretely distributed within 10s as the time delay of each impact unit, and obtain the distribution of all impact units, as shown in Fig.13 and Fig.14 Similarly, the impulse noise component 1 will generate 4 impulse units within 10s, and the 4 impulse units are delayed respectively by 4 random numbers uniformly and discretely distributed within 10, and the distribution of the impulse units of the impulse noise component 1 is as follows: Fig.15 , Fig.16 As shown; the impact signal component 1 and the impact signal component 2 are multiplied by the corresponding amplitude coefficients and then added to synthesize the impact signal. The FFT spectrum average result of the impact signal is shown in Fig.17 ; The FFT-time spectrum analysis results of the impulse signal are as follows Fig.18 As shown, it can be seen that the synthetic impact signal can well simulate the road impact noise component of the real vehicle collected samples in terms of time-frequency characteristics;

[0133] This method generates a time window with the same length as a single impact time, adds a window to the bandpass noise signal, generates a unit impact noise signal, and then delays and superimposes the unit impact noise signal through a uniformly discretely distributed random number to generate impact noise. The time window used in this method must meet the following characteristics: first, the window function value at the start and end time is 0, and second, the window function is second-order differentiable in other intervals except the start and end time; the impact density of the road impact noise signal can be adjusted by setting the distribution number of random numbers; the frequency characteristics of the road impact noise signal can be adjusted by changing the upper and lower limits of the frequency of the bandpass filter; the time length and rise and decay characteristics of the impact unit can be adjusted by changing the window function; the size of the road impact noise can be adjusted by changing the impact noise amplitude coefficient.

[0134] S114, adding the broadband noise simulation signal, the narrowband noise simulation signal, and the impact noise simulation signal to obtain a simulation signal of the noise inside the vehicle traveling at a constant speed, and saving the simulation signal in a format that can be played by electronic equipment for playback and listening evaluation.

[0135] Specifically, the synthesized uniformly moving vehicle interior noise analog signal has the energy of the frequency band where the added signal components are located increased due to the addition of the narrowband noise analog signal and the impact noise analog signal, and the increased frequency band energy information is fed back to restore the energy of each frequency band of the synthesized uniformly moving vehicle interior noise analog signal to the original sample level;

[0136] Wherein, restoring the energy of each frequency band of the synthesized uniform speed vehicle interior noise simulation signal to the original sample level also includes:

[0137] By correcting the sound pressure amplitude calibration coefficient, the corresponding frequency band energy in the broadband noise simulation signal is reduced, and the enhanced energy information is fed back. Through multiple iterations, the energy of each frequency band of the synthesized uniform speed vehicle interior noise simulation signal is restored to the original sample level.

[0138] By adjusting the sound pressure amplitude calibration coefficient, the narrowband amplitude calibration coefficient and the impact amplitude calibration coefficient, the components of the simulated synthetic uniform speed vehicle interior noise are changed to generate different simulation samples;

[0139] The various samples generated after adjustment were played back for listening evaluation.

[0140] Among them, the simulation synthesis of the noise inside the car running at a constant speed is synthesized by superposition of broadband noise simulation signal, narrowband noise simulation signal and impact noise simulation signal;

[0141] The analog signal of the noise inside the car running at a constant speed is obtained by adding the analog signal of broadband noise, the analog signal of narrowband noise, and the analog signal of impact noise:

[0142] Sig_combine=Sig_wb_total+Sig_nb_total+Sig_pu_total

[0143] Amplitude correction of the synthesized signal. The addition of narrowband and impulse signals to the synthesized signal will increase the energy of the 1 / 3 octave where the added signal components are located. The increased energy information needs to be fed back and the amplitude coefficient A_wb of the broadband signal needs to be corrected. n , reduce the corresponding 1 / 3 octave energy in the Sig_wb_total signal, so that the energy of each 1 / 3 octave of the synthesized signal Sig_combine is restored to the original sample level. This process can be completed through multiple iterations, such as Fig.19 As shown;

[0144] The analysis and verification of the simulated synthetic signal is carried out by performing FFT analysis on the synthesized signal. The results are as follows Fig. 20 As shown, it can be seen that the spectrum trends are basically consistent, and the main peak components are well simulated; the synthesized signal is subjected to FFT-time time-frequency analysis, and the results are as follows Fig.21 As shown, it can be seen that the time-frequency characteristics of the simulated synthetic noise signal are similar to the noise signal collected from the real vehicle, and each component is well simulated.

[0145] Playback evaluation of analog synthesized signals, saving analog synthesized signals in a format that can be played by electronic devices for listening playback;

[0146] The simulated synthesized signal is saved in a format that can be played by electronic devices. In this example, it is saved in a .wav file format, and played back using professional acoustic analysis software on a computer. The sound is compared with the noise signal collected from the actual vehicle, and the auditory experience is consistent.

[0147] The application of the method of simulating and synthesizing the noise inside a uniformly moving car is to adjust the parameters A_wb of the broadband signal, narrowband signal and impulse signal in the synthetic signal. n , A_nb n , A_pu n etc., and change the components of the simulated synthetic constant-speed vehicle interior noise to generate different simulation samples;

[0148] Playback and listening evaluation of various samples generated after adjustment can be used to identify the noise problem in a vehicle traveling at a constant speed, formulate optimization goals for the noise in a vehicle traveling at a constant speed, determine the optimization direction, and predict the auditory perception caused by changes in noise components. This example reduces the amplitude of the road impact noise component from 80Hz to 90Hz, increases the amplitude of the 1 / 3 octave frequency band with a center frequency of 500Hz, reduces the narrowband peaks near 200Hz, 400Hz, and 1kHz, and simulates the generation of optimized samples. The FFT spectrum average results of the optimized samples are shown in the figure below. Fig. 22 , Fig.23 As shown;

[0149] The optimized sound samples were played back and compared with the noise collected from the actual vehicle. The overall auditory experience was significantly improved, with the road impact noise component reduced, the mid-frequency area fuller, and the relatively monotonous prominent components basically eliminated. The simulated optimized sample can be used as the optimization target for the uniformly moving interior noise of the example vehicle model. Professionals in the field of automotive noise control can use this method to identify the uniformly moving interior noise problem, formulate uniformly moving interior noise optimization targets, determine the optimization direction, predict the auditory experience caused by changes in noise components, and use this to carry out further work such as automotive development and design.

[0150] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for simulating the noise inside a car when the car is traveling at a constant speed, characterized in that: include: Collect the interior noise signal samples of the real car running at a constant speed, determine the frequency range of the simulated signal, divide the frequency range into multiple continuous frequency bands, calculate the average sound pressure of the interior noise signal samples in each frequency band, the average FFT spectrum result, and the FFT-time time-frequency analysis result; Randomly generate a first white noise signal, calculate the ratio of the average sound pressure of the in-vehicle noise signal sample in each frequency band to the average sound pressure of the first white noise signal in each frequency band, and obtain the sound pressure amplitude calibration coefficient of each frequency band; The first white noise signal is band-pass filtered in each frequency band, and then calibrated with the corresponding sound pressure amplitude calibration coefficient, and then superimposed to obtain a broadband noise simulation signal; According to the FFT spectrum average result of the in-vehicle noise signal, the frequency lower limit, frequency upper limit and amplitude of multiple narrowband components are determined; the narrowband amplitude calibration coefficient is set according to the amplitude of each narrowband component; a second white noise signal is randomly generated, the second white noise signal is filtered according to the frequency lower limit and frequency upper limit of each narrowband component, and the amplitude is calibrated by the narrowband amplitude calibration coefficient and then superimposed to obtain a narrowband noise simulation signal; Based on the FFT-time frequency analysis results of the in-vehicle noise signal, the lower frequency limit, upper frequency limit, density, single impact time and impact amplitude of multiple road impact noise components are respectively determined; the impact amplitude calibration coefficient of the corresponding road impact noise component is set according to the impact amplitude; Randomly generate a third white noise signal, and perform band-pass filtering on the third white noise signal according to the lower frequency limit and the upper frequency limit of the impulse noise component to generate a band-pass filtered signal having the same or similar frequency range as the impulse noise; An overall window function of the impulse noise is generated according to the density of the impulse noise component and the single impulse time. After the bandpass filter signal is overall windowed by the overall window function, the amplitude is calibrated by the impulse amplitude calibration coefficient to generate an impulse noise simulation signal. The broadband noise simulation signal, the narrowband noise simulation signal, and the impact noise simulation signal are added together to obtain a simulation signal of the noise inside the car traveling at a constant speed. The simulation signal is saved in a format that can be played by electronic equipment for playback and listening evaluation.

2. The method for simulating the noise inside a car when the car is traveling at a constant speed according to claim 1, characterized in that: Collect the noise signal samples of the real car running at a constant speed, determine the frequency range of the simulated signal, divide the frequency range into multiple continuous frequency bands, and also include: The frequency range of the analog signal is determined by the lower limit of the starting frequency band and the upper limit of the ending frequency band, or is determined by actual needs; Divide the analog signal frequency range into contiguous frequency bands using integer or fractional octave bands or fixed bandwidth bands.

3. The method for simulating the noise inside a car when the car is traveling at a constant speed according to claim 1, characterized in that: Randomly generate a first white noise signal, calculate the ratio of the average sound pressure of the in-vehicle noise signal sample in each frequency band to the average sound pressure of the first white noise signal in each frequency band, and obtain the sound pressure amplitude calibration coefficient of each frequency band; The first white noise signal is band-pass filtered in each frequency band, and then the sound pressure amplitude is calibrated by the corresponding sound pressure amplitude calibration coefficient, and then superimposed to obtain a broadband noise simulation signal, which also includes: A one-to-one corresponding wideband pass filter is set according to multiple frequency bands, and the first white noise signal is passed through each wideband pass filter to obtain multiple wideband pass signals. Each wideband pass signal is amplitude calibrated by a corresponding sound pressure amplitude calibration coefficient and then superimposed to obtain a wideband noise simulation signal.

4. The method for simulating the noise inside a car when the car is traveling at a constant speed according to claim 1, characterized in that: According to the FFT spectrum average result of the in-vehicle noise signal, the frequency lower limit, frequency upper limit and amplitude of multiple narrowband components are determined; the narrowband amplitude calibration coefficient is set according to the amplitude of each narrowband component; a second white noise signal is randomly generated, the second white noise signal is filtered according to the frequency lower limit and frequency upper limit of each narrowband component, and the second white noise signal is amplitude calibrated by the narrowband amplitude calibration coefficient and then superimposed to obtain a narrowband noise simulation signal, which also includes: The number of the multiple narrowband components is determined by the number of main narrowband components in the FFT spectrum average result of the in-vehicle noise signal and the simulation accuracy requirement; Design narrow-band bandpass filters based on the lower and upper frequency limits of each narrow-band component; The second white noise signal passes through corresponding narrowband bandpass filters respectively to obtain narrowband filtered signals of multiple narrowband components; A plurality of narrowband filter signals are respectively calibrated with their corresponding narrowband amplitude calibration coefficients and then superimposed to obtain a narrowband noise simulation signal.

5. The method for simulating the noise inside a car when the car is traveling at a constant speed according to claim 1, characterized in that: Randomly generating a third white noise signal, performing bandpass filtering on the third white noise signal according to the lower frequency limit and the upper frequency limit of the impulse noise component to generate a bandpass filtered signal having the same or similar frequency range as the impulse noise, further comprising: A bandpass filter is set according to the lower frequency limit and the upper frequency limit of the impulse noise component, and the third white noise is bandpass filtered to generate a bandpass filtered signal with the same or similar frequency range as the impulse noise.

6. The method for simulating the noise inside a car when the car is traveling at a constant speed according to claim 1, characterized in that: The overall window function for generating the impulse noise is used to perform overall windowing on the bandpass filter signal by the overall window function, and then the amplitude is calibrated by the impulse amplitude calibration coefficient to generate an impulse noise simulation signal, which also includes: Set the impulse unit window function for each impulse noise component; The time length of the impact unit window function is the same as the single impact time; Set the window function value at the start and end time of the impact unit window function to zero; The window function is second-order differentiable in intervals other than the start and end times.

7. The method for simulating the noise inside a car when the car is traveling at a constant speed according to claim 6, characterized in that: include: The impact signal is composed of a series of evenly and discretely distributed impact units; The number of shocks in the simulation time is obtained by multiplying the shock density by the total time length of the simulation sample; Generates a random number that is discretely distributed uniformly over the simulation duration based on the number of shocks.

8. The method for simulating the noise inside a car when the car is traveling at a constant speed according to claim 7, characterized in that: include: The impact unit window function is processed by summing the impact number delay times to obtain the overall window function, where the delay time is a random number uniformly and discretely distributed within the simulation duration; The impact noise simulation signal is generated by multiplying the bandpass filtered signal with the overall window function and then multiplying it by the corresponding impact amplitude calibration coefficient.

9. The method for simulating the noise inside a car when the car is traveling at a constant speed according to claim 7, characterized in that: include: Multiply the bandpass filtered signal by the impact unit window function to obtain a single impact unit; The delay time of the impact number is summed up for each impact unit to obtain the impact number impact units with uniform discrete distribution, where the delay time is a random number with uniform discrete distribution within the simulation duration; After a single impact unit is multiplied by the corresponding impact amplitude calibration coefficient, the impact noise simulation signal is obtained by superposition and summation.

10. The method for simulating the noise inside a car when the car is traveling at a constant speed according to claim 2, characterized in that: The broadband noise simulation signal, the narrowband noise simulation signal, and the impact noise simulation signal are added together to obtain a simulation signal of the noise inside the vehicle traveling at a constant speed, and the simulation signal is saved in a format that can be played by electronic equipment for playback and listening evaluation, and also includes: The synthesized uniform speed vehicle interior noise simulation signal has the energy of the frequency band where the added signal components are located increased due to the addition of the narrowband noise simulation signal and the impact noise simulation signal. The increased frequency band energy information is fed back to restore the energy of each frequency band of the synthesized uniform speed vehicle interior noise simulation signal to the original sample level. Wherein, restoring the energy of each frequency band of the synthesized uniform speed vehicle interior noise simulation signal to the original sample level also includes: By correcting the sound pressure amplitude calibration coefficient, the corresponding frequency band energy in the broadband noise simulation signal is reduced, and the enhanced energy information is fed back. Through multiple iterations, the energy of each frequency band of the synthesized uniform speed vehicle interior noise simulation signal is restored to the original sample level. By adjusting the sound pressure amplitude calibration coefficient, the narrowband amplitude calibration coefficient and the impact amplitude calibration coefficient, the components of the simulated synthetic uniform speed vehicle interior noise are changed to generate different simulation samples; The various samples generated after adjustment were played back for listening evaluation.

Citation Information

Patent Citations

  • Water surface target radiation noise simulation method

    CN103323821A

  • Apparatus for providing environmental noise compensation for a synthesized vehicle sound

    CN104648238A

  • Method and system for simulating engine sounding during automobile acceleration and deceleration

    CN111731185A

  • Aircraft cabin noise synthesis method based on auditory perception

    CN113053351A

  • Method for generating a synthetic driving noise characteristic

    DE102020007391A1