Hearing aid acoustic feedback suppression methods, systems, computer equipment, storage media, and hearing aids

By combining a sliding mode observer and a variable step size normalized least mean square algorithm, the problems of acoustic path abrupt change and insufficient nonlinear processing capability in hearing aid acoustic feedback suppression methods are solved, achieving fast, low-energy feedback suppression and high-quality sound output.

CN122676794BActive Publication Date: 2026-11-10SHENZHEN BAILINGNIAO HEARING TECH CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202611141721.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2026-07-30
Publication Date
2026-11-10
Estimated Expiration
2046-07-30

AI Technical Summary

Technical Problem

Existing acoustic feedback suppression methods for hearing aids struggle to meet users' demands for real-time performance and high-quality sound when faced with issues such as abrupt changes in acoustic paths, insufficient nonlinear processing capabilities, and slow response speeds.

Method used

A sliding mode observer is used to detect changes in acoustic feedback path parameters. Combined with the iterative formula of a first- or second-order sliding mode observer and the variable step-size normalized minimum mean square algorithm, the cancellation signal is acquired in real time for compensation. In low-power mode, energy consumption is reduced through fixed-point modification.

Benefits of technology

It achieves rapid suppression of hearing aid feedback, meets the requirements of high-quality sound, and effectively suppresses feedback in its early stages, thus reducing energy consumption.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122676794B_ABST
    Figure CN122676794B_ABST
Patent Text Reader

Abstract

This invention relates to a method, system, computer device, storage medium, and hearing aid for suppressing acoustic feedback. The method includes: real-time acquisition of raw signals within the usage scenario of the hearing aid; obtaining the operating mode of the hearing aid, including a low-power mode and at least one normal mode; if the operating mode is normal mode, then: using a sliding mode observer to detect changes in the acoustic feedback path parameters of the raw signal; if a sudden change occurs, acquiring a cancellation signal; and compensating the output signal of the hearing aid based on the cancellation signal; wherein, if the change in the acoustic feedback path parameters exceeds a preset threshold, a sudden change in the acoustic feedback path parameters is confirmed. This invention can compensate for the output signal, thereby achieving rapid suppression of hearing aid whistling.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of hearing aids, and more particularly to a method, system, computer device, storage medium, and hearing aid for suppressing acoustic feedback in hearing aids. Background Technology

[0002] A hearing aid is a small amplifier that amplifies sounds that are otherwise inaudible. By utilizing the residual hearing of the hearing-impaired person, the sound can be sent to the auditory center of the brain, allowing the person to perceive the sound, thus bringing great convenience to the hearing-impaired.

[0003] Acoustic feedback whistling in hearing aids is a core technical problem that needs to be solved during the use of hearing aids. Existing technologies mainly use adaptive filtering algorithms such as LMS and NLMS to achieve acoustic feedback suppression. These methods generate cancellation signals by identifying acoustic feedback paths and have a basic suppression effect in normal, undisturbed scenarios.

[0004] However, existing adaptive filtering technology has unavoidable technical defects, specifically: First, it has poor robustness to sudden changes in the feedback path. Scenarios such as bringing a mobile phone close to the ear or wearing a hat / glasses, which change the acoustic path, constitute system parameter perturbations and external disturbances. The adaptive filter needs to iterate and converge again, and feedback is very likely to occur during the convergence process, resulting in poor transient response. Second, it has insufficient nonlinear processing capability. The miniature speakers and microphones of hearing aids have severe saturation and harmonic nonlinear distortion. Adaptive filtering is based on the assumption of a linear system, and when processing nonlinear audio signals, algorithm divergence or parameter misadjustment is likely to occur, which greatly reduces the feedback suppression effect. Third, it has a slow response speed. Adaptive filtering belongs to the "error-based learning" mode, which requires multiple iterations to reduce errors. It cannot achieve rapid suppression in the early stage of feedback and is difficult to meet real-time requirements.

[0005] It is evident that existing hearing aid acoustic feedback suppression methods are insufficient to meet users' needs. Summary of the Invention

[0006] In order to solve the above-mentioned technical problems, or at least partially solve the above-mentioned technical problems, the present invention provides a hearing aid acoustic feedback suppression method, system, computer device, storage medium and hearing aid.

[0007] In a first aspect, the present invention provides a method for suppressing acoustic feedback in a hearing aid, the method comprising: Real-time acquisition of raw signals within the usage scenario of the hearing aid; The operating mode of the hearing aid is obtained, including a low-power mode and at least one normal mode; If the operating mode is normal mode, then: A sliding mode observer is used to detect changes in the acoustic feedback path parameters of the original signal. If a sudden change occurs, a cancellation signal is acquired. The output signal of the hearing aid is compensated according to the cancellation signal; If the change in the acoustic feedback path parameters exceeds a preset threshold, it is confirmed that a sudden change has occurred in the acoustic feedback path parameters. Optionally, if the sliding mode observer is a first-order sliding mode observer, then the sliding mode observer is used to detect abrupt changes in the acoustic feedback path parameters of the original signal. If abrupt changes are detected, a cancellation signal is obtained, including: The first acoustic feedback path parameters of the original signal are obtained by using a first-order acoustic feedback path model and a first-order sliding mode observer iterative formula. Detect whether the parameters of the first acoustic feedback path undergo a sudden change; If a sudden change in the parameters of the first acoustic feedback path is detected, a parameter reset signal is generated; After receiving the parameter reset signal, a fast parameter estimate is obtained according to the first-order sliding mode observer iterative formula. The fast parameter estimate includes the feedback path gain estimate and the feedback path delay estimate. The variable step-size normalized least mean square algorithm is used to quickly estimate the reset filter weights through the parameters and obtain the first cancellation signal.

[0008] Optionally, if the sliding mode observer is a second-order sliding mode observer, then the sliding mode observer is used to detect abrupt changes in the acoustic feedback path parameters of the original signal. If abrupt changes are detected, a cancellation signal is obtained, including: The second acoustic feedback path parameters of the original signal are obtained by using a second-order acoustic feedback path model and a second-order sliding mode observer iterative formula. Detect whether the parameters of the second acoustic feedback path undergo a sudden change; If a sudden change is detected in the parameters of the second acoustic feedback path, a sudden change trigger signal and a nonlinear trigger signal are generated. After receiving the mutation trigger signal and the nonlinear trigger signal, the auxiliary observation variable and the equivalent control signal output by the observer are updated according to the second-order sliding mode observer iterative formula, and the nonlinear distortion coefficient is obtained. The first and second control gains are dynamically adjusted according to the second-order sliding mode algorithm formula and the nonlinear distortion coefficient. The original cancellation signal is obtained based on the adjusted first control gain and the adjusted second control gain; An adaptive phase detection algorithm is used to calibrate the phase and amplitude of the original cancellation signal in real time to obtain the first signal; After receiving the mutation trigger signal and the nonlinear trigger signal, the variable step size normalized least mean square algorithm is used to obtain the second signal; The first signal and the second signal are fused together to obtain a second cancellation signal.

[0009] Optionally, the original signal includes a feedback leakage signal. An adaptive phase detection algorithm is used to calibrate the phase and amplitude of the original cancellation signal in real time to obtain the first signal, including: The phase difference between the original cancellation signal and the feedback leakage signal is detected in real time. The original cancellation signal is corrected by delay compensation and replication calibration. The corrected original cancellation signal is used as the first signal. The first signal and the feedback leakage signal satisfy the cancellation condition, which is that the amplitudes are equal and the phases are opposite.

[0010] Optionally, if the operating mode is a low-power mode and the sliding mode observer is a first-order sliding mode observer, then: A time-sharing scheduling and computational cost reduction strategy is adopted to modify the first-order sliding mode observer into a fixed-point model, replacing floating-point operations; If the operating mode is a low-power mode and the sliding mode observer is a second-order sliding mode observer, then: By employing time-sharing scheduling and computational cost reduction strategies, the nonlinear operations in the second-order sliding mode observer and the second-order sliding mode algorithm formula are transformed into fixed-point operations, replacing floating-point operations.

[0011] Optionally, the method further includes: The system switches between low-power and normal modes based on the number of times the feedback error exceeds the error threshold consecutively. Or it can switch between low power and normal modes according to user instructions.

[0012] In a second aspect, a hearing aid acoustic feedback suppression system is provided, employing the method described in any of the preceding claims, the system comprising: The signal acquisition module is used to acquire raw signals in the usage scenario of the hearing aid in real time; A low-power computing module is used to obtain the operating mode of the hearing aid, the operating mode including a low-power mode and at least one normal mode; If the operating mode is normal mode, then: The sliding mode observation module is used to detect abrupt changes in the acoustic feedback path parameters of the original signal using a sliding mode observer. If abrupt changes are detected, a cancellation signal is acquired. The signal output module compensates the output signal of the hearing aid according to the cancellation signal; If the change in the acoustic feedback path parameters exceeds a preset threshold, it is confirmed that a sudden change has occurred in the acoustic feedback path parameters.

[0013] Optionally, the system further includes a mode management unit, used to switch between low power and normal modes based on the number of times the feedback error exceeds the error threshold consecutively, or to switch between low power and normal modes based on user instructions.

[0014] Thirdly, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the method as described in any of the preceding claims.

[0015] Fourthly, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the method as described in any of the preceding claims.

[0016] Fifthly, a hearing aid is provided that applies the method described in any of the preceding methods.

[0017] This invention provides a method, system, computer device, storage medium, and hearing aid for suppressing acoustic feedback. The method includes: real-time acquisition of raw signals within the usage scenario of the hearing aid; obtaining the operating mode of the hearing aid, including a low-power mode and at least one normal mode; if the operating mode is normal mode, then: using a sliding mode observer to detect changes in the acoustic feedback path parameters of the raw signal; if a sudden change occurs, acquiring a cancellation signal and compensating the output signal of the hearing aid based on the cancellation signal; wherein, if the change in the acoustic feedback path parameters exceeds a preset threshold, a sudden change in the acoustic feedback path parameters is confirmed. The method of this invention, by real-time acquisition of raw signals within the usage scenario and employing different methods for suppressing acoustic feedback according to different operating modes, can save energy and meet different user needs. Furthermore, in this embodiment of the invention, when the working mode is normal mode, a sliding mode observer is used to detect abrupt changes in the acoustic feedback path parameters of the original signal. If abrupt changes are detected, a cancellation signal is obtained, and the output signal of the hearing aid is compensated according to the cancellation signal. That is, the method of this embodiment of the invention can obtain a cancellation signal and compensate the output signal when abrupt changes occur in the acoustic feedback path parameters, thereby achieving rapid suppression of the whistling of the hearing aid. Attached Figure Description

[0018] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 The diagram shows a flowchart of the hearing aid acoustic feedback suppression method according to an embodiment of the present invention. Figure 2 The diagram shown is a structural block diagram of the acoustic feedback suppression system for a hearing aid according to an embodiment of the present invention. Figure 3 The diagram shows a flowchart of the hearing aid acoustic feedback suppression method according to an embodiment of the present invention. Figure 4 The diagram shows a flowchart of the hearing aid acoustic feedback suppression method according to an embodiment of the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of 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.

[0022] like Figure 1 As shown, in one embodiment, a flowchart of a hearing aid acoustic feedback suppression method is provided, the method comprising: Step 110: Real-time acquisition of raw signals within the usage scenario of the hearing aid; Different hearing aids may have slightly different usage scenarios, with common types including behind-the-ear, in-the-ear, and bone conduction hearing aids. In this embodiment of the invention, raw signals within the usage scenario are collected in real time while the hearing aid is being worn.

[0023] Step 120: Obtain the operating mode of the hearing aid, the operating mode including a low power mode and at least one normal mode; Hearing aids have two operating modes: a low-power mode for low-frequency use, applications where sound quality requirements are not high, or when the hearing aid's battery is low; and a normal mode for high-frequency use or applications where sound quality requirements are high. The normal mode can also have different modes depending on the application scenario, such as a high-quality normal mode or a mid-quality normal mode.

[0024] The operating mode of a hearing aid can be set by the user or automatically switched by the hearing aid based on the original signal, ambient noise, etc. in the usage scenario.

[0025] Step 130: If the working mode is normal mode, then: A sliding mode observer is used to detect abrupt changes in the acoustic feedback path parameters of the original signal. If abrupt changes are detected, a cancellation signal is acquired. Step 140: Compensate the output signal of the hearing aid according to the cancellation signal.

[0026] In this embodiment of the invention, if no mutation is detected, an adaptive algorithm is used to obtain a cancellation signal, and the output signal is compensated based on the cancellation signal.

[0027] In this embodiment of the invention, if the change in the acoustic feedback path parameters exceeds a preset threshold, it is confirmed that a sudden change has occurred in the acoustic feedback path parameters.

[0028] The method of this invention acquires raw signals in real time within the usage scenario and employs different methods for hearing aid acoustic feedback suppression based on different operating modes, thus saving energy and meeting diverse user needs. Furthermore, in the normal operating mode, a sliding mode observer is used to detect abrupt changes in the acoustic feedback path parameters of the raw signal. If a change is detected, a cancellation signal is acquired, and the output signal of the hearing aid is compensated based on the cancellation signal. In other words, the method of this invention can acquire a cancellation signal and compensate the output signal when abrupt changes occur in the acoustic feedback path parameters, thereby achieving rapid suppression of hearing aid whistling. The method of this invention can rapidly suppress whistling at its nascent stage and can handle nonlinear distortion.

[0029] In this embodiment of the invention, a sliding mode observer and a sliding mode controller may be included. The observer is responsible for estimating feedback path parameters, such as gain, delay, and nonlinear coefficients. The controller applies the Super-twisting algorithm and is responsible for generating the cancellation signal. The sliding mode observer and the sliding mode controller work together.

[0030] In this embodiment of the invention, the sliding mode observer can be a first-order sliding mode observer or a second-order sliding mode observer.

[0031] If the sliding mode observer is a first-order sliding mode observer, in step 130, the sliding mode observer is used to detect abrupt changes in the acoustic feedback path parameters of the original signal. If abrupt changes are detected, a cancellation signal is obtained, including: The first acoustic feedback path parameters of the original signal are obtained by using a first-order acoustic feedback path model and a first-order sliding mode observer iterative formula. Detect whether the parameters of the first acoustic feedback path undergo a sudden change; If a sudden change in the parameters of the first acoustic feedback path is detected, a parameter reset signal is generated; After receiving the parameter reset signal, a fast parameter estimate is obtained according to the first-order sliding mode observer iterative formula. The fast parameter estimate includes the feedback path gain estimate and the feedback path delay estimate. The variable step-size normalized least mean square algorithm is used to quickly estimate the reset filter weights through the parameters and obtain the first cancellation signal.

[0032] In this embodiment of the invention, the first-order acoustic feedback path model is: in, For the first Original signal, For feedback path gain parameters, For the first Output signal, For feedback path delay, For the first Environmental noise; The iterative formula for the first-order sliding mode observer is: in, For the first First acoustic feedback path parameters For the first Feedback path gain estimate For the first Output signal at sampling time; The sampling time is the number of iterations. For the first Feedback path delay estimate, representing the delay. Each sampling time; For the first First feedback error, This is an estimate of the system state. To switch gain, For the sign function on the sliding surface, For the first Feedback path gain estimate For learning rate, For the first Feedback path delay estimate For the error threshold, This is an overflow limit parameter; Among them, if This confirms that a sudden change has occurred in the parameters of the first acoustic feedback path. The first preset threshold value ranges from 0.01 to 0.2 times the root mean square value of the signal (normalized amplitude), preferably 0.05. Learning rate. The value range is 0.01~0.5, preferably 0.1~0.3, and even more preferably 0.1, which can ensure a balance between convergence speed and stability.

[0033] It is an estimated value of the system state, used to obtain the envelope information of the audio signal in real time. It has the characteristics of strong anti-interference ability, fast response speed and no delay, and provides an accurate amplitude reference for the subsequent gain control module.

[0034] The acoustic feedback path model is a mathematical or physical representation of a closed-loop system formed by sound emitted from a loudspeaker, propagating through space, and being picked up again by a microphone. This model is widely used in devices such as sound reinforcement systems, hearing aids, and smart speakers to analyze and suppress feedback problems.

[0035] The first-order acoustic feedback path model in this embodiment of the invention simplifies the parameters by omitting nonlinear distortion terms, reducing computational load and improving processing speed. Furthermore, the range of parameters is limited; for example, the feedback path gain estimate G is a fixed-point value within a finite range, simplifying the model's computational dimensions. Therefore, by omitting redundant terms, the first-order acoustic feedback path model in this embodiment of the invention can significantly reduce computational load, making it suitable for the processing capabilities of low-frequency DSP chips in mid-to-low-end hearing aids. Since the computation process does not require floating-point operations, it can directly match a fixed-point computational architecture.

[0036] In the iterative formula of the first-order sliding mode observer in this embodiment of the invention, an error formula is designed based on the first-order acoustic feedback model of this embodiment. The state matrix and observation matrix of the traditional formula are eliminated, and only the estimated values ​​of gain and delay are retained. , The error calculation is simplified to one multiplication and one subtraction; and the gain update formula adds a constraint term. This prevents parameter overflow. Furthermore, the delay estimate summarized by this formula uses a discretized ±1 step adjustment, which drastically simplifies the error calculation steps compared to continuous integration, allowing the hearing aid's DSP chip to complete the calculation in a single cycle, meeting the rapid detection requirements of intermittent sampling mode; parameter constraints ensure algorithm stability and prevent divergence.

[0037] The method of this invention has no floating-point operations, no redundant parameters, and the DSP completes the core operation in a single cycle, making it compatible with conventional chips with a main frequency of ≥8MHz and achieving lightweight design. In one embodiment of this invention, the intermittent sampling mode (operating once every 100μs) and the standby time increase the system power consumption by only 5% compared to traditional NLMS, with power consumption disclosed. The method of this invention only serves as a "mutation detector + parameter provider" and does not change the core logic of traditional NLMS, making it fully compatible in conventional scenarios. No additional hardware is required; deployment is only through firmware upgrades, and low-to-mid-range hearing aids can be directly adapted, resulting in low industrialization costs.

[0038] In this embodiment of the invention, if the sliding mode observer is a second-order sliding mode observer, step 130 involves using the sliding mode observer to detect abrupt changes in the acoustic feedback path parameters of the original signal. If abrupt changes are detected, a cancellation signal is obtained, including: The second acoustic feedback path parameters of the original signal are obtained by using a second-order acoustic feedback path model and a second-order sliding mode observer iterative formula. Detect whether the parameters of the second acoustic feedback path undergo a sudden change; If a sudden change is detected in the parameters of the second acoustic feedback path, a sudden change trigger signal and a nonlinear trigger signal are generated. After receiving the mutation trigger signal and the nonlinear trigger signal, the auxiliary observation variable and the equivalent control signal output by the observer are updated according to the second-order sliding mode observer iterative formula, and the nonlinear distortion coefficient is obtained. The first and second control gains are dynamically adjusted according to the second-order sliding mode algorithm formula and the nonlinear distortion coefficient. The original cancellation signal is obtained based on the adjusted first control gain and the adjusted second control gain; An adaptive phase detection algorithm is used to calibrate the phase and amplitude of the original cancellation signal in real time to obtain the first signal; After receiving the mutation trigger signal and the nonlinear trigger signal, the variable step size normalized least mean square algorithm is used to obtain the second signal; The first signal and the second signal are fused together to obtain a second cancellation signal.

[0039] In this embodiment of the invention, the original signal includes a feedback leakage signal. An adaptive phase detection algorithm is used to calibrate the phase and amplitude of the original cancellation signal in real time to obtain the first signal, including: The phase difference between the original cancellation signal and the feedback leakage signal is detected in real time. The original cancellation signal is corrected by delay compensation and replication calibration. The corrected original cancellation signal is used as the first signal. The first signal and the feedback leakage signal satisfy the cancellation condition, which is that the amplitudes are equal and the phases are opposite.

[0040] In this embodiment of the invention, the second-order acoustic feedback path model is as follows: in, For the first Original signal, For feedback path gain parameters, For the first Output signal, For nonlinear distortion coefficients, For feedback path delay, For the first Environmental noise.

[0041] In the second-order acoustic feedback path model of this invention, The newly added nonlinear term is the second harmonic distortion of the speaker / microphone, which is the main type of nonlinear distortion in high-end hearing aids.

[0042] A larger value indicates a more severe distortion. The quadratic term of the speaker output signal is used to quantify the degree of harmonic distortion. In the second-order acoustic feedback path model, the sampling rate can be increased to 32kHz and the delay range can be extended to [1,20] to cover the complex acoustic path of high-end hearing aids.

[0043] The second-order acoustic feedback path model of this invention covers all scenarios of linear feedback and nonlinear distortion, and can provide a quantitative basis for nonlinear distortion for subsequent compensation, so as to achieve accurate compensation.

[0044] In this embodiment of the invention, the iterative formula for the second-order sliding mode observer is: in, For the first Second acoustic feedback path parameters For the first Gain estimate; Let be a state variable, representing the th Output signal at sampling time; The sampling time is the number of iterations. For the first Delay estimate, For the first Nonlinear distortion coefficient For the first Feedback error, For the first Auxiliary observation variables, For the first Auxiliary observation variables, For the first observation gain, For the sign function on the sliding surface, The sampling period is The first output of the observer Equivalent control signal, For the second observation gain, As the first learning rate, As the second learning rate, For the first Feedback path delay estimate For the first Feedback path delay estimate For delay gain, In order to make The index of the largest sampling time. This is a direction restriction symbol used to limit the range of delay; Among them, if This confirms that a sudden change has occurred in the parameters of the second acoustic feedback path. The second preset threshold value ranges from 0.02 to 0.3, with 0.08 being preferred. Learning rate. , Related to the input signal power normalization, typically 0.05~0.2. First observation gain. Value range: 0.5~5; Second observation gain The value range is 0.1 to 3; The operation is executed every 5~10ms to reduce the computational load.

[0045] In this embodiment of the invention, It is actually the original output signal of the hearing aid, that is, a clean target audio signal that has not been captured by the microphone and has not experienced acoustic feedback; For a moment The observer output after smoothing filtering, weighting, amplitude limiting, and delay synchronization; Used to find the location where the feedback is most obvious; This is a direction restriction symbol used to prevent delays from exceeding an unreasonable range. and The initial values ​​of all are set to 0 to ensure that the algorithm starts from a zero state.

[0046] In this embodiment of the invention, the second-order sliding mode observer outputs... Simultaneously used for gain and distortion coefficient Updates to achieve joint estimation of multiple parameters; delays Update introduced Enhance sensitivity to delayed changes; add constraints. To prevent delay estimation from exceeding a preset range, the above method in this embodiment of the invention enables jitter-free multi-parameter observation and simultaneous estimation of gain, delay, and nonlinear distortion coefficients; the delay estimation error is ≤1 sampling point, providing an accurate input signal for subsequent compensation.

[0047] In this embodiment of the invention, the formula for the second-order sliding mode algorithm is: in, For the first First control gain, For the first Second control gain, For the first base gain, For the second base gain, This is the first gain adjustment factor. This is the second gain adjustment factor. For the first Nonlinear distortion coefficient For the first Auxiliary variables, For the first Auxiliary variables, For the first Feedback error, For the sign function on the sliding surface, The sampling period is To indicate the first Second-order sliding mode control cancellation signal at the sampling time. For the second control gain, For the first Original cancellation signal.

[0048] Summary of embodiments of the present invention The term is the square root of the absolute value of the error. As the core convergence term of the second-order sliding mode in Super-twisting, it enables the control law to converge quickly when the error is large and to move smoothly when the error is small. It fundamentally eliminates the high-frequency jitter phenomenon of the traditional first-order sliding mode and ensures that the hearing aid outputs noise-free and high-quality audio. For the first The original cancellation signal, generated by the Super-twisting algorithm, is used to cancel the acoustic feedback signal and nonlinear distortion of the hearing aid in real time, achieving fast, jitter-free, and high-precision howling suppression.

[0049] In this embodiment of the invention, the second-order sliding mode algorithm formula uses a fixed control gain. / Change to distortion coefficient Driven dynamic gain , This achieves adaptive control where "the more severe the distortion, the higher the gain, and the stronger the cancellation effect," resulting in an over 80% improvement in suppression rate in nonlinear scenarios. Furthermore, the amplitude limiting function... The output signal is limited to [-0.8, 0.8] by dynamic gain to achieve adaptive compensation for nonlinear distortion; amplitude limitation avoids excessive signal amplitude that could cause secondary distortion in the speaker / microphone, thus ensuring high-quality sound.

[0050] In this embodiment of the invention, when a second-order sliding mode observer is used, the nonlinear distortion coefficient of the receiver of the second-order sliding mode observer is... To achieve control gain Dynamic adjustment is used to specifically handle nonlinear distortion feedback at high volumes, generating a continuous, jitter-free, and adaptive cancellation signal, which, together with the phase compensation module, achieves precise acoustic cancellation.

[0051] In this embodiment of the invention, the above method can achieve ultra-fast suppression, with a howling suppression response time ≤1ms, completely eliminating howling in its incipient stage; based on the distortion coefficient Dynamically adjust the gain to improve the suppression rate by more than 80% in nonlinear scenarios and achieve adaptive control; the output signal is continuous and smooth with no high-frequency harsh noise, meeting the requirements of high-quality sound; the dual-channel fusion method is used for better results.

[0052] In this embodiment of the invention, if the operating mode is a low-power mode and the sliding mode observer is a first-order sliding mode observer, then: A time-sharing scheduling and computational cost reduction strategy is adopted to modify the first-order sliding mode observer into a fixed-point model, replacing floating-point operations; If the operating mode is a low-power mode and the sliding mode observer is a second-order sliding mode observer, then: By employing time-sharing scheduling and computational cost reduction strategies, the nonlinear operations in the second-order sliding mode observer and the second-order sliding mode algorithm formula are transformed into fixed-point operations, replacing floating-point operations.

[0053] The fixed-point arithmetic uses 16-bit or 24-bit fixed-point arithmetic, with a time-sharing scheduling period of 10ms. Each time the observer is woken up, it performs 100 sampling point iterations and then immediately goes into sleep mode. When the observer detects... If the threshold is exceeded three times consecutively, the system automatically switches to normal mode. In low-power mode, the calculation of second-order nonlinear terms can be turned off, and only first-order observations are retained to save energy; the switching threshold can be configured to 2 to 5 times.

[0054] In one embodiment of the present invention, the second-order sliding mode observer and the second-order sliding mode algorithm are driven by a low-power coprocessor to form a closed loop of "observation-computation-control-calibration", achieving seamless linkage with the normalized NLMS module. The core logic is as follows: In a typical linear scenario: the second-order sliding mode observer operates with low power, the second-order sliding mode algorithm module is in standby mode, the normalized NLMS module works independently, and the coprocessor only receives states and does not perform calculations. Path mutation scenario: Observer 50 Internally generated mutation trigger signal, control module 1 Internally generated jitter-free signal, NLMS module parameters are quickly reset, dual-channel cancellation signal fusion output, and howling is immediately eliminated; Highly nonlinear scenarios: Observer estimates distortion coefficients The control module dynamically increases the gain and generates an adaptive cancellation signal. After being precisely calibrated by the phase compensation module, it is fused with the NLMS signal, resulting in an improvement of nonlinear suppression rate of over 80%. After scene restoration: The observer detected and The control module immediately goes into standby mode, the system returns to NLMS independent working mode, the coprocessor resumes low power consumption, and the main DSP always maintains low power operation.

[0055] In this embodiment of the invention, the method further includes: The system switches between low-power and normal modes based on the number of times the feedback error exceeds the error threshold consecutively. Or it can switch between low power and normal modes according to user instructions.

[0056] Figure 2 The diagram shown is a schematic representation of a hearing aid acoustic feedback suppression system according to an embodiment of the present invention. The system applies the method described in any of the preceding claims, and the system includes: Signal acquisition module 210 is used to acquire raw signals in the usage scenario of the hearing aid in real time; The low-power computing module 220 is used to obtain the operating mode of the hearing aid, the operating mode including a low-power mode and at least one normal mode; If the operating mode is normal mode, then: The sliding mode observation module 230 is used to detect abrupt changes in the acoustic feedback path parameters of the original signal using a sliding mode observer. If abrupt changes are detected, a cancellation signal is acquired. The signal output module 240 compensates the output signal of the hearing aid according to the cancellation signal; If the change in the acoustic feedback path parameters exceeds a preset threshold, it is confirmed that a sudden change has occurred in the acoustic feedback path parameters.

[0057] In this embodiment of the invention, the system further includes a mode management unit, which is used to switch between low power and normal modes according to the number of times the feedback error exceeds the error threshold consecutively, or to switch between low power and normal modes according to user instructions.

[0058] Figure 3 The diagram shown is a flowchart of a hearing aid acoustic feedback suppression method according to an embodiment of the present invention. Figure 3 As shown, the method includes: Step 310: The signal acquisition module acquires the raw signals within the usage scenario of the hearing aid in real time; Step 320: The low-power computing module obtains the operating mode of the hearing aid; if it is in low-power mode, proceed to step 350; if it is in normal mode, proceed to step 330.

[0059] Step 330: The first-order sliding mode observation module detects abrupt changes in the acoustic feedback path parameters of the original signal; Step 340: The first-order sliding mode observation module obtains rapid parameter estimates; Step 350: The adaptive filtering module acquires the first cancellation signal; Step 360: The signal output module compensates the output signal of the hearing aid according to the first cancellation signal.

[0060] Figure 4 The diagram shown is a flowchart of a hearing aid acoustic feedback suppression method according to an embodiment of the present invention. Figure 4 As shown, the method includes: Step 410: The signal acquisition module acquires the raw signals in the usage scenario of the hearing aid in real time; Step 420: The low-power computing module obtains the operating mode of the hearing aid; if it is in low-power mode, proceed to step 480; if it is in normal mode, proceed to step 430.

[0061] Step 430: The second-order sliding mode observation module detects abrupt changes in the acoustic feedback path parameters of the original signal, and then proceeds to steps 440 and 450.

[0062] Step 440: The second-order sliding mode algorithm module obtains the second signal and proceeds to step 460.

[0063] Step 450: The adaptive filtering module acquires the first signal.

[0064] Step 460: Combine the first signal with the second signal to obtain the second cancellation signal.

[0065] Step 470: The signal output module compensates the output signal of the hearing aid according to the second cancellation signal.

[0066] Step 480: The first signal is used as the second cancellation signal, and the signal output module compensates the output signal of the hearing aid according to the second cancellation signal.

[0067] This invention also provides a hearing aid that uses the method described in any of the preceding embodiments.

[0068] A computer program on memory and executable on a processor, wherein the processor executes the computer program to implement the following method: real-time acquisition of raw signals within the usage scenario of the hearing aid; acquisition of the operating mode of the hearing aid, the operating mode including a low-power mode and at least one normal mode; if the operating mode is a normal mode, then: using a sliding mode observer to detect changes in the acoustic feedback path parameters of the raw signal; if a sudden change occurs, acquiring a cancellation signal; and compensating the output signal of the hearing aid based on the cancellation signal; wherein, if the change in the acoustic feedback path parameters exceeds a preset threshold, a sudden change in the acoustic feedback path parameters is confirmed.

[0069] This invention also provides a computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements the following method: real-time acquisition of raw signals within the usage scenario of the hearing aid; acquisition of the operating mode of the hearing aid, the operating mode including a low-power mode and at least one normal mode; if the operating mode is a normal mode, then: using a sliding mode observer to detect changes in the acoustic feedback path parameters of the raw signal; if a sudden change occurs, acquiring a cancellation signal; and compensating the output signal of the hearing aid based on the cancellation signal; wherein, if the change in the acoustic feedback path parameters exceeds a preset threshold, it is confirmed that a sudden change has occurred in the acoustic feedback path parameters.

[0070] The above method achieves the beneficial effect of solving the technical problems raised in the background art.

[0071] Figure 1 , Figure 3 and Figure 4 This is a flowchart illustrating a hearing aid acoustic feedback suppression method in one embodiment. It should be understood that, although... Figure 1 , Figure 3 and Figure 4 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 , Figure 3 and Figure 4 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0072] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided by this invention can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0073] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0074] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A method for suppressing acoustic feedback in a hearing aid, characterized in that, The method includes: Real-time acquisition of raw signals within the usage scenario of the hearing aid; The operating mode of the hearing aid is obtained, including a low-power mode and at least one normal mode; If the operating mode is normal mode, then: The change in the acoustic feedback path parameters of the original signal is detected by a sliding mode observer. If the change in the acoustic feedback path parameters exceeds a preset threshold, it is confirmed that the acoustic feedback path parameters have undergone a sudden change. Upon detecting a sudden change, a corresponding cancellation signal is generated based on the sliding mode observer type: If the sliding mode observer is a first-order sliding mode observer, then parameter reset and fast path parameter estimation are performed, and the filter weights are adaptively updated based on the estimation results to generate the first cancellation signal; If the sliding mode observer is a second-order sliding mode observer, a trigger signal is generated and the nonlinear distortion coefficient is obtained. The control parameters are adjusted based on the nonlinear distortion coefficient to generate the original cancellation signal. Phase compensation and adaptive filtering are performed respectively to obtain the first signal and the second signal. The first signal and the second signal are combined to generate the second cancellation signal. The output signal of the hearing aid is compensated according to the cancellation signal; If the operating mode is a low-power mode and the sliding mode observer is a first-order sliding mode observer, then: A time-sharing scheduling and computational cost reduction strategy is adopted to modify the first-order sliding mode observer into a fixed-point model, replacing floating-point operations; If the operating mode is a low-power mode and the sliding mode observer is a second-order sliding mode observer, then: By employing time-sharing scheduling and computational cost reduction strategies, the nonlinear operations in the second-order sliding mode observer and the second-order sliding mode algorithm formula are transformed into fixed-point operations, replacing floating-point operations.

2. The method according to claim 1, characterized in that, The sliding mode observer is a first-order sliding mode observer; the step of using the sliding mode observer to detect changes in the acoustic feedback path parameters of the original signal includes: The first acoustic feedback path parameters of the original signal are obtained by using a first-order acoustic feedback path model and a first-order sliding mode observer iterative formula. Detect whether the parameters of the first acoustic feedback path undergo a sudden change; The process of resetting execution parameters and rapidly estimating path parameters, and adaptively updating filter weights based on the estimation results to generate a first cancellation signal, includes: If a sudden change in the parameters of the first acoustic feedback path is detected, a parameter reset signal is generated; After receiving the parameter reset signal, a fast parameter estimate is obtained according to the first-order sliding mode observer iterative formula. The fast parameter estimate includes the feedback path gain estimate and the feedback path delay estimate. The variable step-size normalized least mean square algorithm is used to quickly estimate the reset filter weights through the parameters and obtain the first cancellation signal.

3. The method according to claim 1, characterized in that, The sliding mode observer is a second-order sliding mode observer; the step of using the sliding mode observer to detect changes in the acoustic feedback path parameters of the original signal includes: The second acoustic feedback path parameters of the original signal are obtained by using a second-order acoustic feedback path model and a second-order sliding mode observer iterative formula. Detect whether the parameters of the second acoustic feedback path undergo a sudden change; The process of generating a trigger signal and obtaining nonlinear distortion coefficients, adjusting control parameters based on the nonlinear distortion coefficients to generate an original cancellation signal, performing phase compensation and adaptive filtering to obtain a first signal and a second signal, and merging the first signal and the second signal to generate a second cancellation signal includes: If a sudden change is detected in the parameters of the second acoustic feedback path, a sudden change trigger signal and a nonlinear trigger signal are generated. After receiving the mutation trigger signal and the nonlinear trigger signal, the auxiliary observation variable and the equivalent control signal output by the observer are updated according to the second-order sliding mode observer iterative formula, and the nonlinear distortion coefficient is obtained. The first and second control gains are dynamically adjusted according to the second-order sliding mode algorithm formula and the nonlinear distortion coefficient. The original cancellation signal is obtained based on the adjusted first control gain and the adjusted second control gain; An adaptive phase detection algorithm is used to calibrate the phase and amplitude of the original cancellation signal in real time to obtain the first signal; After receiving the mutation trigger signal and the nonlinear trigger signal, the variable step size normalized least mean square algorithm is used to obtain the second signal; The first signal and the second signal are fused together to obtain a second cancellation signal.

4. The method according to claim 1, characterized in that, The original signal includes a feedback leakage signal. An adaptive phase detection algorithm is used to calibrate the phase and amplitude of the original cancellation signal in real time to obtain the first signal, including: The phase difference between the original cancellation signal and the feedback leakage signal is detected in real time. The original cancellation signal is corrected by delay compensation and amplitude calibration. The corrected original cancellation signal is used as the first signal. The first signal and the feedback leakage signal satisfy the cancellation condition, which is that the amplitudes are equal and the phases are opposite.

5. The method according to claim 1, characterized in that, The method further includes: The system switches between low-power and normal modes based on the number of times the feedback error exceeds the error threshold consecutively. Or it can switch between low power and normal modes according to user instructions.

6. A hearing aid acoustic feedback suppression system, using the method as described in any one of claims 1 to 5, characterized in that, The system includes: The signal acquisition module is used to acquire raw signals in the usage scenario of the hearing aid in real time; A low-power computing module is used to obtain the operating mode of the hearing aid, the operating mode including a low-power mode and at least one normal mode; If the operating mode is normal mode, then: The sliding mode observation module is used to detect abrupt changes in the acoustic feedback path parameters of the original signal using a sliding mode observer. If the change in the acoustic feedback path parameters exceeds a preset threshold, it is confirmed that abrupt changes have occurred in the acoustic feedback path parameters. Upon detecting a sudden change, a corresponding cancellation signal is generated based on the sliding mode observer type: If the sliding mode observer is a first-order sliding mode observer, then parameter reset and fast path parameter estimation are performed, and the filter weights are adaptively updated based on the estimation results to generate the first cancellation signal; If the sliding mode observer is a second-order sliding mode observer, a trigger signal is generated and the nonlinear distortion coefficient is obtained. The control parameters are adjusted based on the nonlinear distortion coefficient to generate the original cancellation signal. Phase compensation and adaptive filtering are performed respectively to obtain the first signal and the second signal. The first signal and the second signal are combined to generate the second cancellation signal. The signal output module compensates the output signal of the hearing aid according to the cancellation signal.

7. The system according to claim 6, characterized in that, The system also includes a mode management unit, which is used to switch between low power and normal modes based on the number of times the feedback error exceeds the error threshold consecutively, or to switch between low power and normal modes based on user instructions.

8. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 5.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 5.

10. A hearing aid, characterized in that, The method described in any one of claims 1 to 5 is applied.

Citation Information

Patent Citations

  • Hearing-aid with feedback cancellation

    CN105392099A

  • Hearing aid internal disturbance suppression method and device

    CN114598981A