Audio parameter calibration method and device of warning system

Abstract

The invention provides an audio parameter calibration method and device for a warning system, and the method comprises the steps: obtaining an original prompt audio; performing iterative operation, and assigning the current group to the audio adjustment parameter; adjusting the frequency and / or amplitude of the original prompt audio according to the audio adjustment parameter to obtain an output audio; predicting a prompt tone signal corresponding to the output audio based on the transmission path pulse response coefficient corresponding to the transmission path; calculating the fitness of an audio adjustment parameter corresponding to the output audio based on the prompt tone signal; determining the optimal value of each iteration of the audio adjustment parameter according to the fitness; and if the current number of iterations reaches the preset maximum number of iterations, calibrating 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 parameters are optimized through iterative operation, and automatic calibration of the audio parameters of the warning tone of the warning system is achieved; the debugging efficiency of the warning system is improved, and the development process is accelerated.

Description

Technical Field

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

[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, they produce less noise when driving at low speeds, making it difficult to alert pedestrians to the vehicle, increasing the incidence of traffic accidents. In order to reduce safety hazards, the state has formulated relevant laws and regulations / standards requiring electric vehicles to be equipped with pedestrian warning systems to achieve the purpose of issuing warning prompts when electric vehicles are driving at low speeds. The general pedestrian warning system obtains vehicle condition information such as the current vehicle speed and gear position through the vehicle CAN bus, and the control module adjusts the frequency and / or amplitude of the audio signal accordingly and then outputs it to achieve low-speed prompt sound playback. In order to meet the requirements of relevant laws and 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 at present is to stabilize the vehicle speed by manually controlling the accelerator pedal, and manually adjust the parameters and conduct repeated tests to make the warning sound emitted by the speaker of the pedestrian warning system meet the requirements. However, this calibration method has the problems of relying on manual operation, being time-consuming, labor-intensive and inefficient. Summary of the invention

[0004] In view of this, an embodiment of the present invention provides an audio parameter calibration method and device for a warning system to solve the problem that the current audio parameter calibration method is time-consuming, labor-intensive and inefficient.

[0005] To achieve the above objectives, the embodiments of the present invention provide the following technical solutions:

[0006] A first aspect of an embodiment of the present invention discloses a method for calibrating audio parameters of a warning system, the method comprising:

[0007] Get the original prompt audio;

[0008] Perform an iterative operation to assign a current group to an audio adjustment parameter, wherein the audio adjustment parameter includes a frequency adjustment factor and / or an amplitude adjustment factor, the group is a candidate value of the audio adjustment parameter, the group is composed of a preset number of particles, and each of the particles has a corresponding speed value and position value;

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

[0010] Predicting a prompt sound signal corresponding to the output audio based on a transfer path impulse response coefficient corresponding to the transfer path, wherein the prompt sound signal is a signal transmitted from the output audio to a noise measurement point through the transfer path;

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

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

[0013] If the current number of iterations has reached a preset maximum number of iterations, 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 the target audio.

[0014] Preferably, the method further comprises:

[0015] If the current number of iterations does not reach the preset maximum number of iterations, the velocity values ​​and position values ​​corresponding to all the particles in the group are updated according to the optimal value of each iteration of the audio adjustment parameter, 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 execute the iterative operation.

[0016] Preferably, the predicting the prompt sound signal corresponding to the output audio based on the transfer path impulse response coefficient corresponding to the transfer path includes:

[0017] The adaptive filter is used to calculate the pulse response coefficient of the transfer path between the preset controller output signal and the preset noise measurement point;

[0018] A convolution calculation is performed based on the output audio and the transfer path impulse response coefficient to obtain a prompt tone signal corresponding to the output audio.

[0019] Preferably, the step of calculating the transfer path impulse response coefficient between a preset controller output signal and a preset noise measurement point by using an adaptive filter includes:

[0020] Using a preset controller to control a preset speaker arrangement point to play a white noise signal;

[0021] At a preset noise measurement point, using a microphone to collect an acoustic signal corresponding to the white noise signal;

[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 acoustic signal and the output signal;

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

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

[0026] Calculating the total sound pressure level and the sound pressure levels of several preset octaves corresponding to the prompt tone signal;

[0027] Based on a preset frequency requirement, counting a first quantity and a second quantity that meet a preset condition from the sound pressure levels of a plurality of the preset octave bands;

[0028] The adaptability 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.

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

[0030] Based on a preset frequency requirement, comparing the sound pressure level of each of the preset octaves with the standard sound pressure level of the preset octave, wherein the preset frequency requirement at least includes the standard sound pressure level and standard frequency of each of the preset octaves;

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

[0032] Comparing the center frequency corresponding to the qualified sound pressure level with the standard frequency;

[0033] The number of the standard sound pressure levels that meet the first requirement is counted and marked as a first number, and the number of the standard sound pressure levels that meet the second requirement is counted and marked 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.

[0034] A second aspect of an embodiment of the present invention discloses an audio parameter calibration device for a warning system, the device comprising:

[0035] An acquisition unit, used for acquiring original prompt audio;

[0036] an execution unit, configured to perform an iterative operation, assign a current group to an audio adjustment parameter, wherein the audio adjustment parameter includes a frequency adjustment factor and / or an amplitude adjustment factor, the group is a candidate value of the audio adjustment parameter, the group is composed of a preset number of particles, and each particle has a corresponding velocity value and position value;

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

[0038] 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 the transfer path, wherein the prompt sound signal is a signal transmitted from the output audio to a noise measurement point through a transmission path;

[0039] A calculation unit, configured to calculate the fitness of the audio adjustment parameter 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 parameter according to the fitness;

[0041] The calibration unit is used to calibrate 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 if the current number of iterations has reached a preset maximum number of iterations.

[0042] Preferably, the device further comprises:

[0043] The updating unit is used to update the velocity values ​​and position values ​​corresponding to all the particles in the group according to the optimal value of each iteration of the audio adjustment parameter, the preset cognitive learning factor, the preset social learning factor and the preset inertia weight if the current number of iterations does not reach the preset maximum number of iterations, to obtain a new group, take the new group as the current group, and return to execute the iterative operation.

[0044] Preferably, the prediction unit comprises:

[0045] A calculation module, used for calculating the pulse response coefficient of the transfer path 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] The prediction module is used to perform convolution calculation based on the output audio and the transfer path impulse response coefficient to obtain a prompt sound signal corresponding to the output audio.

[0048] Preferably, the calculation module includes:

[0049] A playing submodule, used for controlling a preset loudspeaker arrangement point to play a white noise signal using a preset controller;

[0050] A collection submodule, used to collect the acoustic signal corresponding to the white noise signal using a microphone at a preset noise measurement point;

[0051] A first calculation submodule, used for inputting the white noise signal into an adaptive filter and calculating an output signal of the adaptive filter;

[0052] A second calculation submodule, for calculating the difference between the acoustic signal and the output signal;

[0053] An updating submodule is used to update the weight coefficient of the preset filter based on the white noise signal and the difference using an adaptive filtering algorithm until the adaptive filtering algorithm converges to a steady state, thereby obtaining a transfer path impulse response coefficient between the preset controller output signal and the preset noise measurement point.

[0054] Based on the above-mentioned embodiment of the present invention, a method and device for calibrating the audio parameters of a warning system are provided to obtain the original prompt audio; perform an iterative operation to assign the current group to the audio adjustment parameter; adjust the frequency and / or amplitude of the original prompt audio according to the audio adjustment parameter to obtain the output audio; predict the prompt audio 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 audio signal; determine the optimal value of each iteration of the audio adjustment parameter according to the fitness; if the current number of iterations has reached the preset maximum number of iterations, calibrate 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. The audio adjustment parameters are optimized through iterative operations to achieve automatic calibration of the audio parameters of the warning system prompt sound; improve the debugging efficiency of the warning system and speed up the development process. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0056] Figure 1 A flowchart of an audio parameter calibration method for a warning system provided by an embodiment of the present invention;

[0057] Figure 2A schematic diagram of the minimum sound level limit for a 20 km / h forward driving condition at a constant speed provided by an embodiment of the present invention;

[0058] Figure 3 A schematic diagram of audio data x1[K] provided in an embodiment of the present invention;

[0059] Figure 4 A schematic diagram of audio data x2[K] provided in an embodiment of the present invention;

[0060] Figure 5 A schematic diagram of a transmission path between a loudspeaker arrangement point and a noise measurement point provided in an embodiment of the present invention;

[0061] Figure 6 An example diagram of an impulse response curve of a transfer path provided by an embodiment of the present invention;

[0062] Figure 7 An example diagram of an amplitude-frequency response curve of a transfer path provided by an embodiment of the present invention;

[0063] Figure 8 An example diagram of a phase-frequency response curve of a transfer path provided by an embodiment of the present invention;

[0064] Fig. 9 An example diagram of the 1 / 3 octave sound pressure level of the prompt tone signal corresponding to the output audio provided in an embodiment of the present invention;

[0065] Fig.10 A structural block diagram of an audio parameter calibration device for a warning system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0066] 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.

[0067] In this application, the terms "comprises", "comprising" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprising a ..." does not exclude the presence of other identical elements in the process, method, article or device comprising the element.

[0068] As can be seen from the background technology, the commonly used calibration method at present is to stabilize the vehicle speed by manually controlling the accelerator pedal, and manually adjust the parameters and conduct repeated tests so that the prompt sound emitted by the speaker of the pedestrian warning system meets the requirements. However, this calibration method has the problems of relying on manual operation, being time-consuming, labor-intensive and inefficient.

[0069] Therefore, an embodiment of the present invention provides an audio parameter calibration method and device for a warning system, which obtains the original prompt audio; performs an iterative operation to assign the current group to the audio adjustment parameter; adjusts the frequency and / or amplitude of the original prompt audio according to the audio adjustment parameter to obtain the output audio; predicts the prompt audio signal corresponding to the output audio based on the transfer path impulse response coefficient corresponding to the transfer path; calculates the fitness of the audio adjustment parameter corresponding to the output audio based on the prompt audio signal; determines the optimal value of the audio adjustment parameter for each iteration according to the fitness; if the current number of iterations has reached the preset maximum number of iterations, calibrates 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. By optimizing the audio adjustment parameters through iterative operations, the audio parameters of the warning system prompt sound can be automatically calibrated, which improves the debugging efficiency of the warning system and speeds up the development process.

[0070] See also Figure 1 , shows a flow chart of an audio parameter calibration method for a warning system provided by an embodiment of the present invention.

[0071] It should be noted that in the embodiment of the present invention, based on the national standard "Low-speed Prompt Sound of Electric Vehicles" (GB / T37153-2018), the audio parameters of the pedestrian warning system are calibrated for the 20km / h uniform forward driving condition to achieve the following technical indicators:

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

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

[0074] It can be understood that the audio parameter calibration method provided according to the embodiment of the present invention can also be used to calibrate the audio parameters of the pedestrian warning system for uniform forward driving conditions of other national standards such as 10km / h. The specific implementation method is similar to the implementation process described in the embodiment of the present invention, and no redundant introduction is given here.

[0075] The audio parameter calibration method comprises:

[0076] Step S101: Obtain original prompt audio.

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

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

[0079] In the specific implementation of step S102, a specific iterative operation is performed to assign the position values ​​corresponding to the preset number of particles in the current group to the audio adjustment parameter. In other words, the group is a candidate value of the audio adjustment parameter, and the group is composed of a preset number of particles, wherein each particle has a corresponding speed value and position value.

[0080] It should be noted that the current population is generated in advance based on the particle swarm optimization algorithm. Specifically, the particle swarm optimization algorithm generates a preset number of particles, initializes the speed value and position value corresponding to each particle, and obtains the current population. For example, the speed 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 applications, the audio adjustment parameters include frequency adjustment factors and / or amplitude adjustment factors. Define the frequency adjustment factor F1 and amplitude adjustment factor A1 corresponding to the audio x1[K] with K samples, and the frequency adjustment factor F2 and 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 group is assigned to the audio adjustment parameters F1, A1, F2 and A2.

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

[0084] In the specific implementation of 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 an interpolation method) to obtain the output audio.

[0085] It is understandable that defining the original prompt audio x i The index of [K] is k = 1, 2, ..., K, and the output audio is y i The index of [L] is = 1, 2, ..., L; i = 1, 2; assuming that the frequency of the original audio is magnified by F i The output after multiplication is g i On this basis, the amplitude is enlarged by A i times, which is the final output audio. Define the interpolation point p i , initialize it to 1.

[0086] Then output y i The calculation steps of [L] are as follows:

[0087] 1) Determine the interpolation point p i Does the rounded value belong to index k (k = 1, 2, 3, ...), if so, take out the original prompt audio x i The kth element of [K] is denoted by g i , multiply it by A i , which is the lth sample y of the output audio i [L];

[0088] 2) Determine the interpolation point p i Check whether the rounded value belongs to index k (k = 1, 2, 3, ...). If not, perform interpolation calculation. i The value after rounding up is recorded as r 1i , p i The value after rounding down is recorded as r 0i , take out the original prompt audio x i [K] 1iand r 0i elements, denoted by z 1i and z 0i . Using (r 0i ,z 0i ) and (r 1i ,z 1i ) can calculate the interpolation point p i The value on the straight line or curve passing through the two points is denoted as g i , multiply it by A i , which is the output audio sample y i [L].

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

[0090] 4) Repeat the above steps.

[0091] Step S104: predicting a prompt tone signal corresponding to the output audio based on a transfer path impulse response coefficient corresponding to the transfer path.

[0092] In the process of implementing step S104, the controller, speaker arrangement points and noise measurement points are pre-set, and the transfer path impulse response coefficient between the controller output signal and the noise measurement point is calculated using an adaptive filter; and a convolution calculation is performed based on the output audio and the transfer path impulse response coefficient to obtain a prompt sound signal corresponding to the output audio.

[0093] It should be noted that, based on the output audio and the transfer path impulse response coefficient, a convolution calculation is performed using the following formula (1) to obtain a prompt tone signal d(n) corresponding to the output audio.

[0094]

[0095] Wherein, y(n) is the audio output signal at time n; is the transfer path impulse response coefficient; For FIR filter The order of .

[0096] It is understandable that, referring to the embodiments of the present invention Figure 5 The content shown in the figure, the transmission path between the preset controller output signal and the noise measurement point is specifically: the electronic circuit between the control module of the controller and the speaker arrangement point, and the sound channel between the speaker arrangement point and the noise measurement point (i.e., the "microphone" shown in the figure). Among them, the control module is used to control the speaker to play the white noise signal. The control module and the speaker are arranged in the body of the electric vehicle, and the specific arrangement position depends on the actual situation and is not limited here.

[0097] The following specifically describes the process of calculating the transfer path impulse response coefficient corresponding to the transfer path between the preset controller output signal and the noise measurement point:

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

[0099] Input the white noise signal v(n) into the preset filter (For example, the 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 filter The order of For FIR filter The filter coefficients at the nth time instant.

[0102] According to formula (3), the acoustic signal d is calculated v (n) (i.e. the expected signal) and the output signal y v The difference 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 adjust the preset filter The weight coefficients are updated until the adaptive filtering algorithm converges to a steady state, and the impulse response coefficients of the transfer path between the loudspeaker arrangement point and the noise measurement point are obtained.

[0105]

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

[0107] It should be noted that, in combination with the above content, the example diagram of the pulse response curve of the transmission path can be referred to the embodiment of the present invention. Figure 6 The example diagram of the amplitude-frequency response curve of the transmission path can be found in the embodiment of the present invention. Figure 7 The example diagram of the phase-frequency response curve of the transmission path can be found in the embodiment of the present invention. Figure 8 .

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

[0109] In the process of specifically implementing step S105, the total sound pressure level corresponding to the prompt sound signal of the noise measurement point and the sound pressure levels of several preset octave bands are calculated; based on the preset frequency requirement, a first quantity and a second quantity that meet the preset conditions are counted from the sound pressure levels of several preset octave bands; and the adaptability of the audio adjustment parameters corresponding to the output audio is calculated based on the total sound pressure level, the first quantity and the second quantity.

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

[0111] It should be noted that the specific implementation process of calculating the total sound pressure level corresponding to the prompt sound signal at the noise measurement point and the sound pressure levels of several 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 needs to be corrected, which will not be explained in detail here.

[0114]

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

[0116] According to formula (6), the sound pressure level of several preset octave bands (for example, the sound pressure level of 1 / 3 octave band) corresponding to the prompt tone signal at the noise measurement point is calculated:

[0117]

[0118] Among them, f u represents the upper frequency limit; f l Represents the lower frequency limit; f cRepresents the center frequency.

[0119] It can be understood that, based on the preset frequency requirement, the specific implementation process (process A1 to process A4) of counting the first quantity and the second quantity that meet the preset conditions from the sound pressure levels of a plurality of preset octave bands is as follows:

[0120] A1: Based on the preset frequency requirements, compare the sound pressure level of each preset octave band with the standard sound pressure level of the preset octave band.

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

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

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

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

[0125] For example Figure 2 As shown:

[0126] Compare the sound pressure level a1dB(A) of the 1 / 3 octave band with a center frequency of 160Hz to see if it is greater than the standard sound pressure level of 52dB(A) of the 1 / 3 octave band with a center frequency of 160Hz.

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

[0128] Compare the sound pressure level of a2dB(A) at 1 / 3 octave with a center frequency of 2500Hz to see if it is greater than the standard sound pressure level of 46dB(A) at 1 / 3 octave with a center frequency of 2500Hz.

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

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

[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 1 / 3 octave band with the current center frequency of 800Hz does not meet the standard.

[0132] Compare the sound pressure level of 4dB(A) at 1 / 3 octave with a center frequency of 400Hz to see if it is greater than the standard sound pressure level of 52dB(A) at 1 / 3 octave with a center frequency of 400Hz.

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

[0134] Compare the center frequency of the sound pressure level that meets the standard with the standard frequency, such as comparing the center frequency 160Hz of the sound pressure level a1dB with 1600Hz, the center frequency 2500Hz of the sound pressure level a2dB with 1600Hz, and the center frequency 400Hz of the sound pressure level a4dB with 1600Hz.

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

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

[0137] The objective function is established to calculate the fitness, which is expressed as shown in the following formula (7).

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

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

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

[0141] In the specific implementation of step S106, according to the principle that the smaller the fitness value, the better, the individual optimal position and the global optimal position of each particle in the current group are determined, that is, the optimal value of the audio adjustment parameter for each iteration is determined.

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

[0143] Step S107: If the current number of iterations has reached the preset maximum number of iterations, 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 the target audio.

[0144] In the specific implementation of step S107, if the current number of iterations has reached the preset maximum number of iterations, 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 1 / 3 octave sound pressure level of the prompt sound at the noise measurement point 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, combined with the content described in the above example, when determining the optimal value of the audio adjustment parameter, the frequency adjustment factor F1 = 1.16205653348446 and the amplitude adjustment factor A1 = 0.403482731700248 corresponding to the audio data x1[K], the frequency adjustment factor F2 = 4.04467877313752 and the amplitude adjustment factor A2 = 0.309339452322305 corresponding to the audio data x2[K]. At the same time, the total sound pressure level of the prompt sound signal corresponding to the output audio obtained based on the preset transfer function is 59.999938dB(A), and the first and second numbers that meet the preset conditions are counted from several preset octave sound pressure levels, and the results are as follows: Fig. 9 As shown, there is one frequency band below 1600 Hz that meets the sound level limit requirements, with its center frequency at 1600 Hz and a sound pressure level of 59.373541 dB (A); in the frequency band from 1600 Hz to 5000 Hz, there is one frequency band that meets the sound level limit requirements, with its center frequency at 2000 Hz and a sound pressure level of 50.210700 dB (A).

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

[0147] That is to say, if the current number of iterations does not reach the preset maximum number of iterations, the optimal value of each iteration of the audio adjustment parameter is input into the following update formula (8) and update formula (9) to adjust the position and velocity values ​​of the particles 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] Among them, c1 is the preset cognitive learning factor, c2 is the preset social learning factor, r1 and r2 are random numbers in [0, 1], and W(t) is the 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 specifically calculated according to the following formula (10):

[0152]

[0153] Among them, t max is the maximum number of iterations, 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 update formula, each particle is directed toward the optimal position to obtain a new group, and the new group is used as the current group, and the iteration operation is returned to perform, that is, step S102.

[0155] In the embodiment of the present invention, the audio adjustment parameters are optimized through iterative operations, so that the audio parameters of the warning system prompt sound are automatically calibrated, which improves the debugging efficiency of the warning system and speeds up the development process.

[0156] Corresponding to the audio parameter calibration method of the warning system provided by the above embodiment of the present invention, see Fig.10 , shows a structural block diagram of an audio parameter calibration device for a warning system provided by an embodiment of the present invention. The audio parameter calibration device includes: 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 used to acquire the original prompt tone.

[0158] The execution unit 1002 is used to perform an iterative operation to assign the current group to the audio adjustment parameter, where the audio adjustment parameter includes a frequency adjustment factor and / or an amplitude adjustment factor. The group is a candidate value of the audio adjustment parameter. The group is composed of a preset number of particles, and each particle has a corresponding speed value and position value.

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

[0160] The prediction unit 1004 is used to predict the prompt sound signal corresponding to the output audio based on the transfer path impulse response coefficient corresponding to the transfer path, where the prompt sound signal is a signal transmitted from the output audio to the noise measurement point through the transmission path.

[0161] The calculation unit 1005 is used 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 used to determine the optimal value of the audio adjustment parameter for each iteration according to the fitness.

[0163] The calibration unit 1007 is used to calibrate 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 if the current iteration number has reached the preset maximum iteration number.

[0164] In the embodiment of the present invention, the audio adjustment parameters are optimized through iterative operations, so that the audio parameters of the warning system prompt sound are automatically calibrated, which improves the debugging efficiency of the warning system and speeds up the development process.

[0165] Preferably, combined Fig.10 The content shown in the figure shows that the audio parameter calibration device also includes: an updating unit, which is used to update the speed values ​​and position values ​​corresponding to all particles in the group according to the optimal value of each iteration of the audio adjustment parameter, the preset cognitive learning factor, the preset social learning factor and the preset inertia weight if the current number of iterations does not reach the preset maximum number of iterations, to obtain a new group, use the new group as the current group, and return to execute the iterative operation.

[0166] Preferably, combined Fig.10 The prediction unit 1004 includes: a calculation module and an input module. The implementation principle of each module is as follows:

[0167] The calculation module is used to calculate the pulse response coefficient of the transfer path between the preset controller output signal and the preset noise measurement point by using the adaptive filter.

[0168] The determining module is used to determine the transfer function according to the transfer path impulse response coefficient.

[0169] The prediction module is used to perform 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.

[0170] Preferably, combined Fig.10 As shown in the content, the calculation module in the prediction unit 1004 includes: a playing submodule, a collecting submodule, a first calculation submodule, a second calculation submodule and an updating submodule.

[0171] The playing submodule is used to control the preset speaker arrangement points to play the white noise signal using a preset controller.

[0172] The acquisition submodule is used to collect the acoustic signal corresponding to the white noise signal using a microphone at a preset noise measurement point.

[0173] The first calculation submodule is used to input the white noise signal into the adaptive filter and calculate the output signal of the adaptive filter.

[0174] The second calculation submodule is used to calculate the difference between the acoustic signal and the output signal.

[0175] The updating submodule is used to update the weight coefficients of the preset filter based on the white noise signal and the difference using the adaptive filtering algorithm until the adaptive filtering algorithm converges to a steady state, and obtain the transfer path impulse response coefficient between the preset controller output signal and the preset noise measurement point.

[0176] Preferably, combined Fig.10 As shown in the figure, the calculation unit 1005 includes a sound pressure level calculation module, a first statistical module and a second statistical module.

[0177] The sound pressure level calculation module is used to calculate the total sound pressure level corresponding to the prompt sound signal and the sound pressure levels of several preset octaves.

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

[0179] The fitness calculation module is used to calculate the 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, combined Fig.10 The content shown, the first statistical module in the second computing unit 1005 includes:

[0181] The first comparison submodule is used to compare the sound pressure level of each preset octave with the standard sound pressure level of the preset octave based on the preset frequency requirement, and the preset frequency requirement at least includes the standard sound pressure level and standard frequency of each preset octave.

[0182] The marking submodule is used to mark the sound pressure level of the preset octave band as the standard sound pressure level if the sound pressure level of the preset octave band is greater than the standard sound pressure level of the preset octave band.

[0183] The second comparison submodule is used to compare the center frequency corresponding to the qualified sound pressure level with the standard frequency.

[0184] The statistical submodule is used to count the number of standard sound pressure levels that meet the first requirement, marked as the first number, and count the number of standard sound pressure levels that meet the second requirement, marked as the second number. 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.

[0185] In the embodiment of the present invention, the audio adjustment parameters are optimized through iterative operations, so that the audio parameters of the warning system prompt sound are automatically calibrated, which improves the debugging efficiency of the warning system and speeds up the development process.

[0186] In summary, the embodiments of the present invention provide a method and device for calibrating audio parameters of a warning system, which optimizes audio adjustment parameters through iterative operations to achieve automatic calibration of audio parameters of the warning system prompt sound; improve the debugging efficiency of the warning system and speed up the development process.

[0187] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can refer to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system or system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can refer to the partial description of the method embodiment. The system and system embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without creative work.

[0188] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

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

Claims

1. A method for calibrating audio parameters of a warning system, characterized in that: The method comprises: Get the original prompt audio; Perform an iterative operation to assign a current group to an audio adjustment parameter, wherein the audio adjustment parameter includes a frequency adjustment factor and / or an amplitude adjustment factor, the group is a candidate value of the audio adjustment parameter, the group is composed of a preset number of particles, and each of the particles has a corresponding speed value and position value; Adjust the frequency and / or amplitude of the original prompt audio according to the audio adjustment parameter to obtain output audio; Predicting a prompt sound signal corresponding to the output audio based on a transfer path impulse response coefficient corresponding to the transfer path, wherein the prompt sound signal is a signal transmitted from the output audio to a noise measurement point through the transfer path; Calculating the fitness of the audio adjustment parameter corresponding to the output audio based on the prompt sound signal; Determining the optimal value of each iteration of the audio adjustment parameter according to the fitness; If the current number of iterations has reached a preset maximum number of iterations, 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 the target audio.

2. The method according to claim 1, characterized in that The method further comprises: If the current number of iterations does not reach the preset maximum number of iterations, the velocity values ​​and position values ​​corresponding to all the particles in the group are updated according to the optimal value of each iteration of the audio adjustment parameter, 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 execute the iterative operation.

3. The method according to claim 1, characterized in that The predicting the prompt sound signal corresponding to the output audio based on the transfer path impulse response coefficient corresponding to the transfer path includes: The adaptive filter is used to calculate the pulse response coefficient of the transfer path between the preset controller output signal and the preset noise measurement point; A convolution calculation is performed based on the output audio and the transfer path impulse response coefficient to obtain a prompt tone signal corresponding to the output audio.

4. The method according to claim 3, characterized in that The method of calculating the pulse response coefficient of the transfer path between the preset controller output signal and the preset noise measurement point by using the adaptive filter includes: Using a preset controller to control a preset speaker arrangement point to play a white noise signal; At a preset noise measurement point, using a microphone to collect an acoustic signal corresponding to the white noise signal; Inputting the white noise signal into an adaptive filter, and calculating an output signal of the adaptive filter; calculating a difference between the acoustic signal and the output signal; Based on the white noise signal and the difference, the weight coefficient of the preset filter is updated using an adaptive filtering algorithm until the adaptive filtering algorithm converges to a steady state, thereby obtaining a transfer path impulse response coefficient between the preset controller output signal and the preset noise measurement point.

5. The method according to claim 1, characterized in that The calculating the fitness of the audio adjustment parameter corresponding to the output audio based on the prompt tone signal includes: Calculating the total sound pressure level and the sound pressure levels of several preset octaves corresponding to the prompt tone signal; Based on a preset frequency requirement, counting a first quantity and a second quantity that meet a preset condition from the sound pressure levels of a plurality of the preset octave bands; The adaptability 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.

6. The method according to claim 5, characterized in that The step of counting the first quantity and the second quantity that meet the preset conditions from the sound pressure levels of a plurality of preset octave bands based on the preset frequency requirement includes: Based on a preset frequency requirement, comparing the sound pressure level of each of the preset octaves with the standard sound pressure level of the preset octave, wherein the preset frequency requirement at least includes the standard sound pressure level and standard frequency of each of the preset octaves; If the sound pressure level of the preset octave band is greater than the standard sound pressure level of the preset octave band, marking the sound pressure level of the preset octave band as the standard sound pressure level; Comparing the center frequency corresponding to the qualified sound pressure level with the standard frequency; The number of the standard sound pressure levels that meet the first requirement is counted and marked as a first number, and the number of the standard sound pressure levels that meet the second requirement is counted and marked 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, used for acquiring original prompt audio; an execution unit, configured to perform an iterative operation, assign a current group to an audio adjustment parameter, wherein the audio adjustment parameter includes a frequency adjustment factor and / or an amplitude adjustment factor, the group is a candidate value of the audio adjustment parameter, the group is composed of a preset number of particles, and each particle has a corresponding velocity value and position value; An adjustment unit, configured to adjust the frequency and / or 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 the transfer path, wherein the prompt sound signal is a signal transmitted from the output audio to a noise measurement point through a transmission path; A calculation unit, configured to calculate the fitness of the audio adjustment parameter corresponding to the output audio based on the prompt tone signal; A determination unit, configured to determine an optimal value of each iteration of the audio adjustment parameter according to the fitness; The calibration unit is used to calibrate 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 if the current number of iterations has reached a preset maximum number of iterations.

8. The device according to claim 7, characterized in that The device also includes: The updating unit is used to update the velocity values ​​and position values ​​corresponding to all the particles in the group according to the optimal value of each iteration of the audio adjustment parameter, the preset cognitive learning factor, the preset social learning factor and the preset inertia weight if the current number of iterations does not reach the preset maximum number of iterations, to obtain a new group, take the new group as the current group, and return to execute the iterative operation.

9. The device according to claim 7, characterized in that The prediction unit comprises: A calculation module, used for calculating the pulse response coefficient of the transfer path between a preset controller output signal and a preset noise measurement point by using an adaptive filter; A determination module, configured to determine a transfer function according to the transfer path impulse response coefficient; The prediction module is used to perform convolution calculation based on the output audio and the transfer path impulse response coefficient to obtain a prompt sound signal corresponding to the output audio.

10. The device according to claim 9, characterized in that The computing module comprises: A playing submodule, used for controlling a preset loudspeaker arrangement point to play a white noise signal using a preset controller; A collection submodule, used to collect the acoustic signal corresponding to the white noise signal using a microphone at a preset noise measurement point; A first calculation submodule, used for inputting the white noise signal into an adaptive filter and calculating an output signal of the adaptive filter; A second calculation submodule, for calculating the difference between the acoustic signal and the output signal; An updating submodule is used to update the weight coefficient of the preset filter based on the white noise signal and the difference using an adaptive filtering algorithm until the adaptive filtering algorithm converges to a steady state, thereby obtaining a transfer path impulse response coefficient between the preset controller output signal and the preset noise measurement point.

Images