Intelligent audio test system frequency response analysis method, device and equipment

Abstract

The invention relates to the technical field of audio testing, and discloses an intelligent audio testing system frequency response analysis method, device and equipment, and the method comprises the steps: carrying out the audio parameter initialization of an intelligent audio testing system, and obtaining a testing parameter matrix and an excitation signal; performing dual-channel acquisition and time-frequency feature extraction to obtain a time-frequency feature data set; executing double-branch processing of time domain noise reduction and frequency domain enhancement to obtain a purified feature data set and a feature weighting matrix; performing segmented recursive calculation and feature aggregation to obtain a frequency response feature vector; decomposing the frequency response feature vector according to a test frequency band, and performing environment compensation and nonlinear calibration to obtain a standard frequency response data set; and inputting the standard frequency response data set into the deep neural network for feature recognition and mode classification to obtain a frequency response curve and a phase characteristic curve, thereby improving the test precision, the anti-interference capability, the automation degree and the adaptability of the intelligent audio test system.

Description

Technical Field

[0001] The present application relates to the field of audio testing technology, and in particular to a frequency response analysis method, device and equipment for an intelligent audio testing system. Background Art

[0002] Traditional audio test systems mainly rely on manual operation and simple digital signal processing, and have obvious deficiencies in test accuracy, noise immunity, and automation. Especially in actual test environments, external factors such as environmental noise, temperature and humidity changes will significantly affect the accuracy of test results, and traditional test methods are difficult to effectively deal with these interferences.

[0003] Current audio frequency response test methods generally have problems such as unstable signal quality, insufficient feature extraction, and low data processing efficiency. Especially in complex environments, test results are easily affected by noise interference and environmental changes, resulting in reduced measurement accuracy. At the same time, traditional test systems lack intelligent data processing capabilities and cannot automatically identify and compensate for various errors in the test process. Testers need to have rich professional experience to obtain reliable test results. Existing frequency response analysis methods often use a single signal processing algorithm, which makes it difficult to fully capture the characteristics of audio equipment in different frequency bands. Especially when dealing with complex features such as nonlinear distortion and phase response, the performance of traditional methods is not ideal. In addition, the lack of effective adaptive mechanisms and intelligent analysis methods makes it difficult for the test system to adapt to the test needs of different types of audio equipment, limiting the further improvement of test efficiency and accuracy. Summary of the invention

[0004] The present application provides a method, device and equipment for frequency response analysis of an intelligent audio test system, thereby improving the test accuracy, anti-interference ability, degree of automation and adaptability of the intelligent audio test system.

[0005] The first aspect of the present application provides a frequency response analysis method for an intelligent audio test system, the frequency response analysis method for an intelligent audio test system comprising: Initialize audio parameters of the intelligent audio test system to obtain a test parameter matrix and an excitation signal; Based on the test parameter matrix, dual-channel acquisition and time-frequency feature extraction are performed on the excitation signal and the response signal to obtain a time-frequency feature data set; Performing dual-branch processing of time domain noise reduction and frequency domain enhancement on the time-frequency feature data set to obtain a purified feature data set and a feature weighting matrix; Performing segmented recursive calculation and feature aggregation on the purified feature data set according to the feature weighting matrix to obtain a frequency response feature vector; Decomposing the frequency response characteristic vector according to the test frequency band, and performing environmental compensation and nonlinear calibration to obtain a standard frequency response data set; The standard frequency response data set is input into a deep neural network for feature recognition and pattern classification to obtain a frequency response curve and a phase characteristic curve.

[0006] A second aspect of the present application provides an intelligent audio test system frequency response analysis device, the intelligent audio test system frequency response analysis device comprising: An initialization module is used to initialize the audio parameters of the intelligent audio test system to obtain a test parameter matrix and an excitation signal; An acquisition module, used for performing dual-channel acquisition and time-frequency feature extraction on the excitation signal and the response signal based on the test parameter matrix to obtain a time-frequency feature data set; A processing module, used for performing a dual-branch process of time domain noise reduction and frequency domain enhancement on the time-frequency feature data set to obtain a purified feature data set and a feature weighting matrix; A calculation module, used for performing segmented recursive calculation and feature aggregation on the purified feature data set according to the feature weighting matrix to obtain a frequency response feature vector; A calibration module, used for decomposing the frequency response characteristic vector according to the test frequency band, and performing environmental compensation and nonlinear calibration to obtain a standard frequency response data set; The classification module is used to input the standard frequency response data set into a deep neural network for feature recognition and pattern classification to obtain a frequency response curve and a phase characteristic curve.

[0007] The third aspect of the present application provides an electronic device, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the electronic device executes the above-mentioned intelligent audio test system frequency response analysis method.

[0008] A fourth aspect of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the above-mentioned intelligent audio test system frequency response analysis method.

[0009] Compared with the prior art, the present application has the following beneficial effects: by introducing a dual-branch data processing architecture, the synergistic enhancement of time domain and frequency domain signals is achieved, effectively improving the system's anti-noise ability; by adopting a multi-level feature extraction and recursive analysis strategy, the extraction accuracy of frequency response features is significantly improved; the environmental compensation and nonlinear calibration mechanisms are innovatively designed to effectively eliminate the interference effects of environmental factors; the deep neural network is introduced for feature recognition and pattern classification, realizing the intelligence and automation of the test process. This method ensures the accuracy of full-band testing through segmented recursive calculation and feature aggregation technology; adopts a dual-stream attention network and residual learning mechanism to enhance the system's ability to capture complex frequency response features; through a multi-level calibration and compensation mechanism, the reliability of the test results is ensured; and finally, through the intelligent analysis of the deep learning model, high-precision frequency response curve reconstruction is achieved. The present invention has achieved significant improvements in test accuracy, anti-interference ability, degree of automation and adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] 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 some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0011] The structures, proportions, sizes, etc. illustrated in the drawings of this specification are only used to match the contents disclosed in the specification so as to facilitate understanding and reading by persons familiar with this technology. They are not used to limit the conditions under which the present invention can be implemented, and therefore have no substantive technical significance. Any structural modification, change in proportion or adjustment of size, without affecting the effects and purposes that can be achieved by the present invention, should still fall within the scope of the technical contents disclosed by the present invention.

[0012] Figure 1 It is a flow chart of a frequency response analysis method of an intelligent audio test system provided by an embodiment of the present invention; Figure 2 is a schematic block diagram of the structure of a frequency response analysis device for an intelligent audio test system provided by an embodiment of the present invention; Figure 3 It is a schematic block diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0013] 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 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.

[0014] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0015] It should also be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in this application specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0016] It should be further understood that the term "and / or" used in the specification and appended claims of this application refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations. Figure 1 , an embodiment of the frequency response analysis method of the intelligent audio test system in the embodiment of the present application includes: Step 100: Initialize audio parameters of the intelligent audio test system to obtain a test parameter matrix and an excitation signal; It is understandable that the execution subject of the present application may be an intelligent audio test system frequency response analysis device, or a terminal or a server, which is not limited here. The present application embodiment is described by taking a server as the execution subject as an example.

[0017] Specifically, the processor of the intelligent audio test system is configured with a digital signal processing (DSP) algorithm to optimize and enhance the audio processing capability of the system. In this process, an all-pass filter and a dynamic equalizer (DEQ) are loaded into the intelligent audio test system. The all-pass filter is used to keep the amplitude of the signal unchanged while adjusting its phase response, while the dynamic equalizer is used to adjust the frequency response in real time, so that the system's response to different frequency bands is more balanced and accurate. By loading these audio processing modules, DSP initialization parameters are obtained. Based on the DSP initialization parameters, the system sampling parameters are set to obtain basic sampling parameters. The basic sampling parameters include parameters related to audio signal acquisition such as sampling rate and bit depth. The reasonable setting of these parameters can effectively ensure the accuracy of signal acquisition and the stability of the system. According to the basic sampling parameters, the signal range of the 14 digital gain channels of the system is set to obtain a channel parameter set to ensure that the signal gain range of each channel is reasonable, so that each channel can operate within its predetermined dynamic range without overload or underload. The gain step is set for the channel parameter set to accurately control the amplitude adjustment of each gain channel, and a gain control matrix is ​​generated in this way. The setting of gain step enables the system to be tested under different gain conditions, covering a wider dynamic range to ensure that the test results are representative and comprehensive. Based on the gain control matrix, the dynamic range of the system is verified. By performing the noise floor index measurement, the dynamic range verification data obtained reflects the noise performance and dynamic response capability of the system under different gain conditions. According to the dynamic range verification data, the signal-to-noise ratio index is measured and calculated to obtain the performance parameters of the system. By calculating the signal-to-noise ratio, the anti-interference ability of the system in the face of different noise levels is evaluated. The system performance parameters and the gain control matrix are combined to obtain the test parameter matrix. The test parameter matrix contains all the key parameters of the system during the audio test process, which is used to guide and standardize the entire test process. Based on the generated test parameter matrix, the test signal is generated to obtain the excitation signal. The excitation signal is the core part of the audio test and is used to excite the audio system in order to analyze the frequency response characteristics and dynamic performance of the system. By reasonably configuring the test parameter matrix, the excitation signal can cover all the key frequency points of the test frequency band to ensure the comprehensiveness and accuracy of the test results. The generation of the excitation signal also needs to consider the nonlinear characteristics and dynamic response of the system so that the test can truly reflect the actual performance of the system.

[0018] Step 200: Perform dual-channel acquisition and time-frequency feature extraction on the excitation signal and the response signal based on the test parameter matrix to obtain a time-frequency feature data set; Specifically, the excitation signal and the response signal are converted to digital (A / D) with 24 bits, the sampling points are set to 2048 points / frame, and the signal is sampled 2048 times in each frame to obtain the original sampling data. The original sampling data is allocated to 14 channels based on the test parameter matrix to obtain channel sampling data. Each channel corresponds to a specific frequency band or signal feature, ensuring that the signals of each frequency band can be analyzed and processed independently, thereby improving the resolution and accuracy of different frequency components. The channel sampling data is subjected to 1dB step gain adjustment, and the amplitude of the signal is finely controlled by the processor to obtain gain control data. The 1dB step amplitude can fine-tune the signal under different gain states to ensure that the amplitude of the signal is always within a reasonable range during the test process, avoiding distortion or information loss caused by excessive or weak signals. The gain adjustment step ensures the stability and effectiveness of the subsequent signal processing steps by controlling the amplitude of the signal. The gain control data is input into the all-pass filter for signal preprocessing. The all-pass filter adjusts the phase response without changing the signal amplitude to optimize the time domain characteristics of the signal. By executing the dynamic equalization algorithm, adjusting the frequency response characteristics, eliminating the imbalance in the spectrum, and obtaining the filtered processing data. Performing real-time Fourier transform on the filtered signal, converting the time domain signal into the frequency domain signal, obtaining the frequency domain feature data, and obtaining the amplitude and phase information of the signal at different frequencies. Performing phase feature extraction on the frequency domain feature data, calculating the group delay and phase response, and obtaining the phase feature data. The group delay reflects the time delay when different frequency components pass through the system, while the phase response reflects the phase change of the signal. Performing information fusion operation on the frequency domain feature data and the phase feature data, and performing feature alignment processing on them, to obtain the fused feature data. The fused feature data is standardized to obtain the time-frequency feature data set. The feature values ​​are adjusted to the same dimensional range to avoid deviations caused by different feature scales. The standardized feature data set has better comparability and stability.

[0019] Step 300: performing dual-branch processing of time domain noise reduction and frequency domain enhancement on the time-frequency feature data set to obtain a purified feature data set and a feature weighting matrix; It should be noted that the time-frequency feature data set is branched, and the signal is divided into a time domain processing branch and a frequency domain processing branch to form a dual-branch data set. A 4-level wavelet decomposition operation is performed on the time domain branch in the dual-branch data set. The signal is decomposed into details and approximate components of different scales through multi-level decomposition. The threshold is set to 3 times the standard deviation of the noise to distinguish between signals and noise. In this way, the noise components in the signal are extracted at different scales, so that they are effectively suppressed to obtain time domain denoised data. At the same time, for the frequency domain branch in the dual-branch data set, a spectral subtraction operation is performed to enhance the frequency domain signal characteristics. In the spectral subtraction operation, the frequency-related suppression factor is set to 0.8, which means that 80% of the noise is suppressed in the signal spectrum, while the effective components of the signal are retained. The use of spectral subtraction effectively reduces the influence of background noise and enhances the spectral characteristics of the signal, making the frequency domain characteristics of the signal more prominent and clear. The obtained frequency domain enhanced data has a high spectral signal-to-noise ratio and frequency domain smoothness. The signal-to-noise ratio index is calculated based on the time domain noise reduction data, and the signal-to-noise ratio threshold is set to 40dB to ensure that the signal features are extracted more accurately when the noise level is low. The time domain weight coefficient is obtained through this calculation, which reflects the contribution of time domain processing to the signal. At the same time, the spectrum flatness parameter is calculated based on the frequency domain enhancement data, and the flatness deviation range is set to ±3dB to ensure that the spectrum characteristics have high consistency and smoothness. The frequency domain weight coefficient is obtained through this calculation, which reflects the effect of frequency domain enhancement processing and the smoothness of the signal. The time domain weight coefficient and the frequency domain weight coefficient are weighted and combined, and the weight normalization operation is performed to obtain the feature weighting matrix. The weights of different features are standardized so that the feature contributions of the time domain and frequency domain can be balanced in the subsequent feature fusion to avoid a certain feature occupying too much weight. Based on the feature weighting matrix, the time domain noise reduction data and the frequency domain enhancement data are feature fused to obtain a fused feature set. Based on indicators such as frequency response flatness, phase linearity, and group delay fluctuation, feature screening operations are performed on the fused feature set to obtain a purified feature data set. Frequency response flatness is used to evaluate the consistency of the system at different frequencies, phase linearity is used to evaluate whether the phase response of the system changes linearly, and group delay fluctuation reflects whether the delay characteristics of the system at different frequencies are stable. Through the evaluation of these indicators, the features that have a significant contribution to the frequency response characteristics of the system are screened out from the fusion feature set, and the insignificant or unhelpful parts are removed to obtain a purified feature data set.

[0020] Step 400: performing piecewise recursive calculation and feature aggregation on the purified feature data set according to the feature weighting matrix to obtain a frequency response feature vector; Specifically, the purified feature data set is divided into frequency bands, and the frequency range from 20Hz to 20kHz is divided into 128 frequency bands, each of which represents a specific frequency range for more refined signal analysis. Bandpass filtering is performed in each divided frequency band. The function of bandpass filtering is to retain the signal components in a specific frequency range while suppressing the noise and irrelevant signals in other frequency bands to obtain the frequency band division data. The frequency band division data is encoded and converted according to the feature weighting matrix to construct a 16-dimensional frequency encoding vector. In this encoding vector, the amplitude information occupies 8 dimensions, the phase information occupies 4 dimensions, and the time information occupies 4 dimensions. This encoding method can fully express the various characteristics of frequency, including the strength change of the signal (amplitude), the phase change characteristics (phase), and the dynamic characteristics in the time domain (time). The frequency encoding data finally obtained has multi-dimensional information description capabilities. Recursive sampling processing is performed on the frequency encoding data. The basic sampling density is set to 4 sampling points per octave in the conventional area to ensure the sampling uniformity and coverage of the signal in the entire frequency band. In the area where the frequency response changes suddenly, in order to capture the subtle changes of the signal, the sampling density is increased to 16 sampling points per octave. The dynamic adjustment of the sampling density can effectively deal with the area where the frequency response changes dramatically, making the sampling data more representative and obtaining multi-density sampling data. The multi-density sampling data is input into the three-level feature extraction unit for processing. The first-layer feature extraction unit performs time domain convolution operation to extract the transient characteristics of the signal. The transient characteristics reflect the rapid changes of the signal in time. The second-layer feature extraction unit performs frequency domain transformation to extract the harmonic characteristics of the signal. The harmonic characteristics reflect the structural nature of the signal in the frequency domain. The third-layer feature extraction unit extracts the correlation characteristics of the signal through correlation analysis, analyzes the intrinsic connection between different parts of the signal, and obtains three-dimensional feature data that comprehensively describes the time domain, frequency domain and internal correlation characteristics of the signal. A two-layer prediction unit is constructed for the three-dimensional feature data. The first layer processes the time series relationship through the long short-term memory network (LSTM). The advantage of LSTM is that it can remember the long-term dependency of the signal. The time series prediction value is obtained through the processing of this layer to describe the evolution trend of the signal in time. The second layer fuses the frequency domain information through a feedforward neural network to obtain the frequency response prediction value, which can describe the response characteristics of the signal in frequency. Through the joint processing of these two layers, the predicted feature data is obtained. A three-level residual network is constructed based on the predicted feature data to eliminate the deviation and error in the feature. The first-level residual network calculates the amplitude residual to eliminate the error in the signal amplitude and ensure the accuracy of the amplitude characteristics of the signal. The second-level residual network calculates the phase residual to ensure the consistency and accuracy of the signal in phase through phase correction. The third-level residual network calculates the delay residual, which reflects the delay characteristics of the signal during the propagation process. This step eliminates the error caused by time delay and obtains the residual feature data.The residual feature data is input into the feature fusion network, and the feature dimension reduction and optimization are performed through the fully connected layer. In the fully connected layer, the number of output layer nodes is set to 32 to ensure that the feature retains its main information components while reducing its dimension, and obtain optimized feature data. The optimized feature data is normalized to eliminate the influence of different dimensions between different features. The three-dimensional features are compressed into a one-dimensional vector through weighted summation operation, and finally the frequency response feature vector is obtained.

[0021] Step 500, decomposing the frequency response characteristic vector according to the test frequency band, and performing environmental compensation and nonlinear calibration to obtain a standard frequency response data set; Specifically, the frequency response feature vector is divided into three sections, 20Hz to 200Hz is defined as the low frequency section, 200Hz to 2kHz is defined as the mid-frequency section, and 2kHz to 20kHz is divided into the high frequency section to obtain the segmented frequency response data. Targeted analysis and processing are performed based on the sensitivity of the human ear to different frequencies and the performance characteristics of audio equipment in different frequency bands. An adaptive interval model is constructed based on the segmented frequency response data. By setting the confidence level to 95%, the maximum deviation of the frequency response data is calculated to obtain the interval boundary data of each frequency band. The interval boundary data is used to describe the range of changes in the frequency response characteristics in different frequency bands. The upper and lower limits of these changes are determined by calculating the maximum deviation to reflect whether the frequency response of the system is within the expected range. In order to describe the uncertainty and volatility of the frequency response characteristics, mixed probability modeling is performed on the interval boundary data, in which the weight of the Gaussian distribution is set to 0.7, and the weight of the uniform distribution is set to 0.3. The purpose of mixed probability modeling is to combine the central tendency (Gaussian distribution) and uniform change (uniform distribution) of the signal to better describe the distribution of the frequency response characteristics in different frequency bands and obtain probability distribution data. According to the influence of the actual test environment, an environmental compensation matrix is ​​constructed. Considering that temperature and humidity are important factors affecting the audio test results, the environmental compensation matrix is ​​constructed based on the temperature coefficient ±0.01dB / ℃ and the humidity coefficient ±0.02dB / 10%RH. The compensation operation is performed on the probability distribution data to obtain the environmental compensation data. Changes in temperature and humidity will affect the response characteristics of the audio equipment. The signal is adjusted through the compensation matrix to effectively reduce the error introduced by changes in environmental conditions, making the frequency response data more accurate and reliable. The environmental compensation data is compared with the 1kHz / -20dBFS reference signal, and the error parameters of the system are calculated to obtain the calibration reference data. The 1kHz / -20dBFS reference signal is a commonly used reference signal in audio testing. By comparing it with the compensation data, the error of the system under different conditions is evaluated. According to these error parameters, a 12th-order finite impulse response (FIR) filter is constructed to perform compensation operations for nonlinear errors and obtain compensation calibration data. The FIR filter is a filter used for signal processing. The higher the order, the stronger the signal adjustment ability. The 12th-order FIR filter can effectively compensate for the nonlinear error of the system, so that the calibrated data is more in line with the standard response. Based on the compensation calibration data, the segmented frequency response data is calibrated and corrected, and the correction coefficients of the low frequency band, mid-frequency band and high frequency band are calculated respectively to obtain the frequency band calibration data. The correction coefficients of different frequency bands reflect the amount of compensation required in different frequency ranges. Through precise correction, the frequency response deviation of the equipment within a specific frequency range can be eliminated, making the system response more ideal. After correcting the calibration data of each frequency band, the calibration data of the entire frequency band is integrated to ensure that the transition between the frequency bands is smooth and without mutation, so as to obtain a standard frequency response data set.

[0022] Step 600: Input the standard frequency response data set into the deep neural network for feature recognition and pattern classification to obtain a frequency response curve and a phase characteristic curve.

[0023] Specifically, the standard frequency response data set is input into the feature extraction network, which consists of an 8-layer convolutional neural network. Each convolution layer uses the ReLU activation function. The activation function can introduce nonlinearity and improve the network's ability to express complex patterns. A batch normalization layer is added to each convolution layer to stabilize the training process, accelerate convergence and alleviate the overfitting problem of the network. Through these convolution operations, multi-scale feature maps containing rich frequency response features are extracted to capture the detailed information of the signal in different time and frequency ranges. Feature pyramid operations are performed on the multi-scale feature maps. Through this operation, four feature layers with different resolutions are constructed, namely the original resolution layer and three feature layers with downsampling ratios of 1 / 2, 1 / 4, and 1 / 8. These feature layers reflect the characteristics of the signal at different scales. The use of feature pyramids ensures multi-level analysis of frequency response data, allowing the network to focus on global structure and local details at the same time. Through multi-level feature representation, a more refined frequency response feature description is obtained. The multi-level feature data is input into the two-stream attention network for processing. The dual-stream attention network consists of a frequency channel attention module and a temporal spatial attention module. The frequency channel attention module uses a 1×1 convolutional layer to extract channel weights. These weights dynamically adjust the contributions of different frequency channels, so that the network pays more attention to the frequency bands that are important to the frequency response. The temporal spatial attention module uses a 5×5 convolutional layer to extract spatial position weights, capture the important characteristics of the signal in time and space, and enhance the network's ability to focus on key features. Through the processing of the dual-stream attention network, attention feature data is obtained, which contains weight information of different channels and different positions. Feature aggregation is performed on the attention feature data, and feature maps of different resolutions are weighted fused through a cross-layer feature fusion module. This module learns the fusion weights between feature maps of different layers through a 1×1 convolutional layer, so that the weights of each layer can be reasonably allocated during the fusion process, avoiding feature loss or over-emphasis due to different scales. Through feature fusion, fused feature data is obtained. The fused feature data is input into the classification branch network, which contains three fully connected layers with 512, 256 and 128 nodes respectively. Each fully connected layer is followed by a Dropout layer with a dropout rate of 0.5, which is used to randomly drop half of the nodes to prevent overfitting and improve the generalization ability of the network. Through the processing of the fully connected layer, the frequency classification result is obtained, which is used to determine which type of frequency response mode the input frequency response data belongs to. At the same time, regression branch processing is performed on the fused feature data, and deep features are extracted through three layers of residual blocks. Each residual block contains two 3×3 convolutional layers and a short-circuit connection. The introduction of residual blocks aims to alleviate the gradient vanishing problem in deep network training. Through short-circuit connections, information can be transmitted more smoothly in the deep network, improving the learning ability of deep features. Through the processing of these three layers of residual blocks, regression feature data describing the changing trend of frequency response characteristics is obtained.The regression feature data is input into the dual-head output layer. The first output head contains two fully connected layers for predicting the frequency response value. Through this output head, the amplitude characteristics of the signal at different frequencies are obtained. The second output head contains two fully connected layers for predicting the phase response value and obtaining the phase change characteristics of the signal. Through the processing of the dual-head output layer, the complete prediction data of the frequency response and phase response are obtained, reflecting the frequency response characteristics and phase characteristics of the signal. The frequency classification results and response prediction data are processed continuously by the spline interpolation algorithm. The spline interpolation can construct a smooth curve between discrete prediction points, so that the frequency response curve and the phase characteristic curve are continuous, which meets the frequency response measurement requirements in practical applications, and finally the frequency response curve and the phase characteristic curve are obtained.

[0024] In the embodiment of the present application, by introducing a dual-branch data processing architecture, the coordinated enhancement of time domain and frequency domain signals is achieved, and the anti-noise ability of the system is effectively improved; the multi-level feature extraction and recursive analysis strategy is adopted to significantly improve the extraction accuracy of frequency response features; the environmental compensation and nonlinear calibration mechanism are innovatively designed to effectively eliminate the interference of environmental factors; the deep neural network is introduced for feature recognition and pattern classification to realize the intelligence and automation of the test process. This method ensures the accuracy of full-band testing through segmented recursive calculation and feature aggregation technology; adopts a dual-stream attention network and residual learning mechanism to enhance the system's ability to capture complex frequency response features; through a multi-level calibration and compensation mechanism, the reliability of the test results is ensured; and finally, through the intelligent analysis of the deep learning model, high-precision frequency response curve reconstruction is achieved. The present invention has achieved significant improvements in test accuracy, anti-interference ability, degree of automation and adaptability.

[0025] In a specific embodiment, the process of executing step 100 may specifically include the following steps: Perform DSP algorithm configuration on the processor of the intelligent audio test system, load the all-pass filter and dynamic equalizer DEQ into the intelligent audio test system, and obtain DSP initialization parameters; The system sampling parameters are set based on the DSP initialization parameters to obtain the basic sampling parameters, and the signal ranges of the 14 digital gain channels are set according to the basic sampling parameters to obtain the channel parameter set; Gain step settings are performed on the channel parameter set to obtain a gain control matrix, and the system dynamic range is verified based on the gain control matrix. The noise floor index is measured to obtain dynamic range verification data. According to the dynamic range verification data, the signal-to-noise ratio index is measured and calculated to obtain the system performance parameters, and the system performance parameters and the gain control matrix are combined to obtain the test parameter matrix. Based on the test parameter matrix, the test signal is generated to obtain the excitation signal.

[0026] Specifically, the DSP (digital signal processing) algorithm is configured on the processor of the intelligent audio test system, and the all-pass filter and dynamic equalizer modules are loaded into the intelligent audio test system. The all-pass filter is a filter that adjusts the signal phase while keeping the signal amplitude unchanged, and is expressed in the following transfer function form: ; in, is the transfer function of the filter, is a complex frequency variable, is a filter parameter used to adjust the phase response. The purpose of installing an all-pass filter is to optimize the phase response without affecting the amplitude characteristics. A dynamic equalizer is a filter that dynamically adjusts the frequency response. It is used to control the gain of a specific frequency band so that the signal response at each frequency is more balanced. The function of a dynamic equalizer is expressed using the following formula: ; in, is the gain at a certain frequency, is the basic gain, For frequency The gain adjustment at The gain of each frequency band is adjusted by After loading these modules into the DSP system, the DSP initialization parameters are obtained. Based on the DSP initialization parameters, the sampling parameters of the system are set to obtain the basic sampling parameters. The sampling parameters include key attributes such as sampling frequency and quantization bit depth. The sampling frequency is usually set to kHz, that is, 48,000 samples per second, to ensure sufficient time domain resolution for the audio signal. The quantization bit depth is set to Each sampling point represents Different levels are set to ensure high fidelity of the signal. Based on these basic sampling parameters, the signal range is set for the 14 digital gain channels in the system to obtain a channel parameter set. The gain range setting of each channel is designed to ensure that the signal will not be overloaded or underloaded under different gain conditions. Suppose the gain range is , the channel signal range is expressed by the following formula: ; in, is the output signal voltage, is the input signal voltage, is the gain, and its value range is and By setting a reasonable gain range, stable output characteristics can be maintained under various input conditions. Gain step settings are performed on the channel parameter set to obtain the gain control matrix. Signals at different gain levels are adjusted and tested step by step to find the best gain configuration. The step size of the gain step is set to 1dB, and the adjustment from minimum gain to maximum gain is expressed as: ; in, is the step size, usually set to 1dB, is the number of steps. The gain control matrix generated in this way contains the gain configuration of each channel under different gain states. The dynamic range of the system is verified according to the gain control matrix, and the noise floor index measurement is performed to obtain the dynamic range verification data. The dynamic range is expressed as the ratio of the maximum output voltage of the system to the minimum detectable output voltage, expressed as: ; in, is the dynamic range, is the maximum output voltage, is the noise floor voltage. By measuring the noise floor voltage of the system, the dynamic range of the system is calculated to evaluate the overall performance of the system. Based on the dynamic range verification data, the signal-to-noise ratio index is measured and calculated to obtain the performance parameters of the system. The calculation formula for the signal-to-noise ratio is: ; in, is the voltage of the effective signal, is the noise floor voltage. The signal-to-noise ratio is an important indicator for evaluating the quality of an audio system. The larger its value, the better the signal quality of the system. The system performance parameters are obtained by combining the system performance parameters and the gain control matrix to form a test parameter matrix that contains the performance of the system under different states. The test parameter matrix is ​​used to describe the performance of the system under different gains and different input conditions. It is expressed as a The matrix ,in is the number of gain steps, is the number of each performance parameter (such as dynamic range, signal-to-noise ratio, etc.). Each row of the matrix represents the system performance under a specific gain condition, and each column represents a specific performance parameter. Based on the test parameter matrix, the test signal is generated to obtain the excitation signal. The type of excitation signal is a swept frequency signal, whose frequency changes linearly or logarithmically with time, covering the entire frequency range of the audio test. The mathematical expression of the swept frequency signal is: ; in, The excitation signal is at time The value at is the signal amplitude, is the starting frequency, The generated stimulus signal is used as input to drive the audio device under test, thereby measuring its frequency response characteristics.

[0027] In a specific embodiment, the process of executing step 200 may specifically include the following steps: Perform 24-bit analog-to-digital conversion on the excitation signal and the response signal, set the number of sampling points to 2048 points / frame, obtain the original sampling data, and perform 14-channel signal allocation on the original sampling data based on the test parameter matrix to obtain channel sampling data; Perform 1dB step gain adjustment on the channel sampling data, control the signal amplitude through the processor, and obtain gain control data; The gain control data is input into the all-pass filter for signal preprocessing, and the dynamic equalization DEQ algorithm operation is performed to obtain the filtered processing data, and the filtered processing data is subjected to real-time Fourier transformation to obtain the frequency domain feature data; Perform phase characteristic extraction on the frequency domain feature data, calculate the group delay and phase response, obtain phase feature data, perform information fusion operation on the frequency domain feature data and the phase feature data, perform feature alignment processing, and obtain fused feature data; The fused feature data is standardized to obtain a time-frequency feature data set.

[0028] Specifically, a 24-bit analog-to-digital conversion (A / D conversion) is performed on the excitation signal and the response signal, and the 24-bit quantization depth divides the signal into The number of sampling points is set to 2048 points per frame, and the number of points for sampling the signal in each frame is 2048. The original sampled data obtained is a discrete representation of the signal in the time domain. The original sampled data is allocated to 14 channels based on the test parameter matrix. The test parameter matrix defines specific parameters for different channels, such as gain range, frequency band division, etc. Assuming there are 14 channels, each channel is used for analysis of a specific frequency range, and the allocation is mathematically expressed as: ; in, Indicates The output signal of each channel is is the signal processing function associated with the channel, is the original input signal. In this process, the signal is distributed to different channels for analysis and processing within their respective frequency ranges. The channel sampling data obtained is the specific performance of the signal in each channel. 1dB step gain adjustment is performed on the channel sampling data. Through the step-by-step gain adjustment, the characteristics of the signal under different gain conditions and the nonlinear behavior of the system are effectively tested. The gain adjustment is described by the following formula: ; in, is the signal after gain adjustment, is the gain factor, is the gain value in decibels (dB). The step size is 1dB, which means that in each round of adjustment, the gain value increases or decreases by 1dB. The processor controls the amplitude of the signal to obtain the signal performance under different gains, thereby obtaining the gain control data. The gain control data is input into the all-pass filter for signal preprocessing. The all-pass filter is used to adjust the phase response of the signal while keeping the amplitude unchanged. The transfer function of the all-pass filter is expressed as: ; in, is the transfer function of the filter, is a complex frequency variable, is a filter parameter used to adjust the phase response of the filter. After the all-pass filter adjusts the phase of the signal, it performs the dynamic equalization algorithm. The gain is dynamically adjusted according to the spectral characteristics of the signal to enhance specific frequency components and equalize the frequency response of the signal. The equalized signal is represented by the following formula: ; in, is the signal after equalization processing, is the dynamic gain factor, is the impulse response of the equalizer, is the number of frequency bands. Through equalization processing, the filtered data is obtained. The filtered data is subjected to real-time Fourier transform to convert the signal from the time domain to the frequency domain and obtain the frequency domain feature data. The formula of Fourier transform is: ; in, For the The spectrum value of the frequency component, is the number of Fourier transform points, is the time domain signal, is a complex exponential basis function. Through Fourier transform operation, the amplitude and phase characteristics of the signal in the frequency domain are obtained. The phase characteristics of the frequency domain feature data are extracted to calculate the group delay and phase response of the signal. The phase response describes the phase change of different frequency components of the signal, and the group delay is defined as the derivative of the phase response with respect to the frequency, expressed as: ; in, is the group delay, is the phase response, is the frequency. By calculating the group delay, the delay characteristics of the signal at different frequencies are determined to obtain the phase feature data. The frequency domain feature data and the phase feature data are fused and feature alignment is performed to effectively combine the two. The purpose of feature alignment is to ensure the correspondence between different types of features at the same time and frequency. Information fusion is expressed by the following formula:

[0029] in, To fuse feature data, is the frequency domain feature data, is the phase characteristic data, and is the fusion weight, which is used to control the relative importance of the two features. The fused feature data is standardized to obtain the time-frequency feature data set.

[0030] In a specific embodiment, the process of executing step 300 may specifically include the following steps: The time-frequency feature data set is branched, and the signal is divided into a time domain processing branch and a frequency domain processing branch to obtain a dual-branch data set; A 4-level wavelet decomposition operation is performed on the time domain branch in the dual-branch data set, and the threshold is set to 3 times the noise standard deviation to obtain the time domain denoised data; Perform spectral subtraction on the frequency domain branch in the dual-branch data set, set the frequency-related suppression factor to 0.8, and obtain frequency domain enhanced data; The signal-to-noise ratio index is calculated based on the time-domain noise reduction data, the signal-to-noise ratio threshold is set to 40 dB, and the time-domain weight coefficient is obtained. The spectrum flatness parameter is calculated based on the frequency-domain enhancement data, and the flatness deviation range is set to ±3 dB to obtain the frequency-domain weight coefficient. The time domain weight coefficient and the frequency domain weight coefficient are weighted and combined, and a weight normalization operation is performed to obtain a feature weight matrix. The time domain denoised data and the frequency domain enhanced data are feature fused according to the feature weight matrix to obtain a fused feature set. Based on the frequency response flatness, phase linearity and group delay fluctuation indicators, feature screening operations are performed on the fused feature set to obtain a purified feature data set.

[0031] Specifically, the time-frequency feature data set is branched, and the signal is divided into a time domain processing branch and a frequency domain processing branch to obtain a dual-branch data set. The time domain and frequency domain characteristics of the signal are processed separately to fully extract the important information of the signal in the time and frequency dimensions. Time domain processing is used to analyze and reduce instantaneous noise in the signal, while frequency domain processing is used to enhance the frequency characteristics of the signal and improve the flatness of the spectrum. In the time domain processing branch, a 4-level wavelet decomposition operation is performed on the time domain data to separate the effective information and noise components in the signal. Wavelet decomposition is a time-frequency localization method that decomposes the signal into details and approximate parts of different scales by using wavelet functions. Wavelet decomposition is expressed as the following formula: ; in, represents the original input signal, is a scaling function used to represent low-frequency components. is a wavelet function, which represents the detail components of different frequencies. and are the approximation and detail coefficients respectively. Among them, Represents the decomposition level, which is equal to 4. After the decomposition is completed, in order to remove the noise in the signal, the threshold is set to 3 times the standard deviation of the noise: ; in, represents the threshold of noise suppression, Represents the standard deviation of the noise. Soft threshold processing is performed on the wavelet coefficients, retaining the coefficients exceeding the threshold and suppressing the coefficients less than the threshold to obtain the data after time domain noise reduction. The soft threshold function is expressed as: ; in, is the processed detail coefficient, Indicates the sign of the coefficient. This processing effectively reduces noise interference and obtains a clean time domain signal. At the same time, in the frequency domain processing branch, a spectral subtraction operation is performed on the frequency domain data to enhance the frequency characteristics of the signal. Spectral subtraction is an effective noise reduction method that estimates the spectrum of the noise and subtracts it from the spectrum of the signal to achieve the effect of enhancing the signal. represents the spectrum of the input signal, represents the noise spectrum, then the spectrum subtraction operation is expressed as: ; in, is the signal after frequency domain enhancement, is the frequency-related suppression factor, which is set to 0.8, which means 80% suppression of noise. Through this operation, the proportion of noise in the spectrum is reduced, the clarity of the signal is improved, and the frequency domain enhanced data is obtained. The signal-to-noise ratio (SNR) is calculated based on the time domain noise reduction data, and the signal-to-noise ratio threshold is set to 40dB to obtain the time domain weight coefficient. The calculation formula for the signal-to-noise ratio is: ; Among them, SNR represents the signal-to-noise ratio, represents the power of the signal, Indicates the power of the noise. If the calculated signal-to-noise ratio is higher than 40dB, the signal quality is considered good, corresponding to a higher time domain weight coefficient. The value of the weight coefficient is linearly mapped according to the size of the signal-to-noise ratio, for example: ; in, is the time domain weight coefficient, is the maximum possible value of the signal-to-noise ratio. At the same time, the spectrum flatness parameter is calculated based on the frequency domain enhanced data to evaluate the uniformity of the spectrum. Spectral flatness is an indicator that measures the uniformity of the signal distribution in the frequency domain. Its calculation formula is: ; in, represents the spectrum flatness, is the number of spectral components, Indicates The amplitude of each spectral component. If the flatness deviation range is within ±3dB, the spectrum distribution is considered uniform and the corresponding frequency domain weight coefficient is high. The frequency domain weight coefficient is calculated as follows: ; in, is the frequency domain weight coefficient, is the reference spectrum flatness value, is the flatness deviation range, set to 3dB. The time domain weight coefficient and the frequency domain weight coefficient are weighted and combined, and the weight normalization operation is performed to obtain the feature weight matrix. The weighted combination is expressed as: ; in, is the feature weighting matrix, To prevent the denominator from being zero, the time domain denoising data and the frequency domain enhanced data are fused according to the feature weighting matrix. The fusion process is described by the following formula: ; in, is the time domain denoised data, Enhanced data for the frequency domain. Through feature fusion, the features of the time domain and frequency domain are combined to obtain a more comprehensive and accurate feature description. Feature screening operations are performed on the fused feature data to obtain a purified feature data set. The screening criteria are based on frequency response flatness, phase linearity, and group delay fluctuation indicators. Frequency response flatness is used to measure whether the signal response in different frequency bands is balanced, phase linearity reflects whether the phase response changes linearly, and group delay fluctuation describes whether the delay of different frequency components through the system is consistent. The calculation formula for group delay is: ; in, is the group delay, For frequency When performing feature screening on the fused feature set, if some features cannot meet the above indicators, these features are removed or adjusted to obtain the final purified feature data set.

[0032] In a specific embodiment, the process of executing step 400 may specifically include the following steps: The purified feature data set is divided into frequency bands, the 20Hz-20kHz range is divided into 128 frequency bands, and a bandpass filter operation is performed on each frequency band to obtain the frequency band division data; The frequency band division data is encoded and converted according to the feature weighting matrix to construct a 16-dimensional frequency encoding vector, in which the amplitude occupies 8 dimensions, the phase occupies 4 dimensions, and the time occupies 4 dimensions, thereby obtaining the frequency encoding data; Perform recursive sampling processing on the frequency-encoded data, set the basic sampling density to 4 points / octave, and set the sampling density to 16 points / octave in the frequency response mutation area to obtain multi-density sampling data; The multi-density sampling data is input into a three-level feature extraction unit, the first layer performs time domain convolution operation to extract transient features, the second layer performs frequency domain transformation to extract harmonic features, and the third layer performs correlation analysis to extract associated features to obtain three-dimensional feature data; A two-layer prediction unit is constructed for the three-dimensional feature data. The first layer processes the time series relationship through the LSTM network to obtain the time series prediction value, and the second layer obtains the frequency response prediction value by fusing the frequency domain information through the feedforward network to obtain the predicted feature data. A three-level residual network is constructed based on the predicted feature data. The first level calculates the amplitude residual, the second level calculates the phase residual, and the third level calculates the delay residual to obtain the residual feature data. The residual feature data is input into the feature fusion network, and the feature dimension reduction and optimization are performed through the fully connected layer. The number of output layer nodes is set to 32 to obtain the optimized feature data. The optimized feature data is normalized and the three-dimensional features are compressed into a one-dimensional vector through weighted sum operation to obtain the frequency response feature vector.

[0033] Specifically, the purified feature data set is divided into frequency bands, and the frequency range from 20Hz to 20kHz is divided into 128 frequency bands. Each frequency band after division is expressed by the following formula: ; in, For the The center frequency of the frequency band, is the minimum frequency (20Hz), is the width of the frequency band, and its value is obtained by dividing the total frequency range (20kHz-20Hz) by 128. Bandpass filtering is performed on each frequency band to retain the signal components of the frequency band and suppress signal interference from other frequency bands, which is achieved through the transfer function of the bandpass filter: ; in, For the The transfer function of a bandpass filter is, is the center frequency of the frequency band, is the bandwidth of the frequency band. After bandpass filtering, the data after frequency band division is obtained, and the signal components in each frequency band can be analyzed independently. The frequency band division data is encoded and converted according to the feature weighting matrix to construct a 16-dimensional frequency encoding vector. In this encoding vector, the amplitude feature occupies 8 dimensions and is used to describe the energy distribution of the signal in each frequency band. The amplitude feature is expressed as: ; in, For the The amplitude characteristics of the dimension, The complex spectrum value of the frequency band division data. The phase feature occupies 4 dimensions and is used to describe the phase change of the signal in different frequency bands. It is expressed by the following formula: ; in, is the phase characteristic, Indicates The phase value of each frequency band. The time feature occupies 4 dimensions and is used to describe the transient changes of the signal on the time axis. After encoding, a 16-dimensional frequency encoding vector is obtained. Recursive sampling is performed on the frequency-encoded data, the basic sampling density is set to 4 points / octave, and the sampling density is increased to 16 points / octave in the frequency response mutation area. Higher sampling accuracy is obtained in areas where the frequency response changes dramatically in order to capture more subtle feature changes. The formula for recursive sampling is expressed as:

[0034] in, Indicates The sampling density of the frequency band is is the threshold for frequency response mutation. When the frequency band change amplitude is greater than the threshold, the sampling density is increased to 16 points / octave to ensure that enough information is obtained in these areas. Through the sampling strategy, multi-density sampling data is obtained to provide rich features in different frequency ranges. The multi-density sampling data is input into the three-level feature extraction unit to extract features at different levels. In the first layer, a time domain convolution operation is performed to extract transient features, describing the mutation and short-term changes of the signal in time. The calculation of time domain convolution is expressed as: ; in, is the output after convolution, is the input signal, is the convolution kernel, is the length of the convolution kernel. Transient features help analyze noise bursts and instantaneous changes in the signal. In the second layer, frequency domain transformation is performed to extract the harmonic features of the signal. Harmonic features are implemented through fast Fourier transform to identify harmonic components in the signal, which reflect the frequency structure of the signal. In the third layer, correlation analysis is performed to extract the correlation features between different frequency bands. The results of the correlation analysis are represented by the correlation coefficient matrix: ; in, For the and The correlation coefficient between the frequency bands, and Respectively and The correlation analysis helps to identify the dependency between different frequency components in the signal and obtain three-dimensional feature data. A two-layer prediction unit is constructed for the three-dimensional feature data. In the first layer, the time series relationship is processed by the LSTM network to obtain the time series prediction value. LSTM (Long Short-Term Memory Network) is a recurrent neural network suitable for processing time series data, which can effectively capture the long-term dependency of the signal. The state update formula of the LSTM network is: ; in, is the hidden state at the current moment, is the output gate, is the cell state. In this way, LSTM can process time series data and extract the time series characteristics of the signal. In the second layer, the frequency domain information is fused through the feedforward network to obtain the frequency response prediction value. The calculation of the feedforward network is expressed as: ; in, is the output of the network, is the weight matrix, For input data, is the bias, is the activation function. Through these two layers of processing, the predicted feature data is obtained. A three-level residual network is constructed based on the predicted feature data to calculate the amplitude residual to eliminate the error in the amplitude prediction. The amplitude residual is expressed as: ; in, is the amplitude residual, is the true amplitude value, is the predicted amplitude value. Calculate the phase residual: ; in, is the phase residual, is the true phase value, is the predicted phase value. Calculate the delay residual: ; in, is the delay residual, and are the real and predicted group delays, respectively. Through the three-level residual network, the error in the prediction is effectively corrected to obtain the residual feature data. The residual feature data is input into the feature fusion network, and the feature dimension reduction and optimization are performed through the fully connected layer. The fully connected layer is used to reduce the high-dimensional feature data to a lower dimension for further analysis. The number of nodes in the output layer is set to 32, and the most important information is extracted through dimension reduction while reducing the complexity of the data. The optimized feature data after dimensionality reduction is normalized to eliminate the scale differences between different features. The three-dimensional features are compressed into a one-dimensional vector through a weighted sum operation to obtain the frequency response feature vector. The formula is: ; in, is the final frequency response eigenvector, is the weighting coefficient, is the i-th dimension in the three-dimensional feature data. Through weighted summation, different features are effectively integrated to obtain a frequency response feature vector that describes the frequency response characteristics of the entire system.

[0035] In a specific embodiment, the process of executing step 500 may specifically include the following steps: The frequency response characteristic vector is divided into three sections, 20Hz-200Hz is divided into a low frequency section, 200Hz-2kHz is divided into a medium frequency section, and 2kHz-20kHz is divided into a high frequency section, to obtain segmented frequency response data; An adaptive interval model is constructed based on the segmented frequency response data, the confidence level is set to 95%, the maximum deviation calculation is performed, the interval boundary data is obtained, and mixed probability modeling is performed on the interval boundary data, the Gaussian distribution weight is set to 0.7, the uniform distribution weight is set to 0.3, and the probability distribution data is obtained; An environmental compensation matrix is ​​constructed based on the temperature coefficient ±0.01dB / ℃ and the humidity coefficient ±0.02dB / 10%RH, and compensation operations are performed on the probability distribution data to obtain environmental compensation data; Compare the environmental compensation data with the 1kHz / -20dBFS reference signal, calculate the system error parameters, obtain the calibration reference data, and build a 12th-order FIR filter based on the calibration reference data to perform nonlinear error compensation operations to obtain compensation calibration data; The segmented frequency response data is calibrated and corrected based on the compensation calibration data, and the correction coefficients of the low frequency band, the middle frequency band and the high frequency band are calculated respectively to obtain the frequency band calibration data. The frequency band calibration data is then integrated to obtain a standard frequency response data set.

[0036] Specifically, the frequency range of 20Hz to 20kHz is divided into three different frequency bands, of which 20Hz to 200Hz is defined as the low frequency band, 200Hz to 2kHz is defined as the mid-frequency band, and 2kHz to 20kHz is defined as the high frequency band. Through division, the frequency response characteristics of the low-frequency, mid-frequency and high-frequency bands are analyzed respectively to obtain the segmented frequency response data. An adaptive interval model is constructed based on the segmented frequency response data, the confidence level is set to 95%, and the maximum deviation calculation is performed to determine the response range within each frequency band. The construction of the adaptive interval model is based on the statistical characteristics of the signal. The mean and standard deviation of each frequency band are calculated for constructing a confidence interval. Taking the low frequency band as an example, let its mean be and the standard deviation is , then at the 95% confidence level, its confidence interval is expressed as: ; in, represents the confidence interval of the low frequency band, and 1.96 is the coefficient corresponding to the 95% confidence level. In the same way, the confidence intervals of the mid-frequency band and the high-frequency band are calculated to obtain the interval boundary data of all frequency bands. In order to describe the characteristics of these interval boundaries, mixed probability modeling is performed to combine Gaussian distribution and uniform distribution. The Gaussian distribution weight is set to 0.7 and the uniform distribution weight is set to 0.3 to reflect the concentration and randomness in the signal characteristics. The mixed probability density function is expressed as: ; in, It means the mean , standard deviation is Gaussian distribution, Representation interval The uniform distribution on the surface is obtained through the hybrid model, and more accurate probability distribution data is obtained. Environmental compensation operation is performed on the probability distribution data according to the temperature coefficient and humidity coefficient. Environmental conditions, such as temperature and humidity, will affect the characteristics of the frequency response. An environmental compensation matrix is ​​constructed based on the temperature coefficient ±0.01dB / ℃ and the humidity coefficient ±0.02dB / 10%RH to correct the signal. The compensation operation is expressed as: ; in, is the frequency response data after compensation, and are the changes in temperature and humidity, respectively. and are the compensation coefficients for temperature and humidity, respectively. By performing compensation calculations on the signal, the environmental compensation data obtained better reflects the system performance in the real environment. The environmental compensation data is compared with the 1kHz / -20dBFS reference signal to calculate the error parameters of the system and obtain the calibration reference data. The 1kHz / -20dBFS reference signal is a commonly used reference signal in audio testing. By comparing it with the compensation data, the deviation of the system under different conditions is calculated. The system error is expressed as: ; in, is the error value, is the frequency response characteristic of the reference signal. Based on the error parameters, a 12th-order finite impulse response filter (FIR) is constructed to perform nonlinear error compensation. The design goal of the FIR filter is to make the frequency response characteristics of the system as close to the ideal response as possible. The impulse response of the filter is expressed as: ; in, is the impulse response of the filter, is the filter coefficient, is a unit impulse function. The error signal is compensated by the filter to obtain compensation calibration data, so that the response of the system is closer to the target characteristics. Based on the compensation calibration data, the segmented frequency response data is calibrated and corrected, and the correction coefficients of the low frequency band, mid-frequency band and high frequency band are calculated respectively. The correction coefficients of the low frequency band, mid-frequency band and high frequency band are expressed by the following formula: ; in, For frequency band The correction factor of For frequency band The error data, For frequency band The original response data of the frequency band is obtained by calculating the correction coefficient, and the frequency response characteristics within the frequency band are effectively adjusted so that the response of each frequency band meets the expected standard. Data integration processing is performed on the frequency band calibration data to obtain the standard frequency response data set.

[0037] In a specific embodiment, the process of executing step 600 may specifically include the following steps: The standard frequency response dataset is input into the feature extraction network. The feature extraction network contains 8 layers of convolutional neural network. Each layer uses ReLU activation function and batch normalization layer to obtain multi-scale feature maps. Perform feature pyramid operation on the multi-scale feature map to construct four feature layers with different resolutions. The downsampling ratios of each feature layer are 1, 1 / 2, 1 / 4, and 1 / 8, respectively, to obtain multi-level feature data. The multi-level feature data is input into the two-stream attention network, which includes a frequency channel attention module and a temporal spatial attention module. The frequency channel attention module uses a 1×1 convolution layer to extract channel weights, and the temporal spatial attention module uses a 5×5 convolution layer to extract position weights to obtain attention feature data. Perform feature aggregation on the attention feature data, and perform weighted fusion on feature maps of different resolutions through the cross-layer feature fusion module. The weight coefficient is learned through the 1×1 convolution layer to obtain the fused feature data; The fused feature data is input into the classification branch network. The classification branch network contains three fully connected layers with 512, 256 and 128 nodes respectively. Each layer is followed by a Dropout layer with a dropout rate set to 0.5 to obtain the frequency classification result. Perform regression branch processing on the fused feature data, extract deep features through 3-layer residual blocks, each residual block contains two 3×3 convolutional layers and a short-circuit connection, and obtain regression feature data; The regression feature data is input into the dual-head output layer, the first output head predicts the frequency response value through two fully connected layers, and the second output head predicts the phase response value through two fully connected layers to obtain the response prediction data; The frequency classification results and response prediction data are processed into prediction points through spline interpolation algorithm to obtain frequency response curve and phase characteristic curve.

[0038] Specifically, the standard frequency response dataset is input into the feature extraction network, and a feature extraction module containing 8 layers of convolutional neural network is designed. These convolutional layers are used to gradually extract high-level features of the signal and help the network learn more complex representations layer by layer. Each convolutional layer applies the ReLU activation function to introduce nonlinearity and enable the network to process complex patterns. The formula of the ReLU activation function is: ; in, is the output after activation, is the input of the convolution layer. Through this operation, the gradient vanishing problem is effectively prevented and the network's expressive power is improved. Each convolution layer includes a batch normalization layer, which is used to normalize the data of each small batch to make the network training more stable. After 8 layers of convolution and batch normalization, a multi-scale feature map is obtained, which contains frequency response features at different levels. Feature pyramid operations are performed on the multi-scale feature map. The feature pyramid is used to construct feature layers of different resolutions to capture the characteristics of signals at different scales. Four feature layers of different resolutions are constructed, and the downsampling ratios of each feature layer are 1, 1 / 2, 1 / 4 and 1 / 8 respectively. Downsampling is achieved through pooling operations or strided convolutions, which gradually reduce the resolution while retaining global information to facilitate the fusion and processing of features of different scales. Suppose the input feature map is , No. The feature layer of the layer feature pyramid is , then the downsampling ratio is: ; in, represents the pooling operation, represents the downsampling ratio, which is 1, 1 / 2, 1 / 4 and 1 / 8 respectively. After the feature pyramid operation, multi-level feature data is obtained, which reflects the frequency response characteristics at different scales. The multi-level feature data is input into the two-stream attention network, which consists of a frequency channel attention module and a temporal space attention module. Among them, the frequency channel attention module uses a 1×1 convolution layer to extract the channel weight. The role of the 1×1 convolution is to emphasize the important frequency band features by adjusting the weights of each channel. Suppose the input feature is , the output of channel attention is ,but: ; in, express Convolution operation, is the Sigmoid activation function, which is used to map the output to the range of (0, 1) to adjust the weight of each channel. The temporal spatial attention module uses a 5×5 convolutional layer to extract the weight of the position and capture the important characteristics of the signal in time and space. Let its output be ,but: ; The output of the dual-stream attention network is attention feature data. This mechanism enhances attention to specific frequency bands and time points, making feature extraction more effective. Feature aggregation is performed on the attention feature data, and feature maps of different resolutions are weighted and fused through the cross-layer feature fusion module. During the fusion process, the weight coefficient is calculated by The convolutional layer is learned to ensure the best fusion effect between features of different scales. Let the feature map be , then the fusion feature data It is expressed as: ; in, is the weight coefficient learned through the 1×1 convolutional layer, For the Layer feature map. Through weighted fusion, multi-level feature information is combined to obtain more representative fused feature data. The fused feature data is input into the classification branch network. The classification branch network contains 3 fully connected layers with 512, 256 and 128 nodes respectively. Each fully connected layer is followed by a Dropout layer with a dropout rate set to 0.5. The purpose of Dropout is to randomly drop half of the nodes during training to prevent overfitting of the network. The formula is: ; in, is the output after Dropout, is the output of the fully connected layer, is the discard rate, which is set to 0.5. After being processed by the classification branch network, the frequency classification result is obtained, which is used to identify which frequency response mode the input signal belongs to. At the same time, the regression branch processing is performed on the fused feature data to extract deep features through a three-layer residual block. Each residual block contains two The introduction of residual blocks helps solve the gradient vanishing problem in deep network training and enables information to be transmitted more smoothly in deep networks. The output of the residual block is expressed as: ; in, represents the transformation through two 3×3 convolutional layers, is the input signal, is the output of the residual block. Through the three layers of residual blocks, the deep features of the signal are extracted to obtain the regression feature data. The regression feature data is input into the double-head output layer. The first output head predicts the frequency response value through two fully connected layers, and the second output head predicts the phase response value through two fully connected layers. Assume that the input feature is , then the frequency response value and phase response value are expressed as: ; ; in, and is the weight matrix, and is the bias, and are activation functions for predicting frequency response values ​​and phase response values, respectively. Through the double-headed output layer, the predicted data for frequency response and phase response are obtained. The frequency classification results and response prediction data are processed continuously through the spline interpolation algorithm to obtain the frequency response curve and phase characteristic curve. Spline interpolation is a smoothing method used to connect discrete prediction points into a continuous curve. Let the discrete points be , then the goal of spline interpolation is to construct a smooth function ,satisfy: ; Through spline interpolation, a smooth curve is generated to describe the frequency response and phase characteristics, so that the final output is more consistent with the actual frequency response characteristics.

[0039] The above describes the frequency response analysis method of the intelligent audio test system in the embodiment of the present application. The following describes the frequency response analysis device 10 of the intelligent audio test system in the embodiment of the present application. Figure 2 In the embodiment of the present application, an embodiment of the intelligent audio test system frequency response analysis device 10 includes: Initialization module 11, used to initialize audio parameters of the intelligent audio test system to obtain a test parameter matrix and an excitation signal; An acquisition module 12 is used to perform dual-channel acquisition and time-frequency feature extraction on the excitation signal and the response signal based on the test parameter matrix to obtain a time-frequency feature data set; A processing module 13 is used to perform a dual-branch process of time domain noise reduction and frequency domain enhancement on the time-frequency feature data set to obtain a purified feature data set and a feature weighting matrix; The calculation module 14 is used to perform segmented recursive calculation and feature aggregation on the purified feature data set according to the feature weighting matrix to obtain a frequency response feature vector; The calibration module 15 is used to decompose the frequency response characteristic vector according to the test frequency band, and perform environmental compensation and nonlinear calibration to obtain a standard frequency response data set; The classification module 16 is used to input the standard frequency response data set into the deep neural network for feature recognition and pattern classification to obtain a frequency response curve and a phase characteristic curve.

[0040] Through the synergy of the above components and the introduction of a dual-branch data processing architecture, the synergistic enhancement of time domain and frequency domain signals is achieved, effectively improving the system's anti-noise ability; the multi-level feature extraction and recursive analysis strategies are adopted to significantly improve the extraction accuracy of frequency response features; the environmental compensation and nonlinear calibration mechanisms are innovatively designed to effectively eliminate the interference of environmental factors; deep neural networks are introduced for feature recognition and pattern classification to achieve intelligent and automated testing processes. This method ensures the accuracy of full-band testing through segmented recursive calculation and feature aggregation technology; adopts a dual-stream attention network and residual learning mechanism to enhance the system's ability to capture complex frequency response features; through multi-level calibration and compensation mechanisms, the reliability of test results is ensured; and finally, through intelligent analysis of deep learning models, high-precision frequency response curve reconstruction is achieved. The present invention has achieved significant improvements in test accuracy, anti-interference ability, degree of automation, and adaptability.

[0041] See also Figure 3 , Figure 3 This is a schematic block diagram of the structure of an electronic device 300 provided in an embodiment of the present application. The electronic device 300 includes a processor 301 and a memory 302. The processor 301 and the memory 302 are connected via a device bus 303, wherein the memory 302 may include a non-volatile storage medium and an internal memory.

[0042] The non-volatile storage medium can store a computer program. The computer program includes program instructions, and when the program instructions are executed by the processor 301, the processor 301 can execute any of the above-mentioned frequency response analysis methods of the intelligent audio test system.

[0043] The processor 301 is used to provide computing and control capabilities to support the operation of the entire electronic device 300 .

[0044] The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium. When the computer program is executed by the processor 301, the processor 301 can execute any of the above-mentioned frequency response analysis methods of the intelligent audio test system.

[0045] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a partial structure related to the present application scheme, and does not constitute a limitation on the electronic device 300 involved in the present application scheme. The specific electronic device 300 may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0046] It should be understood that the processor 301 may be a central processing unit (CPU), and the processor 301 may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0047] It should be noted that those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the electronic device 300 described above can refer to the corresponding process of the aforementioned intelligent audio test system frequency response analysis method, which will not be repeated here.

[0048] An embodiment of the present application further provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by one or more processors, the one or more processors implement the frequency response analysis method of the intelligent audio test system provided in the embodiment of the present application.

[0049] The computer-readable storage medium may be an internal storage unit of the electronic device 300 in the aforementioned embodiment, such as a hard disk or memory of the electronic device 300. The computer-readable storage medium may also be an external storage device of the electronic device 300, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped with the electronic device 300.

[0050] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0051] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable an electronic device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc., various media that can store program codes.

[0052] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A frequency response analysis method for an intelligent audio test system, characterized in that: The method comprises: Initialize audio parameters of the intelligent audio test system to obtain a test parameter matrix and an excitation signal; Based on the test parameter matrix, dual-channel acquisition and time-frequency feature extraction are performed on the excitation signal and the response signal to obtain a time-frequency feature data set; Performing dual-branch processing of time domain noise reduction and frequency domain enhancement on the time-frequency feature data set to obtain a purified feature data set and a feature weighting matrix; Performing segmented recursive calculation and feature aggregation on the purified feature data set according to the feature weighting matrix to obtain a frequency response feature vector; Decomposing the frequency response characteristic vector according to the test frequency band, and performing environmental compensation and nonlinear calibration to obtain a standard frequency response data set; The standard frequency response data set is input into a deep neural network for feature recognition and pattern classification to obtain a frequency response curve and a phase characteristic curve.

2. The frequency response analysis method of the intelligent audio test system according to claim 1, characterized in that: The method of initializing audio parameters of the intelligent audio test system to obtain a test parameter matrix and an excitation signal includes: Perform DSP algorithm configuration on the processor of the intelligent audio test system, load the all-pass filter and dynamic equalizer DEQ into the intelligent audio test system, and obtain DSP initialization parameters; The system sampling parameters are set based on the DSP initialization parameters to obtain basic sampling parameters, and the signal ranges of 14 digital gain channels are set according to the basic sampling parameters to obtain a channel parameter set; Performing gain step setting on the channel parameter set to obtain a gain control matrix, verifying the system dynamic range according to the gain control matrix, performing background noise index measurement, and obtaining dynamic range verification data; According to the dynamic range verification data, the signal-to-noise ratio index is measured and calculated to obtain system performance parameters, and the system performance parameters and the gain control matrix are combined to obtain a test parameter matrix. Based on the test parameter matrix, a test signal is generated to obtain an excitation signal.

3. The frequency response analysis method of the intelligent audio test system according to claim 2, characterized in that: The step of performing dual-channel acquisition and time-frequency feature extraction on the excitation signal and the response signal based on the test parameter matrix to obtain a time-frequency feature data set includes: Performing 24-bit analog-to-digital conversion on the excitation signal and the response signal, setting the number of sampling points to 2048 points / frame to obtain original sampling data, and performing 14-channel signal allocation on the original sampling data based on the test parameter matrix to obtain channel sampling data; Performing 1 dB step gain adjustment on the channel sampling data, performing signal amplitude control through a processor, and obtaining gain control data; Input the gain control data into an all-pass filter for signal preprocessing, execute dynamic equalization DEQ algorithm operation to obtain filtered processing data, and perform real-time Fourier transform on the filtered processing data to obtain frequency domain feature data; Performing phase characteristic extraction on the frequency domain feature data, calculating group delay and phase response to obtain phase feature data, and performing information fusion operation on the frequency domain feature data and the phase feature data, performing feature alignment processing to obtain fused feature data; The fused feature data is standardized to obtain a time-frequency feature data set.

4. The frequency response analysis method of the intelligent audio test system according to claim 3, characterized in that: The step of performing dual-branch processing of time domain noise reduction and frequency domain enhancement on the time-frequency feature data set to obtain a purified feature data set and a feature weighting matrix includes: Performing branch division on the time-frequency feature data set, dividing the signal into a time domain processing branch and a frequency domain processing branch, to obtain a dual-branch data set; Performing a 4-level wavelet decomposition operation on the time domain branch in the dual-branch data set, setting the threshold to 3 times the noise standard deviation, to obtain time domain denoised data; Performing a spectral subtraction operation on the frequency domain branch in the dual-branch data set, setting the frequency-related suppression factor to 0.8, and obtaining frequency domain enhanced data; Calculate the signal-to-noise ratio index according to the time domain noise reduction data, set the signal-to-noise ratio threshold to 40 dB, obtain the time domain weight coefficient, and calculate the spectrum flatness parameter based on the frequency domain enhancement data, set the flatness deviation range to ±3 dB, and obtain the frequency domain weight coefficient; Performing a weighted combination on the time domain weight coefficient and the frequency domain weight coefficient, performing a weight normalization operation to obtain a feature weight matrix, and performing feature fusion on the time domain denoised data and the frequency domain enhanced data according to the feature weight matrix to obtain a fused feature set; Based on frequency response flatness, phase linearity and group delay fluctuation indicators, a feature screening operation is performed on the fused feature set to obtain a purified feature data set.

5. The frequency response analysis method of the intelligent audio test system according to claim 4, characterized in that: The step of performing segmented recursive calculation and feature aggregation on the purified feature data set according to the feature weighting matrix to obtain a frequency response feature vector includes: Performing frequency band division on the purified feature data set, dividing the range of 20 Hz-20 kHz into 128 frequency bands, performing a bandpass filtering operation on each frequency band, and obtaining frequency band division data; The frequency band division data is encoded and converted according to the feature weighting matrix to construct a 16-dimensional frequency encoding vector, in which the amplitude occupies 8 dimensions, the phase occupies 4 dimensions, and the time occupies 4 dimensions, to obtain frequency encoding data; Performing recursive sampling processing on the frequency-encoded data, setting the basic sampling density to 4 points / octave, and setting the sampling density to 16 points / octave in the frequency response mutation region, to obtain multi-density sampling data; The multi-density sampling data is input into a three-level feature extraction unit, the first layer performs a time domain convolution operation to extract transient features, the second layer performs a frequency domain transformation to extract harmonic features, and the third layer performs a correlation analysis to extract associated features, thereby obtaining three-dimensional feature data; A two-layer prediction unit is constructed for the three-dimensional feature data, wherein the first layer processes the time series relationship through an LSTM network to obtain a time series prediction value, and the second layer obtains a frequency response prediction value by fusing frequency domain information through a feedforward network to obtain prediction feature data; Based on the predicted feature data, a three-level residual network is constructed, wherein the first level calculates the amplitude residual, the second level calculates the phase residual, and the third level calculates the delay residual to obtain residual feature data; The residual feature data is input into the feature fusion network, and feature dimension reduction and optimization are performed through the fully connected layer. The number of output layer nodes is set to 32 to obtain optimized feature data, and normalization is performed on the optimized feature data. The three-dimensional features are compressed into a one-dimensional vector through a weighted sum operation to obtain a frequency response feature vector.

6. The frequency response analysis method of the intelligent audio test system according to claim 5, characterized in that: Decomposing the frequency response characteristic vector according to the test frequency band, and performing environmental compensation and nonlinear calibration to obtain a standard frequency response data set includes: Performing three-segment frequency division on the frequency response characteristic vector, dividing 20 Hz-200 Hz into a low frequency segment, 200 Hz-2 kHz into a medium frequency segment, and 2 kHz-20 kHz into a high frequency segment, to obtain segmented frequency response data; An adaptive interval model is constructed based on the segmented frequency response data, the confidence level is set to 95%, a maximum deviation calculation is performed to obtain interval boundary data, and mixed probability modeling is performed on the interval boundary data, the Gaussian distribution weight is set to 0.7, and the uniform distribution weight is set to 0.3 to obtain probability distribution data; An environmental compensation matrix is ​​constructed according to a temperature coefficient of ±0.01 dB / °C and a humidity coefficient of ±0.02 dB / 10% RH, and a compensation operation is performed on the probability distribution data to obtain environmental compensation data; Compare the environmental compensation data with a 1kHz / -20dBFS reference signal, calculate system error parameters, obtain calibration reference data, and construct a 12th-order FIR filter based on the calibration reference data, perform nonlinear error compensation operations, and obtain compensation calibration data; The segmented frequency response data is calibrated and corrected based on the compensation calibration data, correction coefficients of the low frequency band, the middle frequency band and the high frequency band are calculated respectively to obtain frequency band calibration data, and data integration processing is performed on the frequency band calibration data to obtain a standard frequency response data set.

7. The frequency response analysis method of the intelligent audio test system according to claim 6, characterized in that: The step of inputting the standard frequency response data set into a deep neural network for feature recognition and pattern classification to obtain a frequency response curve and a phase characteristic curve includes: Inputting the standard frequency response data set into a feature extraction network, wherein the feature extraction network comprises an 8-layer convolutional neural network, each layer uses a ReLU activation function and a batch normalization layer to obtain a multi-scale feature map; Performing feature pyramid operation on the multi-scale feature map to construct four feature layers with different resolutions, with downsampling ratios of each feature layer being 1, 1 / 2, 1 / 4 and 1 / 8 respectively, to obtain multi-level feature data; Input the multi-level feature data into a dual-stream attention network, wherein the dual-stream attention network includes a frequency channel attention module and a temporal space attention module, wherein the frequency channel attention module uses a 1×1 convolution layer to extract channel weights, and the temporal space attention module uses a 5×5 convolution layer to extract position weights, to obtain attention feature data; Performing feature aggregation on the attention feature data, weighted fusion of feature maps of different resolutions through a cross-layer feature fusion module, wherein the weight coefficient is learned through a 1×1 convolutional layer to obtain fused feature data; The fused feature data is input into a classification branch network, wherein the classification branch network comprises three fully connected layers, the number of nodes of which are 512, 256 and 128 respectively, each layer is followed by a Dropout layer, and the dropout rate is set to 0.5, to obtain a frequency classification result; Performing regression branch processing on the fused feature data, extracting deep features through three layers of residual blocks, each residual block including two 3×3 convolutional layers and a short-circuit connection, to obtain regression feature data; The regression feature data is input into a dual-head output layer, the first output head predicts a frequency response value through two fully connected layers, and the second output head predicts a phase response value through two fully connected layers to obtain response prediction data; The frequency classification result and the response prediction data are processed into prediction point continuity by using a spline interpolation algorithm to obtain a frequency response curve and a phase characteristic curve.

8. An intelligent audio test system frequency response analysis device, characterized in that: Used to execute the intelligent audio test system frequency response analysis method according to any one of claims 1 to 7, the intelligent audio test system frequency response analysis device comprises: An initialization module is used to initialize the audio parameters of the intelligent audio test system to obtain a test parameter matrix and an excitation signal; An acquisition module, used for performing dual-channel acquisition and time-frequency feature extraction on the excitation signal and the response signal based on the test parameter matrix to obtain a time-frequency feature data set; A processing module, used for performing a dual-branch process of time domain noise reduction and frequency domain enhancement on the time-frequency feature data set to obtain a purified feature data set and a feature weighting matrix; A calculation module, used for performing segmented recursive calculation and feature aggregation on the purified feature data set according to the feature weighting matrix to obtain a frequency response feature vector; A calibration module, used for decomposing the frequency response characteristic vector according to the test frequency band, and performing environmental compensation and nonlinear calibration to obtain a standard frequency response data set; The classification module is used to input the standard frequency response data set into a deep neural network for feature recognition and pattern classification to obtain a frequency response curve and a phase characteristic curve.

9. An electronic device, characterized in that: The electronic device comprises: a memory and at least one processor, wherein instructions are stored in the memory; The at least one processor calls the instruction in the memory to enable the electronic device to execute the frequency response analysis method of the intelligent audio test system according to any one of claims 1 to 7.

10. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by the processor, the frequency response analysis method of the intelligent audio test system according to any one of claims 1 to 7 is implemented.

Images