How to convert USB Audio stereo to 5.1 channel audio

Abstract

The application provides a kind of implementation method of USB Audio stereo to 5.1 channel audio, to solve the problems of traditional conversion mode sound field distortion, low channel separation, feedback regulation lag, etc.The method is applied to USB Audio audio processing system, by collecting USB stereo signal and preprocessing, initial 5.1 channel signal is obtained through initial matrix mapping;Real-time acquisition of output end and sound field two-dimensional feedback signal, extract multiple feature parameters, through multi-level complex matrix feedback mechanism, calculate feedback weight matrix and dynamic adjustment matrix, realize dynamic adaptation of mapping matrix after stability check;At the same time, introduce USB transmission compensation and exception handling mechanism, adapt to multiple scenes.The application does not need additional hardware, realizes high-fidelity conversion, has strong anti-interference ability, and is suitable for multiple scenes such as home and vehicle.

Description

Technical Field

[0001] This invention relates to the field of audio signal processing technology, specifically to a method for converting USB Audio stereo to 5.1 channel audio. It is particularly suitable for scenarios that rely on USB interfaces to transmit audio and require immersive multi-channel playback, such as home theaters and in-vehicle entertainment. The core of this invention lies in achieving efficient and high-fidelity conversion of stereo signals to 5.1 channel signals through a complex matrix feedback mechanism, solving the technical pain points of traditional conversion methods such as sound field distortion, low channel separation, and lag in feedback adjustment. Background Technology

[0002] With the popularization of multimedia technology and immersive audio-visual experiences, 5.1 channel audio, due to its advantages of strong stereo surround sound and rich sound field layering, has been widely used in various audio playback devices. Currently, a large number of audio sources on the market (such as music, video, game sound effects, etc.) still exist in stereo (left channel L, right channel R) format. At the same time, the USB interface has become the mainstream transmission interface between audio devices due to its stable transmission and strong universality, and the USB Audio protocol has become the standard protocol for transmitting audio signals through the USB interface.

[0003] In existing technologies, the implementation methods of USB Audio stereo to 5.1 channel conversion are mainly divided into two categories. The first category is a simple signal distribution conversion, which directly distributes or simply superimposes the stereo left and right channel signals onto the 5.1 channel front left (FL), front right (FR), center (C), left surround (SL), right surround (SR), and low frequency effect (LFE) channels. This method is simple in structure and low in cost, but it has serious sound field distortion problems. The signals of each channel are highly correlated and have low separation, which makes it impossible to present a real surround sound field. In addition, it does not consider the attenuation and interference during signal transmission, resulting in poor playback effect. The second type is the fixed matrix-based conversion method, which maps stereo signals to 5.1 channels using preset matrix coefficients. However, the matrix coefficients in this method are fixed values ​​and cannot be dynamically adjusted according to the frequency characteristics, amplitude changes, and characteristics of the audio signal and the playback device. Furthermore, it lacks an effective feedback mechanism. When there are fluctuations in the input stereo signal, delays or interference in USB transmission, or differences in the channel response of the playback device, the converted 5.1 channel signal will suffer from amplitude imbalance, phase deviation, howling, and other problems, which seriously affect the fidelity and immersion of audio playback.

[0004] Furthermore, even the few existing conversion methods that incorporate feedback adjustment are mostly simple, single-dimensional feedback mechanisms (such as adjusting only the amplitude of the center channel). They cannot achieve multi-channel, multi-parameter, end-to-end dynamic feedback calibration, resulting in slow response times, low adjustment precision, and difficulty in adapting to complex audio scenarios and diverse playback devices. Simultaneously, issues such as audio data frame synchronization deviations and signal attenuation during USB Audio protocol transmission further exacerbate channel distortion after conversion. Existing conversion methods lack corresponding feedback compensation mechanisms designed for USB transmission characteristics, leading to conversion results being significantly affected by USB transmission quality. Summary of the Invention

[0005] To address the aforementioned problems, this invention proposes a method for converting USB Audio stereo to 5.1 channel audio. This method combines the transmission characteristics of USB Audio and uses a complex matrix feedback mechanism to achieve high-fidelity conversion from stereo to 5.1 channel audio. It solves the technical defects of existing technologies, such as sound field distortion, low channel separation, lag in feedback adjustment, and weak anti-interference ability, thereby improving the immersive audio playback experience.

[0006] Specifically, this invention provides a method for converting USB Audio stereo to 5.1 channel audio. This method is based on a system architecture that includes a USB audio input module, a stereo preprocessing module, a matrix conversion module, a 5.1 channel output module, a feedback acquisition module, and a feedback control module. The method includes the following steps: Step S1: The USB audio input module receives the USB Audio stereo signal transmitted from an external device via the USB interface, and converts it according to the USB... The audio protocol parses the stereo signal, extracts the raw audio data, and transmits this raw audio data to the stereo preprocessing module for preprocessing to obtain a standardized stereo signal. Step S2: The standardized stereo signal is transmitted to the matrix conversion module, which has a built-in initial mapping matrix M0. The initial mapping matrix M0 maps the standardized stereo signal to an initial 5.1 channel signal. Step S3: The initial 5.1 channel signal is transmitted to the 5.1 channel output module. The 5.1 channel output module amplifies and compensates the initial 5.1 channel signal before outputting it to an external playback device. The feedback acquisition module acquires feedback signals in real time in two dimensions: the first dimension is the output signal F1 from the 5.1 channel output module, and the second dimension is the sound field feedback signal F2 from the external playback device. Step S4: The feedback control module receives the feedback signals F1 and F2 and... The two feedback signals are analyzed and processed to extract feedback feature parameters. Based on the feedback feature parameters, the feedback weight matrix W is calculated, and the feedback attenuation coefficient λ is introduced to calculate the attenuated feedback weight matrix W'. Based on the initial mapping matrix M0 and the attenuated feedback weight matrix W', the dynamic adjustment matrix M is calculated. Step S5: The matrix conversion module receives the dynamic adjustment matrix M, replaces the initial mapping matrix M0, and remaps and converts the standardized stereo signal obtained in step S1 through the dynamic adjustment matrix M to obtain the optimized 5.1 channel signal. The feedback control module monitors the changes in the feedback signal in real time. When the changes in the feedback feature parameters are all less than the preset threshold, it is determined that the 5.1 channel signal has reached a steady state, and the matrix conversion module keeps the current dynamic adjustment matrix M unchanged. When the changes in the feedback feature parameters exceed the preset threshold, steps S3 and S4 are repeated.

[0007] Preferably, the method further includes step S6 after step S5: when the feedback acquisition module detects that the USBAudio stereo signal is interrupted, the feedback signal is abnormal, or the 5.1 channel output is abnormal, the feedback control module triggers the abnormal handling mechanism. Under this mechanism, the mapping matrix of the matrix conversion module is immediately switched back to the initial mapping matrix M0, and the power of the 5.1 channel output module is reduced to avoid damage to the device. After the abnormality is resolved, the system automatically returns to the dynamic matrix conversion mode and readjusts the feedback.

[0008] Preferably, the raw audio data includes amplitude, frequency, and phase information.

[0009] Preferably, in step S1, the stereo preprocessing module performs noise reduction, amplitude normalization, phase calibration, and frame synchronization processing on the raw audio data to obtain a standardized stereo signal. The frame synchronization processing is used to compensate for the audio data frame offset that occurs during USB transmission, ensuring the synchronization of the left and right channel signals of the USB Audio stereo signal. By detecting the synchronization flag bit of the USB audio frame, the transmission timing of the left and right channel data is adjusted so that the phase difference between the left and right channel signals is controlled within a certain range.

[0010] Preferably, in step S2, the initial mapping matrix M0 is a 6×2 dimensional matrix, corresponding to the 6 output channels of the 5.1 channel and the 2 input channels of the stereo signal. The initial mapping matrix M0 maps the normalized stereo signal to an initial 5.1 channel signal. The 5.1 channel signal includes the left front channel (FL), right front channel (FR), center channel (C), left surround channel (SL), right surround channel (SR), and low-frequency effects channel (LFE). The expression for the initial mapping matrix M0 is:

[0011]

[0012] Wherein, FL0, FR0, C0, SL0, SR0, and LFE0 are the signal amplitudes of each channel in the initial 5.1 channel, and L and R are the signal amplitudes of the left and right channels of the standardized stereo, respectively.

[0013] Preferably, the matrix elements of the initial mapping matrix M0 are preset initial mapping coefficients, which are used to realize the initial allocation of stereo signals in each channel of the 5.1 channel. The initial mapping coefficients are preset according to the sound field layout of the 5.1 channel and satisfy the following condition: FL0+FR0+C0+SL0+SR0+LFE0= L+R.

[0014] Preferably, the first dimension is the output signal F1 of the 5.1 channel output module, which includes the real-time amplitude, phase, and frequency information of each channel. The second dimension is the sound field feedback signal F2 of the external playback device, which includes the sound field distribution uniformity, the channel separation, and the feedback detection signal. The sampling frequency of signal F1 is consistent with the sampling frequency of the USB Audio stereo signal. Signal F2 is acquired by a sound field detection sensor. The sound field distribution uniformity is calculated by detecting the amplitude deviation of each channel playback signal within a preset sound field range. The channel separation is calculated by detecting the cross-interference of adjacent channel signals. The feedback detection signal is obtained by detecting a specific frequency peak in the feedback signal. This specific frequency peak is in the high frequency band of feedback. When the frequency peak detected in the high frequency band of feedback exceeds a preset threshold, it is determined that there is feedback interference.

[0015] Preferably, in step S4, the feedback characteristic parameters include: the deviation ΔA between the real-time amplitude and the initial amplitude of each channel, and the phase deviation ΔA of each channel. The parameters are: sound field distribution uniformity ε, channel separation σ, and howling interference intensity η. Based on the extracted feedback feature parameters, the mapping matrix coefficients of the matrix transformation module are dynamically adjusted through a multi-level, multi-dimensional complex matrix feedback mechanism to obtain the dynamic adjustment matrix M.

[0016] Preferably, step S4 includes: Step S41: Normalizing the extracted feedback feature parameters, mapping all feedback feature parameters to the [0,1] interval, eliminating the influence of different dimensions of the parameters, and obtaining normalized feedback feature parameters ΔA' and ΔA'. The normalized expressions are: ε', σ', η';

[0017]

[0018] Where x is the original feedback feature parameter, encompassing parameters ΔA and Δ ε, σ, η, and x' are the normalized feedback characteristic parameters, where x is the input parameter. min x is the minimum threshold of this parameter. max This is the maximum threshold for this parameter.

[0019] Step S42: Based on the normalized feedback feature parameters, calculate the feedback weight matrix W. The feedback weight matrix W is a 6×6 matrix, corresponding to the 6 channels of the 5.1 channel. Each channel corresponds to a weight coefficient, which is used to adjust the mapping coefficient of that channel to adjust the weight. The matrix elements W are... ij The calculation expression is:

[0020]

[0021] Where i and j are both 1~6, corresponding to 6 audio channels, ΔA' i Δ ' i Let α, β, γ, δ, and θ be the normalized amplitude deviation and normalized phase deviation of the i-th channel, respectively; α, β, γ, δ, and θ be the weighting coefficients of each feedback characteristic parameter, and α+β+γ+δ+θ=1. At the same time, a feedback attenuation coefficient λ is introduced, 0<λ<1, to attenuate the feedback weight matrix W. The attenuated feedback weight matrix W' is: W' = λ·W + (1-λ)·I, where I is a 6×6 identity matrix, and the value of λ is dynamically adjusted according to the fluctuation frequency of the feedback signal.

[0022] Sub-step S43: Based on the initial mapping matrix M0 and the decayed feedback weight matrix W', calculate the dynamic adjustment matrix M, which is a 6×2 dimensional matrix, and its calculation expression is as follows:

[0023]

[0024] Among them, K is the feedback regulation gain coefficient, where 1 < K < 2, which is used to adjust the sensitivity of the feedback regulation; ΔM is the matrix correction amount, and the matrix elements of ΔM, ΔM ij , where i = 1 to 6, j = 1 to 2, are jointly determined by the normalized feedback characteristic parameters and the initial mapping coefficient. The specific expression is:

[0025]

[0026] Among them, M 0ij is the element in the i-th row and j-th column of the initial mapping matrix M0. A USB transmission compensation term ΔT is introduced to compensate for the influence of signal attenuation and delay during USB transmission on matrix conversion. The calculation of ΔT is based on the transmission delay time τ of the USB audio frame, and the expression is: ΔT = kτ·M0, where k is the delay compensation coefficient, which is dynamically adjusted according to the USB transmission rate. The finally dynamically adjusted matrix M is corrected to: M = W'×M0 + K×ΔM + ΔT; Sub-step S44: Perform a stability check on the calculated dynamically adjusted matrix M.

[0027] Preferably, in the stability check, if the check standard is met, the dynamically adjusted matrix M is sent to the matrix conversion module; if the check standard is not met, the weight coefficients α, β, γ, δ, θ of the feedback weight matrix W and the feedback regulation gain coefficient K are readjusted, and sub-steps S42 to S44 are repeated until the check passes. Among them, the check standard is that the absolute values of all elements of the dynamically adjusted matrix M are within the interval [0, 1], and the condition number cond(M) of this matrix ≤ 10.

[0028] Generally speaking, the present invention discloses a method for implementing USB Audio stereo to 5.1-channel audio, aiming to solve problems such as sound field distortion, low channel separation, and lag in feedback regulation in traditional conversion methods. This method is applied to the USB Audio audio processing system. By collecting and preprocessing the USB stereo signal, an initial 5.1-channel signal is obtained through initial matrix mapping; two-dimensional feedback signals at the output end and the sound field are collected in real time, multiple types of characteristic parameters are extracted, and through a multi-level complex matrix feedback mechanism, the feedback weight matrix and the dynamically adjusted matrix are calculated, and the mapping matrix is dynamically adapted after stability verification; at the same time, a USB transmission compensation and exception handling mechanism is introduced to adapt to multiple scenarios. The present invention does not require additional hardware, realizes high-fidelity conversion, has strong anti-interference ability, adapts to multiple scenarios such as home and vehicle, and has high practicality and promotion value. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be discussed below. Obviously, the technical solutions described in conjunction with the accompanying drawings are only some embodiments of the present invention. For those skilled in the art, other embodiments and their accompanying drawings can be obtained based on the embodiments shown in these drawings without creative effort.

[0030] Figure 1 A general flowchart of the method for implementing USB Audio stereo to 5.1 channel audio according to the present invention is shown. Detailed Implementation

[0031] The technical solutions of various embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments described in the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0032] In summary, to overcome the shortcomings of existing USB Audio stereo to 5.1 channel conversion methods, such as sound field distortion, low channel separation, limited feedback adjustment, and weak anti-interference capabilities, this invention provides a method for converting USB Audio stereo to 5.1 channel audio. By designing a multi-level, multi-dimensional matrix feedback mechanism and combining it with the characteristics of USB Audio signal transmission, this invention achieves dynamic, accurate, and high-fidelity conversion of stereo signals to 5.1 channel signals. It also possesses strong anti-interference capabilities and adaptability, making it suitable for different types of USB audio devices and playback scenarios.

[0033] The following text will combine Figure 1 The method provided by this invention will be described in detail step by step. Figure 1 A general flowchart of the method for implementing USB Audio stereo to 5.1 channel audio according to the present invention is shown.

[0034] The USB Audio stereo to 5.1 channel audio conversion method provided by this invention has a specific application system architecture including a USB audio input module, a stereo preprocessing module, a matrix conversion module, a 5.1 channel output module, a feedback acquisition module, and a feedback control module. Based on this architecture, the method is implemented step by step as follows.

[0035] Step S1 aims to acquire and preprocess the USB Audio stereo signal.

[0036] In this step, the USB audio input module receives the USB Audio stereo signal (including the left channel L and the right channel R) transmitted from the external device through the USB interface, and parses the stereo signal according to the USB Audio protocol to extract the original audio data (including amplitude, frequency, and phase information).

[0037] The extracted raw audio data is then transmitted to the stereo preprocessing module. This module performs noise reduction, amplitude normalization, phase calibration, and frame synchronization on the raw audio data to obtain a standardized stereo signal. Frame synchronization is used to compensate for audio data frame offsets that occur during USB transmission, ensuring the synchronization of the left and right channel signals. Specifically, it detects the synchronization flag of the USB audio frame and adjusts the transmission timing of the left and right channel data to keep the phase difference between the left and right channel signals within a certain range, such as ±5°.

[0038] The next step is step S2, which is designed to perform the initial matrix mapping transformation.

[0039] In step S2, the standardized stereo signal obtained in step S1 is transmitted to the matrix conversion module. The matrix conversion module has a built-in initial mapping matrix M0, which is a 6×2-dimensional matrix (corresponding to the 6 output channels of 5.1 channel and the 2 input channels of stereo). The standardized stereo signal is mapped to an initial 5.1 channel signal through the initial mapping matrix M0. The 5.1 channel signal includes the left front channel FL, the right front channel FR, the center channel C, the left surround channel SL, the right surround channel SR, and the low-frequency effects channel LFE.

[0040] The expression for the initial mapping matrix M0 is:

[0041]

[0042] Wherein, FL0, FR0, C0, SL0, SR0, and LFE0 are the signal amplitudes of each channel in the initial 5.1 channel, and L and R are the signal amplitudes of the standardized stereo left and right channels, respectively.

[0043] The matrix elements of the initial mapping matrix M0 are preset initial mapping coefficients, which are used to achieve the initial allocation of stereo signals in each channel of the 5.1 channel. The initial mapping coefficients are preset according to the sound field layout of the 5.1 channel and satisfy the following conditions: FL0+FR0+C0+SL0+SR0+LFE0= L+R (ensuring the conservation of total signal energy). Preferably, the signal amplitude C0 of the center channel is 0.6 to 0.8 times the average amplitude of the left and right channels, the amplitude of the low frequency effect channel LFE0 is 0.3 to 0.5 times the maximum amplitude of the left and right channels, and the amplitudes of the surround channels SL0 and SR0 are 0.4 to 0.6 times the corresponding amplitudes of the left and right channels.

[0044] The next step is step S3, which aims to complete the 5.1 channel signal output and feedback acquisition.

[0045] In this step, the matrix conversion module transmits the initial 5.1 channel signal to the 5.1 channel output module. The 5.1 channel output module amplifies and compensates the power of the initial 5.1 channel signal before outputting it to an external playback device (such as a 5.1 channel speaker).

[0046] Meanwhile, the feedback acquisition module collects feedback signals in two dimensions in real time: the first dimension is the output signal of the 5.1 channel output module (denoted as feedback signal F1, including the real-time amplitude, phase and frequency information of each channel), and the second dimension is the sound field feedback signal of the external playback device (denoted as feedback signal F2, including the sound field distribution uniformity, the separation of each channel and the howling detection signal).

[0047] Preferably, the sampling frequency of the feedback signal F1 is consistent with the sampling frequency of the USB Audio signal (adjustable from 44.1kHz to 192kHz) to ensure real-time acquisition. The feedback signal F2 is acquired by a sound field detection sensor. The uniformity of the sound field distribution is calculated by detecting the amplitude deviation of the playback signal of each channel within a preset sound field range. The channel separation is calculated by detecting the cross-interference of adjacent channel signals. The feedback detection signal is obtained by detecting specific frequency peaks (2kHz to 8kHz, the high frequency band of feedback) in the feedback signal. When the frequency peak in this band exceeds a preset threshold, feedback interference is determined to exist.

[0048] Next, we come to step S4, which is the core step and aims to achieve complex matrix feedback regulation.

[0049] In this step, the feedback control module receives feedback signals F1 and F2 transmitted by the feedback acquisition module, analyzes and processes the feedback signals, and extracts feedback feature parameters, including: the deviation ΔA between the real-time amplitude and the initial amplitude of each channel, and the phase deviation ΔA of each channel. The parameters are: sound field distribution uniformity ε, channel separation σ, and howling interference intensity η. Based on the extracted feedback feature parameters, the mapping matrix coefficients of the matrix transformation module are dynamically adjusted through a multi-level, multi-dimensional complex matrix feedback mechanism to obtain the dynamic adjustment matrix M. Since this step is the core step, it is described in detail as a sub-step.

[0050] In step S4, sub-step S41 is first executed to perform feedback feature parameter normalization processing.

[0051] In sub-step S41, the extracted feedback feature parameters are first normalized, mapping all feedback feature parameters to the [0,1] interval to eliminate the influence of different dimensions of the parameters, thus obtaining the normalized feedback feature parameters ΔA' and ΔA'. The normalized expressions are: ε', σ', η';

[0052]

[0053] Where x is the original feedback feature parameter, and x' is the normalized feedback feature parameter, x min x is the minimum threshold of this parameter. max This is the maximum threshold for this parameter.

[0054] For example, the minimum threshold of amplitude deviation ΔA is 0 (no deviation), and the maximum threshold is 50% of the initial amplitude (too large deviation). The above formula normalizes ΔA to ΔA' = (ΔA-0) / (50%-0).

[0055] The process proceeds further to sub-step S42, which aims to calculate the feedback weight matrix.

[0056] Based on the normalized feedback characteristic parameters, the feedback weight matrix W is calculated. The feedback weight matrix W is a 6×6 matrix (corresponding to the 6 channels of a 5.1 channel, with each channel corresponding to a weight coefficient used to adjust the mapping coefficient of that channel). The matrix elements W... ij The calculation expression for (i and j are both 1~6, corresponding to 6 channels) is:

[0057]

[0058] Wherein, ΔA' i Δ ' i α, β, γ, δ, and θ are the normalized amplitude deviation and normalized phase deviation of the i-th channel, respectively; α, β, γ, δ, and θ are the weighting coefficients of each feedback characteristic parameter, and α+β+γ+δ+θ=1.

[0059] The weighting coefficients can be dynamically adapted according to the playback scenario. For example, in home audio-visual scenarios, the weighting coefficients (γ, δ) of sound field distribution uniformity ε and channel separation σ are set to 0.3~0.4, and the weighting coefficient θ of howling interference intensity η is set to 0.1~0.2; in in-vehicle entertainment scenarios, the weighting coefficient θ of howling interference intensity η is set to 0.3~0.4 to ensure anti-interference capability.

[0060] Meanwhile, a feedback attenuation coefficient λ (0 < λ < 1) is introduced to attenuate the feedback weight matrix W, avoiding signal oscillation caused by excessive feedback regulation. The attenuated feedback weight matrix W' is: W' = λ·W + (1 - λ)·I, where I is a 6×6 identity matrix. The value of λ is dynamically adjusted according to the fluctuation frequency of the feedback signal. The higher the signal fluctuation frequency, the smaller the value of λ (for example, 0.3 - 0.5), and the lower the signal fluctuation frequency, the larger the value of λ (for example, 0.6 - 0.8).

[0061] Next, sub-step S43 is executed, which aims to implement dynamic adjustment matrix calculation.

[0062] In sub-step S43, based on the initial mapping matrix M0 and the attenuated feedback weight matrix W', the dynamic adjustment matrix M is calculated. The dynamic adjustment matrix M is a 6×2 matrix, and its calculation expression is:[[]]

[0063]

[0064] where K is the feedback regulation gain coefficient (1 < K < 2), which is used to adjust the sensitivity of feedback regulation; ΔM is the matrix correction amount, and the matrix elements ΔM ij (i = 1~6, j = 1~2) are jointly determined by the normalized feedback characteristic parameters and the initial mapping coefficients. The specific expression is:[[]]

[0065]

[0066] where M 0ij is the element in the i-th row and j-th column of the initial mapping matrix M0.

[0067] The role of the matrix correction amount ΔM is to accurately correct each element of the initial mapping matrix according to the feedback characteristic parameters, ensuring that the mapping coefficients of each channel can be dynamically adapted according to the feedback signal.

[0068] In addition, a USB transmission compensation term ΔT is introduced to compensate for the impact of signal attenuation and delay in the USB transmission process on matrix conversion. The calculation of ΔT is based on the transmission delay time τ of the USB audio frame, and the expression is: ΔT = kτ·M0 (k is the delay compensation coefficient, which is dynamically adjusted according to the USB transmission rate). The final dynamically adjusted matrix M is corrected to: M = W'×M0 + K×ΔM + ΔT.

[0069] Next, sub-step S44 is executed, which aims to implement matrix stability verification.

[0070] The calculated dynamic adjustment matrix M is subjected to stability verification. In particular, a certain verification standard needs to be set for this stability verification. The setting of the verification standard will be described in detail below.

[0071] In the corresponding verification standard settings, the absolute values ​​of all elements of the dynamically adjusted matrix M are within the range of [0,1] (to ensure that the mapping coefficients are reasonable and that the signal amplitude is not overloaded), and the condition number of the matrix, cond(M), is ≤10 (to ensure the stability of the matrix and avoid signal oscillation caused by feedback adjustment).

[0072] If the above verification criteria are met, i.e. the verification passes, the dynamically adjusted matrix M will be sent to the matrix conversion module.

[0073] If the above verification criteria are not met, i.e. the verification fails, the weight coefficients α, β, γ, δ, θ of the feedback weight matrix W and the feedback adjustment gain coefficient K are readjusted, and sub-steps S42 to S44 are repeated until the verification passes.

[0074] Next, step S5 is executed, which aims to achieve dynamic matrix transformation and steady-state output.

[0075] In step S5, the matrix conversion module receives the dynamic adjustment matrix M sent by the feedback control module, replaces the initial mapping matrix M0, and remaps and converts the standardized stereo signal obtained in step S1 through the dynamic adjustment matrix M to obtain the optimized 5.1 channel signal.

[0076] Simultaneously, the feedback control module monitors changes in the feedback signal in real time. When the changes in the feedback characteristic parameters are all less than a preset threshold (e.g., ΔA' < 0.05, Δ...), the system will detect and control the changes. When ε'<0.05, σ'<0.1, η'<0.1, and η'<0.05, the 5.1 channel signal is determined to have reached a steady state. The matrix conversion module keeps the current dynamic adjustment matrix M unchanged and continues to output the optimized 5.1 channel signal.

[0077] When the change in the feedback characteristic parameter exceeds the preset threshold, repeat steps S3 to S4 to readjust the feedback and matrix, ensuring the stability and fidelity of the output signal.

[0078] Next, we proceed to the optimization step S6, which aims to achieve exception handling and adaptive adaptation.

[0079] In this step, when the feedback acquisition module detects a USB Audio signal interruption, an abnormal feedback signal (such as a feedback signal amplitude of 0 or a howling interference intensity η' ≥ 0.8), or an abnormal 5.1 channel output, the feedback control module triggers an abnormality handling mechanism. Under this mechanism, the mapping matrix of the matrix conversion module is immediately switched back to the initial mapping matrix M0, and the power of the 5.1 channel output module is reduced to avoid damage to the equipment. After the abnormality is resolved, it automatically returns to the dynamic matrix conversion mode and readjusts the feedback.

[0080] In addition, the feedback control module has a built-in scene adaptation library, which stores preset values ​​of the initial mapping matrix coefficients, feedback weight coefficients α~θ, and feedback adjustment gain coefficient K for different playback scenarios (home audio-visual, in-car entertainment). Users can set playback scenarios through the USB interface, and the feedback control module will automatically call the preset parameters in the scene adaptation library according to the user-set scenario to achieve adaptive adaptation to different scenarios and improve the targeting of the conversion effect.

[0081] This concludes the general overview of the basic structure and technical content of this invention. As can be seen from the above technical description, this invention has many beneficial effects compared to existing technologies.

[0082] First, this invention adopts a multi-level, multi-dimensional matrix feedback mechanism, which breaks through the limitations of simple single-dimensional feedback in the existing technology. By collecting feedback information from two dimensions, namely the output signal and the sound field feedback signal, multiple types of feedback feature parameters are extracted, and a feedback weight matrix and a dynamic adjustment matrix are constructed. This achieves accurate and dynamic adjustment of the mapping matrix coefficients, effectively solving the problems of sound field distortion and low channel separation in traditional conversion methods, and improving the fidelity and immersion of 5.1 channel audio.

[0083] Secondly, this invention introduces a USB transmission compensation term and frame synchronization processing to specifically compensate for problems such as signal attenuation, delay, and frame offset that exist in the USB Audio protocol transmission process. This reduces the impact of USB transmission quality on the conversion effect, improves the anti-interference capability of the method, and ensures stable conversion in different USB transmission environments.

[0084] Furthermore, this invention features a matrix stability verification mechanism and an anomaly handling mechanism to prevent signal oscillations caused by excessive feedback adjustment, while also preventing damage to the device from abnormal situations such as USB signal interruption and howling interference, thus improving the stability and reliability of the method.

[0085] Furthermore, the present invention incorporates a scene adaptation library, which can adaptively adjust the feedback parameters and initial parameters of the mapping matrix according to different playback scenarios, exhibiting strong adaptability and meeting the usage needs of various scenarios such as home use and in-vehicle use.

[0086] Furthermore, in this invention, the entire conversion process is based on digital signal processing, requiring no additional hardware devices and can be implemented solely through software algorithms. It is low-cost, easy to integrate, and can be widely applied to various audio devices that support the USB Audio protocol, thus possessing high practicality and promotional value.

[0087] To facilitate understanding, specific embodiments will be provided below to further illustrate the present invention. It should be noted that the specific values, module models, and related protocols involved in these embodiments are only for the purpose of making the explanation more intuitive, and do not constitute a limitation on the present invention.

[0088] In this embodiment, a framework for a USB Audio stereo to 5.1 channel audio processing system that implements the method of the present invention is provided, including a USB audio input module, a stereo preprocessing module, a matrix conversion module, a 5.1 channel output module, a feedback acquisition module, and a feedback control module.

[0089] The USB audio input module uses a USB 2.0 interface, supports the USB Audio 1.0 / 2.0 protocol, and can receive stereo signals with a sampling frequency of 44.1kHz and quantization precision of 16bit. The stereo preprocessing module uses a DSP processor to perform noise reduction, normalization, phase calibration, and frame synchronization processing of the audio signal. The matrix conversion module is integrated into the DSP processor to perform matrix mapping conversion of stereo signals to 5.1 channel signals. The 5.1 channel output module uses a multi-channel power amplifier (output power ≥20W×6) and supports 6 independent channel outputs. The feedback acquisition module includes 6 channel signal acquisition sensors (for acquiring the output feedback signal F1) and 1 sound field detection sensor for acquiring the sound field feedback signal F2. The feedback control module uses an ARM processor and communicates with the DSP processor through an SPI interface to perform feedback signal parsing, matrix coefficient adjustment, and control logic execution.

[0090] Based on the above system architecture, this embodiment provides a method for converting USB Audio stereo to 5.1 channel audio. Based on the above system, the specific steps include:

[0091] The first step is S1, as described above. This step aims to achieve USB Audio stereo signal acquisition and preprocessing.

[0092] In this step, the USB audio input module receives the USB Audio stereo signal (left and right channel signals) transmitted from the computer via the USB 2.0 interface, with a sampling frequency set to 44.1kHz and a quantization precision of 16bit. The stereo signal is parsed according to the USBAudio 2.0 protocol to extract the raw audio data of the left and right channels (amplitude range 0~32767, frequency range 20Hz~20kHz, phase range 0~360°). The extracted raw audio data is then transmitted to the stereo preprocessing module 2 for the following preprocessing operations.

[0093] The first step is noise reduction, which uses an adaptive Wiener filtering algorithm to remove high-frequency noise (above 20kHz) and power line interference (50Hz) from the original audio data. The filter window size is set to 256 points.

[0094] The second step is amplitude normalization. Specifically, the amplitudes of the left and right channels are mapped to the [0,1] interval. The normalization formula is: A_norm = A_original / 32767, where A_original is the original amplitude and A_norm is the normalized amplitude.

[0095] The next step is phase calibration, which involves detecting the phase difference between the left and right channels. If the phase difference exceeds ±5°, the phase of the right channel is adjusted to keep the phase difference between the left and right channels within ±5°.

[0096] Finally, there is frame synchronization processing, which involves detecting the synchronization flag of the USB audio frame. When an abnormality is detected in the synchronization flag, the transmission timing of the left and right channel data is adjusted to compensate for the frame offset that occurs during USB transmission, ensuring that the left and right channel signals are transmitted synchronously to the matrix conversion module. After preprocessing, the standardized stereo signal (L_norm, R_norm) is obtained.

[0097] The next step is step S2, as described above, which aims to perform the initial matrix mapping transformation.

[0098] In this step, the matrix transformation module has a built-in initial mapping matrix M0, which is a 6×2 dimensional matrix. Initial mapping coefficients are preset based on the home audio-visual scenario. The specific values ​​of the initial mapping matrix M0 are:

[0099]

[0100] The normalized stereo signals (L_norm, R_norm) are mapped to initial 5.1 channel signals using the initial mapping matrix M0. The calculation process is as follows:

[0101] FL0 = 0.7×L_norm + 0.1×R_norm

[0102] FR0 = 0.1×L_norm + 0.7×R_norm

[0103] C0 = 0.6×L_norm + 0.6×R_norm

[0104] SL0 = 0.4×L_norm + 0.0×R_norm

[0105] SR0 = 0.0×L_norm + 0.4×R_norm

[0106] LFE0 = 0.3×L_norm + 0.3×R_norm

[0107] After the initial 5.1 channel signal is calculated, it is transmitted to the 5.1 channel output module.

[0108] The next step is step S3, as described above, which aims to achieve 5.1 channel signal output and feedback acquisition.

[0109] The 5.1 channel output module amplifies the initial 5.1 channel signal (amplification factor set to 10x) and performs frequency compensation on the low-frequency effect channel LFE (increasing the amplitude of frequencies below 100Hz), and then outputs it to the 5.1 channel speakers (external playback device); at the same time, the feedback acquisition module acquires feedback signals F1 and F2 in real time at a acquisition frequency of 44.1kHz.

[0110] For feedback signal F1, acquire the channel signals from the output of the 5.1 channel output module, and extract the real-time amplitude A_FL, A_FR, A_C, A_SL, A_SR, and A_LFE of each channel, as well as the real-time phase. _FL、 _FR、 _C、 _SL、 _SR、 _LFE, real-time frequencies f_FL, f_FR, f_C, f_SL, f_SR, f_LFE.

[0111] For feedback signal F2, the sound field distribution uniformity ε (calculating the amplitude deviation of each channel signal within a 1m×1m×1m sound field range, ε=0.05~0.2), channel separation σ (calculating the cross-interference of adjacent channel signals, σ=0.7~0.95), and howling interference intensity η (detecting the signal amplitude in the 2kHz~8kHz frequency band, η=0~0.3) are collected by the sound field detection sensor; the collected feedback signals F1 and F2 are transmitted to the feedback control module.

[0112] The next step is S4, as described above, which aims to achieve complex matrix feedback regulation.

[0113] In this step, the feedback control module analyzes the feedback signals F1 and F2, extracts the feedback feature parameters, and then dynamically adjusts the mapping matrix coefficients through a complex matrix feedback mechanism. Specifically, it includes the following sub-steps.

[0114] The first step is sub-step S41, which aims to normalize the feedback feature parameters.

[0115] Specifically, the threshold ranges for each feedback characteristic parameter are set: amplitude deviation ΔA (0~0.5), phase deviation Δ (0~30°), sound field distribution uniformity ε (0.05~0.2), channel separation σ (0.7~0.95), howling interference intensity η (0~0.8); according to the normalization formula The feedback feature parameters are normalized to obtain the normalized feedback feature parameters:

[0116] ΔA'_FL = (A_FL - FL0) / (0.5×FL0), ΔA'_FR = (A_FR - FR0) / (0.5×FR0), and so on, to obtain ΔA' for each channel. i .

[0117] Δ '_FL = ( _FL - 0_FL) / 30, Δ '_FR = ( _FR - 0_FR) / 30, and so on, to obtain the Δ for each channel. ' i ( 0_FL、 0_FR is the initial phase);

[0118] ε' = (ε - 0.05) / (0.2 - 0.05) = (ε - 0.05) / 0.15;

[0119] σ' = (σ - 0.7) / (0.95 - 0.7) = (σ - 0.7) / 0.25;

[0120] η' = η / 0.8;

[0121] After normalization, all feedback feature parameters are mapped to the [0,1] interval.

[0122] The next step is sub-step S42, which is designed to perform the calculation of the feedback weight matrix.

[0123] This embodiment is for a home audio-visual scenario, and the weighting coefficients of each feedback feature parameter are set as follows: α=0.25, β=0.25, γ=0.3, δ=0.15, θ=0.05 (α+β+γ+δ+θ=1).

[0124] The feedback attenuation coefficient λ = 0.7 (for lower signal fluctuation frequencies, a larger value is used); according to the formula... Calculate the feedback weight matrix W (6×6 dimension), and then calculate the attenuated feedback weight matrix W' (I is a 6×6 dimension identity matrix) by W' = λ·W + (1-λ)·I.

[0125] Assume the normalized feedback feature parameter obtained analytically is: ΔA' i =0.1 (mean of each channel), Δ ' i =0.08 (mean of each channel), ε'=0.2, σ'=0.8, η'=0.1, then the weight element W ij =0.25×0.1 + 0.25×0.08 +0.3×0.2 + 0.15×0.8 + 0.05×0.1=0.025+0.02+0.06+0.12+0.005=0.23, W' after attenuation ij =0.7×0.23 + 0.3×1=0.161+0.3=0.461, and so on, to obtain the complete feedback weight matrix W'.

[0126] Then comes sub-step S43, which aims to achieve dynamic adjustment of the matrix calculation.

[0127] Set the feedback adjustment gain coefficient K=1.5, USB transmission delay time τ=1ms, and delay compensation coefficient k=0.001; according to the formula Calculate the matrix correction ΔM, for example, ΔM 11 (Corresponding to the mapping correction between FL and L channels) = 0.7 × (0.1 + 0.08 - 0.2 + 0.8 - 0.1) = 0.7 × 0.68 = 0.476; USB transmission compensation term ΔT = kτ·M0 = 0.001 × 1 × M0 = 0.001 × M0; Then, according to the formula M = W' × M0 + K × ΔM + ΔT, the dynamic adjustment matrix M is calculated, and the specific values ​​are as follows (example):

[0128]

[0129] This is followed by sub-step S44, which is designed to perform matrix stability verification.

[0130] In this sub-step, the stability of the dynamically adjusted matrix M is checked. All elements of M are within the interval [0,1] (all elements in the above example satisfy this condition). Then, the condition number cond(M) of matrix M is calculated as 5.2≤10, which meets the stability requirements. The check passes, and the dynamically adjusted matrix M is then sent to the matrix conversion module via the SPI interface.

[0131] The next step is S5, which aims to achieve dynamic matrix transformation and steady-state output.

[0132] In this step, the matrix conversion module receives the dynamically adjusted matrix M, replaces the initial mapping matrix M0, and remaps and converts the normalized stereo signal (L_norm, R_norm) through M to obtain the optimized 5.1 channel signal.

[0133] FL = 0.72×L_norm + 0.11×R_norm

[0134] FR = 0.11×L_norm + 0.72×R_norm

[0135] C = 0.63×L_norm + 0.63×R_norm

[0136] SL = 0.42×L_norm + 0.01×R_norm

[0137] SR = 0.01×L_norm + 0.42×R_norm

[0138] LFE = 0.31×L_norm + 0.31×R_norm

[0139] The feedback control module monitors changes in feedback characteristic parameters in real time. When ΔA' < 0.05 and ΔA' < 0.05 for 500 ms consecutively, the response is triggered. When ε'<0.05, σ'<0.1, σ'<0.1, and η'<0.05, the 5.1 channel signal is determined to have reached a steady state. The matrix conversion module keeps the current dynamic adjustment matrix M unchanged and continues to output the optimized 5.1 channel signal. When the change in the feedback characteristic parameters exceeds the preset threshold, steps S3~S4 are repeated to re-adjust the feedback and matrix.

[0140] The next step is S6, which aims to achieve exception handling and adaptive adaptation.

[0141] In this step, when the feedback acquisition module detects a USB audio signal interruption (no USB audio frames are acquired for 100ms), the feedback control module triggers an exception handling mechanism, immediately switching the mapping matrix of the matrix conversion module back to the initial mapping matrix M0, and simultaneously reducing the power of the 5.1 channel output module to 50%; after the USB signal is restored, it automatically returns to the dynamic matrix conversion mode and readjusts the feedback.

[0142] If the user switches the playback scene to the in-vehicle entertainment scene via the USB interface, the feedback control module calls the preset parameters of the in-vehicle scene from the scene adaptation library: α=0.2, β=0.2, γ=0.15, δ=0.15, θ=0.3, λ=0.5, K=1.8, and recalculates the feedback weight matrix and dynamic adjustment matrix to achieve adaptive adaptation of the in-vehicle scene.

[0143] The above description is merely an exemplary embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for converting USB Audio stereo to 5.1 channel audio, characterized in that, This method is based on a system architecture that includes a USB audio input module, a stereo preprocessing module, a matrix conversion module, a 5.1 channel output module, a feedback acquisition module, and a feedback control module. The method includes the following steps: Step S1: The USB audio input module receives the USB Audio stereo signal transmitted from the external device through the USB interface, parses the stereo signal according to the USB Audio protocol, extracts the raw audio data, and transmits the raw audio data to the stereo preprocessing module for preprocessing to obtain a standardized stereo signal. Step S2: The standardized stereo signal is transmitted to the matrix conversion module, which has a built-in initial mapping matrix M0. The initial mapping matrix M0 maps the standardized stereo signal into an initial 5.1 channel signal. Step S3: The initial 5.1 channel signal is transmitted to the 5.1 channel output module. The 5.1 channel output module amplifies and compensates the power of the initial 5.1 channel signal and then outputs it to the external playback device. The feedback acquisition module collects feedback signals in two dimensions in real time. The first dimension is the output signal F1 of the 5.1 channel output module, and the second dimension is the sound field feedback signal F2 of the external playback device. Step S4: The feedback control module receives feedback signals F1 and F2, analyzes and processes these two feedback signals, extracts feedback feature parameters, calculates the feedback weight matrix W based on the feedback feature parameters, and introduces the feedback attenuation coefficient λ to calculate the attenuated feedback weight matrix W'. Based on the initial mapping matrix M0 and the attenuated feedback weight matrix W', the dynamic adjustment matrix M is calculated. Step S5: The matrix conversion module receives the dynamic adjustment matrix M, replaces the initial mapping matrix M0, and remaps and converts the standardized stereo signal obtained in step S1 through the dynamic adjustment matrix M to obtain the optimized 5.1 channel signal. The feedback control module monitors the changes in the feedback signal in real time. When the changes in the feedback characteristic parameters are all less than the preset threshold, it is determined that the 5.1 channel signal has reached a steady state, and the matrix conversion module keeps the current dynamic adjustment matrix M unchanged. When the changes in the feedback characteristic parameters exceed the preset threshold, steps S3 and S4 are repeated.

2. The method according to claim 1, characterized in that, The method further includes step S6 after step S5: when the feedback acquisition module detects that the USB Audio stereo signal is interrupted, the feedback signal is abnormal, or the 5.1 channel output is abnormal, the feedback control module triggers the abnormal handling mechanism. Under this mechanism, the mapping matrix of the matrix conversion module is immediately switched back to the initial mapping matrix M0, and the power of the 5.1 channel output module is reduced to avoid damage to the device. After the abnormality is resolved, it automatically returns to the dynamic matrix conversion mode and readjusts the feedback.

3. The method according to claim 1, characterized in that, Raw audio data includes amplitude, frequency, and phase information.

4. The method according to claim 1, characterized in that, In step S1, the stereo preprocessing module performs noise reduction, amplitude normalization, phase calibration, and frame synchronization processing on the raw audio data to obtain a standardized stereo signal. The frame synchronization processing is used to compensate for the audio data frame offset that occurs during USB transmission, ensuring the synchronization of the left and right channel signals of the USB Audio stereo signal. By detecting the synchronization flag bit of the USB audio frame, the transmission timing of the left and right channel data is adjusted so that the phase difference between the left and right channel signals is controlled within a certain range.

5. The method according to claim 1, characterized in that, In step S2, the initial mapping matrix M0 is a 6×2 dimensional matrix, corresponding to the 6 output channels of the 5.1 channel and the 2 input channels of the stereo signal. The initial mapping matrix M0 maps the normalized stereo signal to an initial 5.1 channel signal. The 5.1 channel signal includes the left front channel (FL), right front channel (FR), center channel (C), left surround channel (SL), right surround channel (SR), and low-frequency effects channel (LFE). The expression for the initial mapping matrix M0 is: Wherein, FL0, FR0, C0, SL0, SR0, and LFE0 are the signal amplitudes of each channel in the initial 5.1 channel, and L and R are the signal amplitudes of the left and right channels of the standardized stereo, respectively.

6. The method according to claim 5, characterized in that, The matrix elements of the initial mapping matrix M0 are preset initial mapping coefficients, which are used to realize the initial allocation of stereo signals in each channel of the 5.1 channel. The initial mapping coefficients are preset according to the sound field layout of the 5.1 channel and satisfy the following condition: FL0+FR0+C0+SL0+SR0+LFE0= L+R.

7. The method according to claim 1, characterized in that, The first dimension is the output signal F1 of the 5.1 channel output module. Signal F1 includes the real-time amplitude, phase, and frequency information of each channel. The second dimension is the sound field feedback signal F2 of the external playback device. Signal F2 includes the sound field distribution uniformity, channel separation, and feedback detection signal. The sampling frequency of signal F1 is consistent with the sampling frequency of the USB Audio stereo signal. Signal F2 is acquired by a sound field detection sensor. The sound field distribution uniformity is calculated by detecting the amplitude deviation of each channel playback signal within a preset sound field range. The channel separation is calculated by detecting the cross-interference of adjacent channel signals. The feedback detection signal is obtained by detecting a specific frequency peak in the feedback signal. This specific frequency peak is in the high-frequency band of feedback. When the frequency peak detected in this high-frequency band exceeds a preset threshold, feedback interference is determined to exist.

8. The method according to claim 1, characterized in that, In step S4, the feedback characteristic parameters include: the deviation ΔA between the real-time amplitude and the initial amplitude of each channel, and the phase deviation ΔA of each channel. The parameters are: sound field distribution uniformity ε, channel separation σ, and howling interference intensity η. Based on the extracted feedback feature parameters, the mapping matrix coefficients of the matrix transformation module are dynamically adjusted through a multi-level, multi-dimensional complex matrix feedback mechanism to obtain the dynamic adjustment matrix M.

9. The method according to claim 8, characterized in that, Step S4 includes: Step S41: Normalize the extracted feedback feature parameters, mapping all feedback feature parameters to the [0,1] interval to eliminate the influence of different dimensions of the parameters, and obtain the normalized feedback feature parameters ΔA' and ΔA'. The normalized expressions are: ε', σ', η'; Where x is the original feedback feature parameter, encompassing parameters ΔA and Δ ε, σ, η, and x' are the normalized feedback characteristic parameters, where x is the input parameter. min x is the minimum threshold of this parameter. max This is the maximum threshold value for this parameter. Step S42: Based on the normalized feedback feature parameters, calculate the feedback weight matrix W. The feedback weight matrix W is a 6×6 matrix, corresponding to the 6 channels of the 5.1 channel. Each channel corresponds to a weight coefficient, which is used to adjust the mapping coefficient of that channel to adjust the weight. The matrix elements W are... ij The calculation expression is: Where i and j are both 1~6, corresponding to 6 audio channels, ΔA' i Δ ' i Let be the normalized amplitude deviation and normalized phase deviation of the i-th channel, respectively; α, β, γ, δ, and θ are the weighting coefficients of each feedback characteristic parameter, and α+β+γ+δ+θ=1. Simultaneously, a feedback attenuation coefficient λ is introduced, 0 < λ < 1, to attenuate the feedback weight matrix W. The attenuated feedback weight matrix W' is: W' = λ·W + (1-λ)·I, where I is a 6×6 identity matrix, and the value of λ is dynamically adjusted according to the fluctuation frequency of the feedback signal. Sub-step S43: Based on the initial mapping matrix M0 and the decayed feedback weight matrix W', calculate the dynamic adjustment matrix M, which is a 6×2 dimensional matrix, and its calculation expression is as follows: Among them, K is the feedback regulation gain coefficient, where 1 < K < 2, which is used to adjust the sensitivity of the feedback regulation; ΔM is the matrix correction amount, and the matrix element ΔM ij , where i = 1~6 and j = 1~2, is jointly determined by the normalized feedback characteristic parameters and the initial mapping coefficient, and the specific expression is: Among them, M 0ij The element in the i-th row and j-th column of the initial mapping matrix M0, A USB transmission compensation term ΔT is introduced to compensate for the impact of signal attenuation and delay during USB transmission on matrix transformation. The calculation of ΔT is based on the transmission delay time τ of the USB audio frame, and the expression is: ΔT = kτ·M0, where k is the delay compensation coefficient, which is dynamically adjusted according to the USB transmission rate. The final dynamic adjustment matrix M is corrected as: M = W'×M0+ K×ΔM +ΔT. Sub-step S44: Perform stability verification on the calculated dynamic adjustment matrix M.

10. The method according to claim 9, characterized in that, During stability verification, if the verification criteria are met, the dynamically adjusted matrix M is sent to the matrix transformation module; if the verification criteria are not met, the weight coefficients α, β, γ, δ, θ of the feedback weight matrix W and the feedback adjustment gain coefficient K are readjusted, and sub-steps S42~S44 are repeated until the verification passes. The verification criterion is that the absolute values ​​of all elements of the dynamically adjusted matrix M are within the interval [0,1], and the condition number cond(M) of the matrix is ​​≤10.

Images