Audio parameter calibration method and device of warning system

Abstract

The application provides an audio parameter calibration method and device of a warning system. The method is as follows: an original prompt audio is obtained; an iteration operation is performed to assign a current group to an audio adjustment parameter; the frequency and / or amplitude of the original prompt audio is adjusted according to the audio adjustment parameter to obtain an output audio; a prompt sound signal corresponding to the output audio is predicted based on a transfer path impulse response coefficient corresponding to a transfer path; the fitness of the audio adjustment parameter corresponding to the output audio is calculated based on the prompt sound signal; the optimal value of each iteration of the audio adjustment parameter is determined according to the fitness; if the current iteration number has reached a preset maximum iteration number, the frequency and / or amplitude of the original prompt audio is calibrated according to the current optimal value of the audio adjustment parameter to obtain a target audio. The audio adjustment parameter is optimized through the iteration operation, the audio parameter of the prompt sound of the warning system is automatically calibrated, and the debugging efficiency of the warning system is improved, thereby accelerating the development process.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electric vehicles, and in particular to an audio parameter calibration method and device for a warning system. BACKGROUND

[0002] With the rapid development of the new energy vehicle industry, the sales of electric vehicles continue to grow. Since electric vehicles do not use internal combustion engines, the noise is low when driving at low speed, making it difficult to alert pedestrians to the vehicle and increasing the incidence of accidents. To reduce safety hazards, the state has developed relevant regulations / standards, requiring electric vehicles to be equipped with a pedestrian warning system to achieve the purpose of playing a warning prompt sound when the electric vehicle is driving at low speed. Generally, the pedestrian warning system obtains current vehicle conditions such as speed and gear position through the vehicle CAN bus, and adjusts the frequency and / or amplitude of the audio signal according to the control module to output, so as to realize low-speed prompt sound playing. In order to meet the requirements of relevant regulations / standards, it is necessary to calibrate the key parameters (such as frequency and amplitude) of the audio signal in the pedestrian warning system under different working conditions.

[0003] The commonly used calibration method is to stabilize the vehicle speed by manually controlling the accelerator pedal, and manually adjusting the parameters and repeatedly testing to make the prompt sound emitted by the loudspeaker of the pedestrian warning system meet the requirements. However, this calibration method has the problems of relying on manual operation, time-consuming and laborious, and low efficiency. SUMMARY

[0004] Therefore, the embodiments of the present application provide an audio parameter calibration method and device for a warning system to solve the problems of time-consuming, laborious and low efficiency of the current audio parameter calibration method.

[0005] To achieve the above-mentioned purpose, the embodiments of the present application provide the following technical solutions:

[0006] The first aspect of the embodiments of the present application discloses an audio parameter calibration method for a warning system, the method comprising:

[0007] obtaining an original prompt audio;

[0008] performing an iteration operation to assign a current population to an audio adjustment parameter, the audio adjustment parameter including a frequency adjustment factor and / or an amplitude adjustment factor, the population being candidate values of the audio adjustment parameter, the population being composed of a preset number of particles, each of the particles having a corresponding speed value and a position value;

[0009] adjusting the frequency and / or amplitude of the original prompt audio according to the audio adjustment parameter to obtain an output audio;

[0010] predict a prompt sound signal corresponding to the output audio based on a transfer path impulse response coefficient corresponding to a transfer path, the prompt sound signal being a signal of the output audio propagating to a noise measurement point through the transfer path;

[0011] calculate a fitness of the audio adjustment parameter corresponding to the output audio based on the prompt sound signal;

[0012] determine an optimal value of each iteration of the audio adjustment parameter according to the fitness;

[0013] if the current iteration number has reached a preset maximum iteration number, calibrate the frequency and / or amplitude of the original prompt audio according to the current optimal value of the audio adjustment parameter, to obtain a target audio.

[0014] Preferably, the method further comprises:

[0015] if the current iteration number has not reached the preset maximum iteration number, update the speed value and position value corresponding to all the particles in the group according to the optimal value of each iteration of the audio adjustment parameter, a preset cognitive learning factor, a preset social learning factor and a preset inertia weight, to obtain a new group, take the new group as a current group, and return to perform the iteration operation.

[0016] Preferably, the method further comprises:

[0017] calculating a transfer path impulse response coefficient between a preset controller output signal and a preset noise measurement point by using an adaptive filter;

[0018] performing convolution calculation on the output audio and the transfer path impulse response coefficient, to obtain the prompt sound signal corresponding to the output audio.

[0019] Preferably, the method further comprises:

[0020] controlling a preset loudspeaker to play a white noise signal at a preset speaker arrangement point by using a preset controller;

[0021] collecting a sound signal corresponding to the white noise signal at a preset noise measurement point by using a microphone;

[0022] inputting the white noise signal into an adaptive filter, and calculating an output signal of the adaptive filter;

[0023] calculating a difference between the sound signal and the output signal;

[0024] Based on the white noise signal and the difference value, the weight coefficient of the preset filter is updated by using an adaptive filtering algorithm until the adaptive filtering algorithm converges to a steady state, so as to obtain a transfer path impulse response coefficient between the preset controller output signal and the preset noise measuring point.

[0025] Preferably, the step of calculating the fitness of the audio adjustment parameter corresponding to the output audio based on the prompt tone signal comprises:

[0026] calculating a total sound pressure level corresponding to the prompt tone signal and sound pressure levels of a plurality of preset octaves;

[0027] based on a preset frequency requirement, counting a first number and a second number of sound pressure levels of the plurality of preset octaves that meet preset conditions;

[0028] calculating the fitness of the audio adjustment parameter corresponding to the output audio based on the total sound pressure level, the first number and the second number.

[0029] Preferably, the step of counting the first number and the second number of sound pressure levels of the plurality of preset octaves that meet preset conditions based on a preset frequency requirement comprises:

[0030] comparing the sound pressure level of each preset octave with a standard sound pressure level of the preset octave based on a preset frequency requirement, the preset frequency requirement at least including the standard sound pressure level of each preset octave and a standard frequency;

[0031] if the sound pressure level of the preset octave is greater than the standard sound pressure level of the preset octave, marking the sound pressure level of the preset octave as a standard sound pressure level;

[0032] comparing a center frequency corresponding to the standard sound pressure level with the standard frequency;

[0033] counting a number of the standard sound pressure levels that meet a first requirement and marking the number as a first number, and counting a number of the standard sound pressure levels that meet a second requirement and marking the number as a second number, the first requirement being that the center frequency corresponding to the standard sound pressure level is not greater than the standard frequency, and the second requirement being that the center frequency corresponding to the standard sound pressure level is greater than the standard frequency.

[0034] The second aspect of the embodiment of the present application discloses an audio parameter calibration device of a warning system, and the device comprises:

[0035] an acquisition unit configured to acquire an original prompt audio;

[0036] an execution unit configured to execute an iteration operation of assigning a current population to audio adjustment parameters, the audio adjustment parameters comprising a frequency adjustment factor and / or an amplitude adjustment factor, the population being candidate values of the audio adjustment parameters, the population being composed of a preset number of particles, each of the particles having a corresponding velocity value and a position value;

[0037] an adjustment unit configured to adjust a frequency and / or an amplitude of the original prompt audio according to the audio adjustment parameters to obtain an output audio;

[0038] a prediction unit configured to predict a prompt tone signal corresponding to the output audio based on a transfer path impulse response coefficient corresponding to a transfer path, the prompt tone signal being a signal of the output audio propagating to a noise measurement point through the transfer path;

[0039] a calculation unit configured to calculate a fitness of the audio adjustment parameters corresponding to the output audio based on the prompt tone signal;

[0040] a determination unit configured to determine an optimal value of each iteration of the audio adjustment parameters according to the fitness;

[0041] a calibration unit configured to calibrate a frequency and / or an amplitude of the original prompt audio according to the current optimal value of the audio adjustment parameters to obtain a target audio if a current iteration number has reached a preset maximum iteration number.

[0042] Preferably, the device further comprises:

[0043] an update unit configured to update the velocity value and the position value corresponding to all the particles in the population according to the optimal value of each iteration of the audio adjustment parameters, a preset cognitive learning factor, a preset social learning factor and a preset inertia weight to obtain a new population if the current iteration number has not reached the preset maximum iteration number, the new population being used as a current population, and the execution iteration operation being returned.

[0044] Preferably, the prediction unit comprises:

[0045] a calculation module configured to calculate a transfer path impulse response coefficient between a preset controller output signal and a preset noise measurement point by using an adaptive filter;

[0046] a determination module configured to determine a transfer function according to the transfer path impulse response coefficient;

[0047] a prediction module configured to perform convolution calculation based on the output audio and the transfer path impulse response coefficient to obtain the prompt tone signal corresponding to the output audio.

[0048] Preferably, the calculation module comprises:

[0049] a playing sub-module, configured to play a white noise signal at a preset speaker arrangement point by using a preset controller;

[0050] a collecting sub-module, configured to collect a sound signal corresponding to the white noise signal at a preset noise measuring point by using a microphone;

[0051] a first calculating sub-module, configured to input the white noise signal into an adaptive filter and calculate an output signal of the adaptive filter;

[0052] a second calculating sub-module, configured to calculate a difference between the sound signal and the output signal;

[0053] an updating sub-module, configured to update a weight coefficient of the preset filter by using an adaptive filtering algorithm based on the white noise signal and the difference until the adaptive filtering algorithm converges to a steady state, so as to obtain a transfer path impulse response coefficient between a preset controller output signal and a preset noise measuring point.

[0054] The audio parameter calibration method and device for the warning system provided in the above embodiment of the present application obtain an original prompt audio; perform an iteration operation to assign a current group to an audio adjustment parameter; adjust the frequency and / or amplitude of the original prompt audio according to the audio adjustment parameter to obtain an output audio; predict a prompt sound signal corresponding to the output audio based on the transfer path impulse response coefficient corresponding to the transfer path; calculate the fitness of the audio adjustment parameter corresponding to the output audio based on the prompt sound signal; determine the optimal value of each iteration of the audio adjustment parameter according to the fitness; and if the current iteration number has reached a preset maximum iteration number, calibrate the frequency and / or amplitude of the original prompt audio according to the current optimal value of the audio adjustment parameter to obtain a target audio. The audio adjustment parameter is optimized through the iteration operation, and the audio parameters of the prompt sound of the warning system are automatically calibrated, which improves the debugging efficiency of the warning system and speeds up the development process. BRIEF DESCRIPTION OF DRAWINGS

[0055] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, brief descriptions will be given to the drawings needed to be used in the embodiments or prior art descriptions. Obviously, the drawings in the following description are only embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative labor on the basis of the provided drawings.

[0056] Figure 1 The flowchart of the audio parameter calibration method for the warning system provided in the embodiment of the present application;

[0057] Figure 2The lowest sound level limit diagram of 20km / h uniform forward driving working condition provided for the embodiment of the present application is shown in the figure;

[0058] Figure 3 The audio data x1[K] diagram provided for the embodiment of the present application is shown in the figure;

[0059] Figure 4 The audio data x2[K] diagram provided for the embodiment of the present application is shown in the figure;

[0060] Figure 5 The transfer path diagram between the loudspeaker arrangement point and the noise measuring point provided for the embodiment of the present application is shown in the figure;

[0061] Figure 6 The impulse response curve example diagram of the transfer path provided for the embodiment of the present application is shown in the figure;

[0062] Figure 7 The amplitude-frequency response curve example diagram of the transfer path provided for the embodiment of the present application is shown in the figure;

[0063] Figure 8 The phase-frequency response curve example diagram of the transfer path provided for the embodiment of the present application is shown in the figure;

[0064] Figure 9 The 1 / 3 octave sound pressure level example diagram of the prompt tone signal corresponding to the output audio provided for the embodiment of the present application is shown in the figure;

[0065] Figure 10 The structural block diagram of the audio parameter calibration device of a warning system provided for the embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0066] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all the other embodiments obtained by those skilled in the art without creative work are within the protection scope of the present application.

[0067] In the present application, the terms “comprising”, “containing” or any other variants thereof are intended to cover the non-exclusive containing, so that the process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or device. Without more limitations, the element defined by the sentence “including a…” does not exclude the presence of other identical elements in the process, method, article or device including the element.

[0068] As the background technology shows, the commonly used calibration method involves manually controlling the accelerator pedal to stabilize the vehicle speed, while simultaneously manually adjusting parameters and conducting repeated tests to ensure that the warning sound emitted by the pedestrian warning system's speaker meets the requirements. However, this calibration method relies on manual operation, is time-consuming, labor-intensive, and inefficient.

[0069] Therefore, embodiments of the present invention provide an audio parameter calibration method and apparatus for an alarm system, comprising: acquiring the original alarm audio; performing an iterative operation to assign the current group value to the audio adjustment parameter; adjusting the frequency and / or amplitude of the original alarm audio according to the audio adjustment parameter to obtain the output audio; predicting the alarm tone signal corresponding to the output audio based on the impulse response coefficient of the transmission path corresponding to the transmission path; calculating the fitness of the audio adjustment parameter corresponding to the output audio based on the alarm tone signal; determining the optimal value of the audio adjustment parameter for each iteration based on the fitness; if the current iteration number has reached the preset maximum iteration number, calibrating the frequency and / or amplitude of the original alarm audio according to the current optimal value of the audio adjustment parameter to obtain the target audio. By optimizing the audio adjustment parameter through iterative operation, the audio parameters of the alarm system alarm tone are automatically calibrated, improving the debugging efficiency of the alarm system and accelerating the development process.

[0070] See Figure 1 The diagram shows a flowchart of an audio parameter calibration method for an alarm system provided by an embodiment of the present invention.

[0071] It should be noted that, in this embodiment of the invention, the audio parameters of the pedestrian warning system are calibrated based on the national standard "Low-Speed ​​Warning Sound for Electric Vehicles" (GB / T37153-2018) for a constant forward speed of 20 km / h, so that it meets the following technical indicators:

[0072] (1) The total sound pressure level corresponding to the pedestrian warning system's alert tone shall not be less than 58 dB(A);

[0073] (2) At least two frequency bands in the 1 / 3 octave band sound pressure level corresponding to the pedestrian warning system's alert sound meet the requirements (i.e., as provided in the embodiments of the present invention). Figure 2 As shown), and at least one of them is less than 1600Hz. For example, the number of 1 / 3 octave band sound pressure levels below 1600Hz (inclusive) that meet the standard is 1, i.e., num1=1; the number of 1 / 3 octave band sound pressure levels above 1600Hz that meet the standard is 1, i.e., num2=1; the expected total sound pressure level is set to SPL=60dB(A).

[0074] It is understood that the audio parameter calibration method provided in the embodiments of the present invention can also be used to calibrate the audio parameters of the pedestrian warning system for uniform forward driving conditions such as 10km / h in other national standards. The specific implementation method is similar to the implementation process described in the embodiments of the present invention, and will not be described redundantly here.

[0075] The audio parameter calibration method includes:

[0076] Step S101: Obtain the original prompt audio.

[0077] In the specific implementation step S101, the original prompt audio is obtained from the pedestrian warning system. For example, using... Figure 3 The audio data x1[K] shown is similar to... Figure 4 The original prompt audio is synthesized from the audio data x2[K]. The audio data is presented as an array containing K samples. The audio data x1[K] contains frequency components of 200Hz, 270Hz, and 340Hz; the audio data x2[K] contains frequency components of 400Hz, 540Hz, and 680Hz; the number of samples K = 39690.

[0078] Step S102: Perform an iterative operation to assign the current group value to the audio adjustment parameter.

[0079] In the specific implementation step S102, a specific iterative operation is performed, assigning the position values ​​corresponding to a preset number of particles in the current swarm to the audio adjustment parameters. That is, the swarm is a candidate value for the audio adjustment parameters, and the swarm consists of a preset number of particles, where each particle has a corresponding velocity value and position value.

[0080] It should be noted that the current swarm is pre-generated based on the Particle Swarm Optimization (PSO) algorithm. Specifically, the PSO algorithm generates a predetermined number of particles, initializes the velocity and position values ​​for each particle, and thus obtains the current swarm. For example, the velocity value v... i (t)=[v i1 (t),v i2 (t),v i3 (t),v i4 (t) and position value x i (t)=[x i1 (t),x i2 (t),x i3 (t),x i4 (t)].

[0081] In practical application, the audio adjustment parameters include frequency adjustment factors and / or amplitude adjustment factors. Define the frequency adjustment factor F1 and the amplitude adjustment factor A1 corresponding to the audio x1[K], and the frequency adjustment factor F2 and the amplitude adjustment factor A2 corresponding to the audio x2[K].

[0082] It can be understood that the position value corresponding to each particle in the current population is assigned to the audio adjustment parameters F1, A1, F2 and A2.

[0083] Step S103: Adjust the frequency and / or amplitude of the original prompt audio according to the audio adjustment parameters to obtain the output audio.

[0084] In the process of implementing step S103, the frequency and / or amplitude of the original prompt audio is adjusted according to the audio adjustment parameters by using an adjustment method (such as interpolation method) to obtain the output audio.

[0085] It can be understood that the original prompt audio x i [K] is defined with the index index k = 1, 2, …, K, and the output audio is y i [L] with the index index = 1, 2, …, L; i = 1, 2; assuming that the frequency of the original audio is amplified F i times, and the output is g i , and on this basis, the amplitude is amplified A i times, which is the final output audio. Define the interpolation point p i , which is initialized to 1.

[0086] Then the output y i [L] is calculated as follows:

[0087] 1) Determine whether the value obtained by rounding up the interpolation point p i belongs to the index index k (k = 1, 2, 3, …), if yes, take the kth element of the original prompt audio x i [K], denoted as g i , and multiply it by A i , which is the lth sample of the output audio y i [L];

[0088] 2) Determine whether the value obtained by rounding up the interpolation point p i belongs to the index index k (k = 1, 2, 3, …), if not, perform interpolation calculation. Denote the value obtained by rounding up the interpolation point p i as r 1i , and denote the value obtained by rounding down the interpolation point p i as r 0i , take the r i th element of the original prompt audio x 1iand the rth 0i There are n elements, denoted as z. 1i and z 0i Using (r) 0i ,z 0i ) and (r 1i ,z 1i The interpolation point p can be calculated. i The value on the straight line or curve passing through these two points is denoted as g. i Multiply it by A i y is the i-th sample of the output audio. i [L].

[0089] 3) At the current interpolation point p i Add frequency amplification factor F to the base i This is used to update the interpolation points.

[0090] 4) Repeat the above steps.

[0091] Step S104: Predict the prompt tone signal corresponding to the output audio based on the impulse response coefficient of the transmission path.

[0092] In the specific implementation step S104, the controller, speaker placement points and noise measurement points are set in advance, and the transmission path impulse response coefficient between the controller output signal and the noise measurement points is calculated using an adaptive filter; and convolution calculation is performed based on the output audio and the transmission path impulse response coefficient to obtain the prompt tone signal corresponding to the output audio.

[0093] It should be noted that, based on the output audio and the impulse response coefficient of the transmission path, the output audio corresponding to the prompt signal d(n) is obtained by convolution calculation using the following formula (1).

[0094]

[0095] Where y(n) is the audio output signal at time n; The impulse response coefficient of the transmission path; For FIR filters The order of.

[0096] It is understood that, referring to the embodiments of the present invention Figure 5 The transmission path between the pre-set controller output signal and the noise measurement point, as shown in the diagram, is specifically as follows: the electronic circuit between the controller's control module and the speaker placement point, and the acoustic channel between the speaker placement point and the noise measurement point (i.e., the "microphone" shown in the diagram). The control module is used to control the speaker to play white noise signals. The control module and speaker are located within the electric vehicle body; their specific placement depends on the actual situation and is not limited here.

[0097] The following details the process of calculating the impulse response coefficient of the transmission path corresponding to the transmission path between the pre-set controller output signal and the noise measurement point:

[0098] Two points in a specified three-dimensional space are designated as the loudspeaker placement point and the noise measurement point. A white noise signal v(n) is played through the pre-set loudspeaker placement point; at the pre-set noise measurement point, the corresponding acoustic signal d is collected using a microphone. v (n)(Specifically, the microphone of the microphone is used to collect the sound signal d corresponding to the white noise signal.) v (n)).

[0099] The white noise signal v(n) is input into a preset filter. (e.g., length is) In the adaptive FIR filter, the preset filter is calculated according to formula (2). The output signal y v (n):

[0100]

[0101] in, For FIR filters The order of; For FIR filters The filter coefficients at time n.

[0102] The acoustic signal d is calculated according to formula (3). v (n) (i.e., the desired signal) and the output signal y v The difference e between (n) v (n) (i.e., error).

[0103] e v (n)=d v (n)-y n (n) (3)

[0104] Based on the white noise signal v(n) and the difference e v (n), using an adaptive filtering algorithm (such as the NLMS algorithm, i.e., the following formula (4)) to apply the preset filter. The weight coefficients are updated until the adaptive filtering algorithm converges to a steady state, yielding the impulse response coefficients of the transmission path between the loudspeaker placement point and the noise measurement point.

[0105]

[0106] Where μ is the step size, Let n represent the weight vector at time n. The input signal vector.

[0107] It should be noted that in combination with the above, the impulse response curve example of the transfer path can refer to the impulse response curve example of the transfer path provided by the embodiment of the application Figure 6 ; the amplitude-frequency response curve example of the transfer path can refer to the amplitude-frequency response curve example of the transfer path provided by the embodiment of the application Figure 7 ; and the phase-frequency response curve example of the transfer path can refer to the phase-frequency response curve example of the transfer path provided by the embodiment of the application Figure 8 .

[0108] Step S105: Calculate the fitness of the audio adjustment parameter corresponding to the output audio based on the prompt tone signal.

[0109] In the process of specifically implementing step S105, the total sound pressure level corresponding to the prompt tone signal of the noise measurement point and the sound pressure levels of a plurality of preset octaves are calculated; based on the preset frequency requirement, the first number and the second number of the sound pressure levels of the plurality of preset octaves that meet the preset condition are counted; and the fitness of the audio adjustment parameter corresponding to the output audio is calculated based on the total sound pressure level, the first number and the second number.

[0110] It can be understood that the sound pressure level of each preset octave corresponds to a center frequency, and the preset octave can be a 1 / 3 octave.

[0111] It should be noted that the specific implementation process of calculating the total sound pressure level corresponding to the prompt tone signal of the noise measurement point and the sound pressure levels of a plurality of preset octaves is as follows:

[0112] The total sound pressure level q corresponding to the output audio is calculated through a preset filter (such as a weighting network).

[0113] It should be noted that the total sound pressure level calculated by the following formula (5) is the sound pressure level that has not been processed by the weighting network. If the weighted sound pressure level is to be calculated, it also needs to be corrected, which will not be specifically illustrated here.

[0114]

[0115] Wherein, P i is the root mean square sound pressure of the prompt tone signal of the noise measurement point; P n is the reference sound pressure, which is 2*10 -5 Pa.

[0116] According to formula (6), the sound pressure levels of a plurality of preset octaves (for example, the sound pressure levels of 1 / 3 octaves) corresponding to the prompt tone signal of the noise measurement point are calculated:

[0117]

[0118] Wherein, f u represents the upper limit frequency; f l represents the lower limit frequency; f cRepresentative center frequency.

[0119] It can be understood that, based on the preset frequency requirement, the implementation process of counting the first quantity and the second quantity of the specific implementation process (process A1 to process A4) of the preset frequency requirement from the sound pressure level of several preset octaves that meet the preset condition is as follows:

[0120] A1: Based on the preset frequency requirement, compare the size between the sound pressure level of each preset octave and the standard sound pressure level of the preset octave.

[0121] It can be understood that the preset frequency requirement at least includes the standard sound pressure level of each preset octave and the standard frequency.

[0122] A2: If the sound pressure level of the preset octave is greater than the standard sound pressure level of the preset octave, mark the sound pressure level of the preset octave as a standard sound pressure level.

[0123] A3: Compare the center frequency corresponding to the standard sound pressure level with the standard frequency.

[0124] A4: Count the number of standard sound pressure levels that meet the first requirement and mark it as the first quantity, and count the number of standard sound pressure levels that meet the second requirement and mark it as the second quantity. Wherein, the first requirement is that the center frequency corresponding to the standard sound pressure level is not greater than the standard frequency, and the second requirement is that the center frequency corresponding to the standard sound pressure level is greater than the standard frequency.

[0125] For example Figure 2 As shown in the figure:

[0126] Compare the sound pressure level a1dB(A) of the 1 / 3 octave with the center frequency of 160Hz, whether it is greater than the standard sound pressure level 52dB(A) of the 1 / 3 octave with the center frequency of 160Hz;

[0127] If the sound pressure level a1dB(A) is greater than the standard sound pressure level 52dB(A), mark the current sound pressure level a1dB(A) of the 1 / 3 octave with the center frequency of 160Hz as a standard sound pressure level.

[0128] Compare the sound pressure level a2dB(A) of the 1 / 3 octave with the center frequency of 2500Hz, whether it is greater than the standard sound pressure level 46dB(A) of the 1 / 3 octave with the center frequency of 2500Hz;

[0129] If the sound pressure level a2dB(A) is greater than the standard sound pressure level 46dB(A), mark the current sound pressure level a2dB(A) of the 1 / 3 octave with the center frequency of 2500Hz as a standard sound pressure level.

[0130] whether the sound pressure level a3dB(A) of the 1 / 3 octave band with a center frequency of 800 Hz is greater than a standard sound pressure level 53dB(A) of the 1 / 3 octave band with a center frequency of 800 Hz;

[0131] If the sound pressure level a3dB(A) is less than the standard sound pressure level 53dB(A), it indicates that the sound pressure level a3dB(A) of the current 1 / 3 octave band with a center frequency of 800 Hz does not meet the standard.

[0132] whether the sound pressure level a4dB(A) of the 1 / 3 octave band with a center frequency of 400 Hz is greater than a standard sound pressure level 52dB(A) of the 1 / 3 octave band with a center frequency of 400 Hz;

[0133] If the sound pressure level a4dB(A) is greater than the standard sound pressure level 52dB(A), the sound pressure level a4dB(A) of the current 1 / 3 octave band with a center frequency of 400 Hz is marked as a passing sound pressure level.

[0134] The center frequencies of the passing sound pressure levels are compared with the standard frequencies, such as the center frequency 160 Hz of the sound pressure level a1dB compared with 1600 Hz, the center frequency 2500 Hz of the sound pressure level a2dB compared with 1600 Hz, and the center frequency 400 Hz of the sound pressure level a4dB compared with 1600 Hz.

[0135] The number of passing sound pressure levels meeting a first requirement is counted and marked as a first number, and the number of passing sound pressure levels meeting a second requirement is counted and marked as a second number. The first requirement is that the center frequency corresponding to the passing sound pressure level is not greater than the standard frequency, and the second requirement is that the center frequency corresponding to the passing sound pressure level is greater than the standard frequency. The first number is 2, and the second number is 1.

[0136] It should be noted that the specific implementation process of calculating the fitness of the audio adjustment parameter corresponding to the output audio based on the total sound pressure level, the first number and the second number is as follows:

[0137] A target function is established for calculating the fitness, which is shown in the following formula (7).

[0138] f=abs(h1-num1)+abs(h2-num2)+abs(q-SPL) (7)

[0139] num1 is the expected number of 1 / 3 octave sound pressure levels below 1600Hz (inclusive) that meet the standard, num2 is the expected number of 1 / 3 octave sound pressure levels above 1600Hz that meet the standard, SPL is the expected total sound pressure level; h1 is the number of sound pressure levels that meet the first requirement, i.e., the center frequency corresponding to the sound pressure level is not greater than the standard frequency; h2 is the number of sound pressure levels that meet the second requirement, i.e., the center frequency corresponding to the sound pressure level is greater than the standard frequency.

[0140] Step S106: determining the optimal value of the audio adjustment parameter in each iteration according to the fitness.

[0141] In the process of implementing step S106, the individual optimal position and the global optimal position of each particle in the current population are determined according to the principle that the smaller the fitness value is, the better it is, i.e., the optimal value of the audio adjustment parameter in each iteration is determined.

[0142] It should be noted that the individual optimal position of each particle in the current population can be represented as p pi (t) = [p pi1 , p pi2 , p pi3 , p pi4 ]; and the global optimal position can be represented as p G (t) = [p G1 , p G2 , p G3 , p G4 ].

[0143] Step S107: if the current iteration number has reached the preset maximum iteration number, calibrating the frequency and / or amplitude of the original prompt audio according to the current optimal value of the audio adjustment parameter to obtain the target audio.

[0144] In the process of implementing step S107, if the current iteration number has reached the preset maximum iteration number, it indicates that the current optimal value of the audio adjustment parameter is the optimal parameter that minimizes the objective function, and at the same time, the optimal value of the current audio adjustment parameter makes the number of 1 / 3 octave sound pressure levels of the prompt sound at the noise measurement point and the total sound pressure level reach or approach the expected value. Therefore, the frequency and / or amplitude of the original prompt audio are calibrated according to the optimal value of the audio adjustment parameter to obtain the target audio.

[0145] It should be noted that, in combination with the content explained in the above example, when the optimal value of the audio adjustment parameter is determined, the frequency adjustment factor F1 corresponding to the audio data x1[K] is 1.16205653348446, the amplitude adjustment factor A1 is 0.403482731700248, the frequency adjustment factor F2 corresponding to the audio data x2[K] is 4.04467877313752, and the amplitude adjustment factor A2 is 0.309339452322305. At the same time, the total sound pressure level of the output audio corresponding to the prompt tone signal obtained based on the preset transfer function is 59.999938 dB(A), and the first number and the second number that meet the preset condition are counted from a plurality of preset octave sound pressure levels, and the results are as shown in Table 1. Figure 9 As shown in Table 1, there is one band below 1600 Hz that meets the sound level limit requirement, the center frequency of which is 1600 Hz, and the sound pressure level is 59.373541 dB(A); in the frequency band of 1600 Hz to 5000 Hz, there is one band that meets the sound level limit requirement, the center frequency of which is 2000 Hz, and the sound pressure level is 50.210700 dB(A).

[0146] In some embodiments, if the current iteration number does not reach the preset maximum iteration number, the velocity value and the position value corresponding to all particles in the group are updated according to the optimal value of the audio adjustment parameter in each iteration, the preset cognitive learning factor, the preset social learning factor and the preset inertia weight, a new group is obtained, the new group is taken as the current group, and the iteration operation is returned to be executed, that is, step S102.

[0147] That is, if the current iteration number does not reach the preset maximum iteration number, the optimal value of the audio adjustment parameter in each iteration is input into the following update formula (8) and update formula (9) to adjust the position value and the velocity value of the particle in the current group.

[0148] v i (t+1)=W(t)·v i (t)+r1c1(p pi (t)-p i (t))+r2c2(p G (t)-p i (t)) (8)

[0149] p i (t+1)=p i (t)+v i (t+1) (9)

[0150] Wherein, c1 is a preset cognitive learning factor, c2 is a preset social learning factor, r1 and r2 are random numbers in [0, 1], and W(t) is a preset inertia weight.

[0151] It should be noted that the preset inertia weight W(t) is controlled by a nonlinear adaptive adjustment method, and is calculated according to the following formula (10):

[0152]

[0153] Wherein, t max is the maximum iteration number, the maximum value of the inertia weight W1=0.9, and the minimum value of the inertia weight W2=0.4.

[0154] Based on the above updating formula, each particle is directed to the optimal position, a new population is obtained, and the new population is taken as the current population, and the iteration operation is returned to be executed, that is, step S102.

[0155] In the embodiment of the application, the audio adjustment parameter is optimized through the iteration operation, the audio parameter of the warning system prompt sound is automatically calibrated, the debugging efficiency of the warning system is improved, and the development process is accelerated.

[0156] Corresponding to the audio parameter calibration method of the warning system provided in the above embodiment of the application, referring to Figure 10 , a structural block diagram of an audio parameter calibration device of a warning system provided in an embodiment of the application is shown. The audio parameter calibration device comprises an acquisition unit 1001, an execution unit 1002, an adjustment unit 1003, a prediction unit 1004, a calculation unit 1005, a determination unit 1006 and a calibration unit 1007.

[0157] The acquisition unit 1001 is configured to acquire an original prompt sound.

[0158] The execution unit 1002 is configured to perform an iteration operation, assign a current population to an audio adjustment parameter, and include a frequency adjustment factor and / or an amplitude adjustment factor in the audio adjustment parameter. The population is a candidate value of the audio adjustment parameter, and the population is composed of a preset number of particles. Each particle has a corresponding speed value and a position value.

[0159] The adjustment unit 1003 is configured to adjust the frequency and / or the amplitude of the original prompt audio according to the audio adjustment parameter, and obtain an output audio.

[0160] The prediction unit 1004 is configured to predict a prompt sound signal corresponding to the output audio based on a transfer path impulse response coefficient corresponding to a transfer path.

[0161] The calculation unit 1005 is configured to calculate the fitness of the audio adjustment parameter corresponding to the output audio based on the prompt sound signal.

[0162] The determination unit 1006 is configured to determine the optimal value of the audio adjustment parameter in each iteration according to the fitness.

[0163] The calibration unit 1007 is configured to calibrate the frequency and / or amplitude of the original prompt audio according to the current optimal value of the audio adjustment parameter if the current iteration number reaches the preset maximum iteration number, to obtain the target audio.

[0164] In the embodiment of the present application, the audio adjustment parameter is optimized through the iteration operation, so that the audio parameter of the prompt audio of the warning system is automatically calibrated, and the debugging efficiency of the warning system is improved, and the development process is accelerated.

[0165] Preferably, in combination with the content shown in the figure, the audio parameter calibration device further comprises an update unit configured to update the speed value and the position value corresponding to all particles in the group according to the optimal value of the audio adjustment parameter in each iteration, the preset cognitive learning factor, the preset social learning factor and the preset inertia weight, to obtain a new group, take the new group as the current group, and return to perform the iteration operation if the current iteration number does not reach the preset maximum iteration number. Figure 10

[0166] Preferably, in combination with the content shown in the figure, the prediction unit 1004 comprises a calculation module and an input module, and the implementation principles of each module are as follows: Figure 10 The calculation module is configured to calculate the transfer path impulse response coefficient between the preset controller output signal and the preset noise measurement point by using the adaptive filter.

[0167] The determination module is configured to determine the transfer function according to the transfer path impulse response coefficient.

[0168] The prediction module is configured to perform convolution calculation based on the output audio and the transfer path impulse response coefficient to obtain the prompt audio signal corresponding to the output audio.

[0169] Preferably, in combination with the content shown in the figure, the calculation module in the prediction unit 1004 comprises a playing sub-module, a collecting sub-module, a first calculation sub-module, a second calculation sub-module and an updating sub-module.

[0170] Figure 10 The playing sub-module is configured to play the white noise signal by using the preset controller to control the preset speaker arrangement point.

[0171] The collecting sub-module is configured to collect the sound signal corresponding to the white noise signal by using the microphone at the preset noise measurement point.

[0172] The first calculation sub-module is configured to input the white noise signal into the adaptive filter and calculate the output signal of the adaptive filter.

[0173] The second calculation sub-module is configured to calculate the transfer path impulse response coefficient between the output signal of the adaptive filter and the sound signal collected by the microphone at the preset noise measurement point.

[0174] ​​A second calculating sub-module is configured to calculate a difference between the sound signal and the output signal.

[0175] An updating sub-module is configured to update the weight coefficient of the preset filter based on the white noise signal and the difference by using the adaptive filtering algorithm until the adaptive filtering algorithm converges to a steady state, so as to obtain a transfer path impulse response coefficient between the preset controller output signal and the preset noise measuring point.

[0176] Preferably, in combination with Figure 10 As shown, the computing unit 1005 includes a sound pressure level calculating module, a first statistical module and a second statistical module.

[0177] The sound pressure level calculating module is configured to calculate a total sound pressure level corresponding to the prompt tone signal and sound pressure levels of a plurality of preset octaves.

[0178] The first statistical module is configured to count a first quantity and a second quantity of sound pressure levels that meet preset conditions from the sound pressure levels of the plurality of preset octaves based on preset frequency requirements.

[0179] The fitness calculating module is configured to calculate a fitness of the audio adjustment parameter corresponding to the output audio based on the total sound pressure level, the first quantity and the second quantity.

[0180] Preferably, in combination with Figure 10 As shown, the first statistical module in the second computing unit 1005 includes:

[0181] The first comparison sub-module is configured to compare the sound pressure level of each preset octave and a standard sound pressure level of the preset octave based on preset frequency requirements, the preset frequency requirements at least including the standard sound pressure level of each preset octave and a standard frequency.

[0182] The marking sub-module is configured to mark the sound pressure level of the preset octave as a qualified sound pressure level if the sound pressure level of the preset octave is greater than the standard sound pressure level of the preset octave.

[0183] The second comparison sub-module is configured to compare a center frequency corresponding to the qualified sound pressure level with the standard frequency.

[0184] The statistical sub-module is configured to count a number of the qualified sound pressure levels that meet a first requirement and mark the number as the first quantity, and count a number of the qualified sound pressure levels that meet a second requirement and mark the number as the second quantity, the first requirement being that the center frequency corresponding to the qualified sound pressure level is not greater than the standard frequency, and the second requirement being that the center frequency corresponding to the qualified sound pressure level is greater than the standard frequency.

[0185] In the embodiment of the present application, the audio adjustment parameter is optimized through the iteration operation, so that the audio parameter of the prompt tone of the warning system is automatically calibrated, and the debugging efficiency of the warning system is improved, and the development process is accelerated.

[0186] In summary, the embodiment of the present application provides an audio parameter calibration method and device of a warning system, which realizes automatic calibration of audio parameters of a prompt sound of the warning system by optimizing the audio adjustment parameters through iterative operation, and improves the debugging efficiency of the warning system and speeds up the development process.

[0187] Each embodiment in the specification is described in a progressive manner, and the same or similar parts between each embodiment can be referred to each other. Each embodiment focuses on the difference from other embodiments. In particular, for the system or system embodiments, since it is basically similar to the method embodiments, it is described more simply, and the related parts can be referred to the part of the method embodiments. The above-described system and system embodiments are only illustrative, and the units described as separate components can be or can not be physically separated, and the components displayed as units can be or can not be physical units, that is, they can be located in one place or distributed on multiple network units. Part or all of the modules can be selected to achieve the purpose of the embodiment scheme according to actual needs. Those skilled in the art can understand and implement without creative labor.

[0188] The skilled person can further realize that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be realized in electronic hardware, computer software or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the components and steps of each example have been described in the above description. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical scheme. The skilled person can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0189] The above description of the disclosed embodiments enables a person skilled in the art to implement or use the present application. Various modifications to the embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An audio parameter calibration method for a warning system, characterized in that, The method comprises: acquiring an original prompt audio; performing an iteration operation, assigning a current population to an audio adjustment parameter, the audio adjustment parameter comprising a frequency adjustment factor and / or an amplitude adjustment factor, the population being a candidate value of the audio adjustment parameter, the population being composed of a preset number of particles, each of the particles having a corresponding speed value and a position value; adjusting the frequency and / or the amplitude of the original prompt audio according to the audio adjustment parameter to obtain an output audio; predicting a prompt sound signal corresponding to the output audio based on a transfer path impulse response coefficient corresponding to a transfer path, the prompt sound signal being a signal of the output audio propagating to a noise measurement point through the transfer path; calculating a fitness of the audio adjustment parameter corresponding to the output audio based on the prompt sound signal; determining an optimal value of the audio adjustment parameter in each iteration according to the fitness; if a current iteration number has reached a preset maximum iteration number, calibrating the frequency and / or the amplitude of the original prompt audio according to the current optimal value of the audio adjustment parameter to obtain a target audio.

2. The method of claim 1, wherein, The method further comprises: if the current iteration number has not reached the preset maximum iteration number, updating the speed value and the position value corresponding to all the particles in the population according to the optimal value of the audio adjustment parameter in each iteration, a preset cognitive learning factor, a preset social learning factor and a preset inertia weight to obtain a new population, taking the new population as the current population and returning to perform the iteration operation.

3. The method of claim 1, wherein, The method further comprises: calculating a transfer path impulse response coefficient between a preset controller output signal and a preset noise measurement point by using an adaptive filter; performing convolution calculation on the output audio and the transfer path impulse response coefficient to obtain the prompt sound signal corresponding to the output audio.

4. The method of claim 3, wherein, The method further comprises: playing a white noise signal at a preset speaker arrangement point by using a preset controller; collecting a sound signal corresponding to the white noise signal at a preset noise measurement point by using a microphone; inputting the white noise signal into an adaptive filter to calculate an output signal of the adaptive filter; calculating a difference value between the sound signal and the output signal; updating a weight coefficient of the adaptive filter by using an adaptive filtering algorithm based on the white noise signal and the difference value until the adaptive filtering algorithm converges to a steady state to obtain the transfer path impulse response coefficient between the preset controller output signal and the preset noise measurement point.

5. The method of claim 1, wherein, The method further comprises: calculating a total sound pressure level corresponding to the prompt sound signal and a sound pressure level of a plurality of preset octaves; based on a preset frequency requirement, counting a first number and a second number that meet a preset condition from the sound pressure levels of the plurality of preset octaves. Calculate fitness of the audio adjustment parameter corresponding to the output audio based on the total sound pressure level, the first number and the second number.

6. The method of claim 5, wherein, The first number and the second number of the sound pressure levels of the plurality of preset frequency bands meeting preset conditions are counted based on preset frequency requirements, and the method comprises the steps of: Comparing the sound pressure level of each of the plurality of preset frequency bands with the standard sound pressure level of the plurality of preset frequency bands based on preset frequency requirements, wherein the preset frequency requirements at least include the standard sound pressure level of each of the plurality of preset frequency bands and a standard frequency; If the sound pressure level of the preset frequency band is greater than the standard sound pressure level of the preset frequency band, the sound pressure level of the preset frequency band is marked as a standard sound pressure level; Comparing the center frequency corresponding to the standard sound pressure level with the standard frequency; Counting the number of the standard sound pressure levels meeting a first requirement and marking the number as a first number, and counting the number of the standard sound pressure levels meeting a second requirement and marking the number as a second number, wherein the first requirement is that the center frequency corresponding to the standard sound pressure level is not greater than the standard frequency, and the second requirement is that the center frequency corresponding to the standard sound pressure level is greater than the standard frequency.

7. An audio parameter calibration device for a warning system, characterized in that The device comprises: An acquisition unit configured to acquire an original prompt audio; An execution unit configured to perform an iteration operation, assign a current population to an audio adjustment parameter, the audio adjustment parameter comprising a frequency adjustment factor and / or an amplitude adjustment factor, the population being candidate values of the audio adjustment parameter, the population being composed of a preset number of particles, each of the particles having a corresponding speed value and a position value; An adjustment unit configured to adjust the frequency and / or the amplitude of the original prompt audio according to the audio adjustment parameter to obtain an output audio; A prediction unit configured to predict a prompt sound signal corresponding to the output audio based on a transfer path impulse response coefficient corresponding to a transfer path, the prompt sound signal being a signal of the output audio propagating to a noise measurement point through a propagation path; A calculation unit configured to calculate fitness of the audio adjustment parameter corresponding to the output audio based on the total sound pressure level, the first number and the second number. A determination unit configured to determine an optimal value of the audio adjustment parameter at each iteration according to the fitness; A calibration unit configured to calibrate the frequency and / or the amplitude of the original prompt audio according to the current optimal value of the audio adjustment parameter to obtain a target audio if a current iteration number has reached a preset maximum iteration number.

8. The apparatus of claim 7, wherein, The device further comprises: An update unit configured to update the speed value and the position value corresponding to all the particles in the population according to the optimal value of the audio adjustment parameter at each iteration, a preset cognitive learning factor, a preset social learning factor and a preset inertia weight to obtain a new population if the current iteration number has not reached the preset maximum iteration number, take the new population as a current population, and return to perform the iteration operation.

9. The apparatus of claim 7, wherein, The prediction unit comprises: A calculation module configured to calculate a transfer path impulse response coefficient between a pre-set controller output signal and a pre-set noise measurement point using an adaptive filter; A determination module configured to determine a transfer function based on the transfer path impulse response coefficient; A prediction module is configured to perform a convolution calculation based on the output audio and the transfer path impulse response coefficient to obtain a prompt tone signal corresponding to the output audio.

10. The apparatus of claim 9, wherein, The calculation module comprises: A playing sub-module is configured to control a pre-set loudspeaker arrangement point to play a white noise signal by using a pre-set controller; A collecting sub-module is configured to collect a sound signal corresponding to the white noise signal by using a microphone at a pre-set noise measuring point; A first calculation sub-module is configured to input the white noise signal into an adaptive filter to calculate an output signal of the adaptive filter; A second calculation sub-module is configured to calculate a difference between the sound signal and the output signal; An updating sub-module is configured to update a weight coefficient of the adaptive filter by using an adaptive filtering algorithm based on the white noise signal and the difference until the adaptive filtering algorithm converges to a steady state to obtain a transfer path impulse response coefficient between a pre-set controller output signal and a pre-set noise measuring point.

Images