Headset adaptive hybrid active noise control method

Abstract

The application discloses a headset adaptive hybrid active noise reduction control method and belongs to the technical field of audio equipment and active noise reduction. The application synchronously collects external environmental noise and residual noise signals in ear covers, adopts a multi-feature fusion technology to realize rapid identification of transient noise, constructs a feedforward-feedback hybrid noise reduction model based on a filter least mean square algorithm, adaptively adjusts filter response parameters and feedforward and feedback noise reduction weight distribution according to noise types, generates an inverted cancellation signal to realize active noise reduction. The application also integrates frequency domain and time domain optimization in a transparent mode, double-microphone call noise reduction and noise reduction effect adaptive calibration functions, effectively solves the problems of insufficient transient noise suppression and distorted hearing in a transparent mode in the prior art, significantly improves noise reduction effect and user experience in different complex scenes, and can be widely applied to various types of Bluetooth headset products.

Description

Technical Field

[0001] This invention relates to the field of audio equipment and active noise cancellation technology, specifically to an adaptive hybrid active noise cancellation control method for headphones, which is particularly suitable for Bluetooth headphones that need to provide a high-quality audio experience in complex noise environments. Background Technology

[0002] With the rapid development of mobile internet and consumer electronics technology, over-ear Bluetooth headphones have become essential electronic products for daily commutes, work, study, and leisure. Active noise cancellation technology, as a key technology for enhancing the core competitiveness of headphones, uses electronic means to generate sound waves with the opposite phase to ambient noise, canceling out noise entering the ear and creating a quiet listening environment for users. Currently, mainstream active noise cancellation technologies are mainly divided into three architectures: feedforward noise cancellation, feedback noise cancellation, and hybrid noise cancellation combining both. Among them, the hybrid noise cancellation architecture combines the wide-band advantage of feedforward noise cancellation with the high-precision advantage of feedback noise cancellation, and has become a standard technology in high-end over-ear headphones.

[0003] The existing technology has the following problems: 1. Existing hybrid active noise cancellation technology has a lag in response to transient noise and limited suppression capabilities, making it unable to effectively filter out common sudden high-frequency noise in daily life, thus affecting the user's immersive listening experience; 2. Existing signal processing algorithms for transparent mode lack fine frequency and time domain correction, resulting in low environmental sound reproduction, timbre distortion, and time misalignment, which reduces the user's ability to perceive the environment. Summary of the Invention

[0004] This invention provides an adaptive hybrid active noise cancellation control method for headphones to solve the problems mentioned in the background art.

[0005] To address the aforementioned technical problems, the technical solution adopted in this invention is as follows: A hybrid active noise cancellation system capable of adapting to environmental changes is constructed. This system achieves rapid response to sudden noise through multi-dimensional transient noise detection technology, optimizes noise cancellation effects under different noise types through a dynamic weight allocation mechanism, enhances the listening quality of the transparency mode through refined frequency and time domain processing, and integrates adaptive calibration and call noise cancellation functions to comprehensively improve the overall performance and user experience of the headphones.

[0006] Specifically, the adaptive hybrid active noise cancellation control method for headphones includes the following steps: The first step is noise signal acquisition: external ambient noise signals are acquired through a feedforward microphone located on the outside of the earphone shell, and residual noise signals inside the earcups are acquired through a feedback microphone located inside the earcups near the ear. The two microphone signals are sampled synchronously, and after anti-aliasing filtering and analog-to-digital conversion, they are sent to a digital signal processor for processing.

[0007] The second step is the rapid transient noise identification process: the external noise signal collected by the feedforward microphone is processed into frames, and a Hamming window is used to reduce spectral leakage. For each frame, three characteristic parameters are calculated: short-time energy, spectral entropy, and zero-crossing rate. Short-time energy reflects the intensity change of the signal, spectral entropy reflects the spectral distribution complexity of the signal, and zero-crossing rate reflects the frequency change of the signal. By setting a multi-feature joint judgment threshold, when the short-time energy of a frame exceeds the energy threshold, the spectral entropy is lower than the entropy threshold, and the zero-crossing rate is higher than the zero-crossing rate threshold, the frame is determined to contain transient noise. This multi-feature fusion detection method can effectively distinguish between transient noise and steady-state noise, reducing the false detection rate and the false negative rate.

[0008] The next step is to construct the hybrid noise reduction model: Based on the Filtered Least Mean Square (FXLMS) algorithm, a feedforward-feedback hybrid noise reduction model is constructed. The feedforward filter receives the external noise signal collected by the feedforward microphone and predicts the noise that will enter the earcups. The feedback filter receives the residual noise signal collected by the feedback microphone and corrects the prediction error of the feedforward filter. The coefficients of the two filters are updated iteratively in real time according to the error signal to ensure that the noise reduction effect is always kept in the optimal state.

[0009] Then comes the adaptive parameter adjustment step: when transient noise is detected, the transient noise suppression mechanism is immediately triggered, compressing the step response time of the filter to within 3 milliseconds and increasing the noise reduction gain in the mid-to-high frequency band from 1kHz to 8kHz. An inverse cancellation signal is quickly generated to suppress transient noise. At the same time, the spectral characteristics and stability of the ambient noise are analyzed in real time, and the output weights of the feedforward filter and the feedback filter are dynamically adjusted. For low-frequency steady-state noise such as air conditioning noise and traffic noise, the weight of the feedforward filter is increased to give full play to the advantages of feedforward noise reduction in the low-frequency band. For mid-to-high frequency non-steady-state noise such as human voice and music, the weight of the feedback filter is increased to improve the suppression effect by utilizing the high-precision characteristics of feedback noise reduction.

[0010] The next step is the generation and playback of the cancellation signal: the outputs of the feedforward filter and the feedback filter are weighted and fused according to the dynamically adjusted weights to generate the final noise cancellation signal. After digital-to-analog conversion and power amplification, the signal drives the headphone driver unit to play, which cancels out the ambient noise entering the earcups, thus achieving active noise cancellation.

[0011] It also includes a transparency mode signal processing step: When the user turns on transparency mode, the system switches to the transparency mode processing flow. After the external microphone collects the ambient sound signal, it first undergoes frequency domain equalization compensation processing. The frequency response curve of the passive noise isolation of the headphones under different wearing conditions is measured in advance through experiments to generate the corresponding reverse equalization filter. The digitally collected ambient sound signal is filtered to compensate for the high frequency attenuation and low frequency gain caused by passive noise isolation. Then, time domain delay alignment processing is performed to calculate the propagation delay time of the sound wave through the ear cup foam and the shell. The digitally processed ambient sound signal is delayed by the same time so that it reaches the human ear synchronously with the original ambient sound that enters the ear cup through the physical path. Finally, dynamic range control is performed to avoid the sound pressure signal causing listening discomfort. Then, it is superimposed with the original sound field for output.

[0012] In addition, it includes a call noise reduction processing step and an adaptive calibration step: the call noise reduction processing uses a dual-microphone array to form a directional beam to enhance the voice signal from the direction of the user's mouth, while suppressing environmental noise from other directions. The adaptive calibration step runs periodically during the use of the headphones, collects residual noise signals inside the earcups, analyzes the changes in the noise reduction effect, automatically updates the filter coefficients, and compensates for the attenuation of the noise reduction effect caused by factors such as earcup foam aging and changes in wearing tightness.

[0013] Due to the adoption of the above technical solution, the technical progress achieved by this invention compared to the prior art is as follows: 1. This invention achieves millisecond-level response to sudden noise through multi-feature fusion transient noise detection technology. Combined with an adaptive filter parameter adjustment mechanism, it significantly improves the ability to suppress transient noise and effectively eliminates common sudden interferences such as door closing sounds and keyboard sounds, providing users with a more immersive listening experience.

[0014] 2. This invention adopts a feedforward-feedback hybrid noise reduction architecture with dynamic weight allocation, which can automatically optimize the noise reduction strategy according to the type and characteristics of environmental noise. It can achieve deep noise reduction in steady-state noise environment and maintain good speech intelligibility in non-steady-state noise environment, thus comprehensively improving the noise reduction effect in different scenarios.

[0015] 3. This invention introduces a refined frequency domain equalization compensation and time domain delay alignment algorithm in the transparency mode, which effectively solves the problems of environmental sound distortion and time misalignment in the prior art, restores the natural and realistic sound field effect, and enables users to accurately perceive the sound details and spatial orientation of the external environment when the transparency mode is turned on.

[0016] 4. This invention integrates a call noise reduction function based on a dual-microphone array. Through spatial beamforming and adaptive noise cancellation technology, it can effectively separate human voice from background noise in noisy environments, significantly improve the call signal-to-noise ratio, and ensure clear and smooth call quality.

[0017] 5. This invention features an adaptive calibration mechanism for noise reduction, which automatically compensates for the decrease in noise reduction effect caused by changes in earcup sealing and component aging, extending the effective service life of the headphones and ensuring performance stability during long-term use.

[0018] 6. The noise reduction control method of the present invention is completely independent of other functional modules of the headphones, and can also be used in combination with other functions such as magnetofluid visual effects and personalized sound effect adjustment, and has good compatibility and scalability. Attached Figure Description

[0019] Figure 1 This is an overall flowchart of the present invention; Figure 2 This is a detailed flowchart of the rapid transient noise identification steps of the present invention; Figure 3 This is a schematic diagram of the feedforward-feedback hybrid noise reduction model of the present invention; Figure 4 This is a flowchart of the transparent mode signal processing steps of the present invention; Figure 5 This is a schematic diagram illustrating the principle of the call noise reduction processing steps of the present invention; Figure 6 This is a flowchart of the adaptive calibration steps for noise reduction effect in this invention; Figure 7 This is a flowchart illustrating the noise reduction process of the present invention. Detailed Implementation

[0020] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0021] Implementation Plan like Figures 1-7 As shown, this implementation provides an adaptive hybrid active noise cancellation control method for headphones, applied to a typical Bluetooth headphone. The hardware system of the headphone mainly includes: two earcup units, a headband assembly, a Bluetooth communication module, a digital signal processor (DSP), an audio codec, a power amplifier, a speaker unit, a microphone array, and a power management module.

[0022] The microphone array includes two feedforward microphones, two feedback microphones, and two talk microphones. The feedforward microphones are installed on the top outer side of the two earcups to collect external ambient noise. The feedback microphones are installed inside the earcups near the sound-generating unit to collect residual noise inside the earcups. The talk microphones are installed on the lower front side of the earcups, forming a directional array to collect voice signals. The sound-generating unit uses a 40mm diameter dynamic driver with wide frequency response and low distortion characteristics. The digital signal processor uses a high-performance audio-specific DSP, which can run all the algorithms of this invention in real time.

[0023] The adaptive hybrid active noise cancellation control method for headphones is implemented on the aforementioned hardware platform as follows: When the user presses and holds the power button to turn on the headphones and enters active noise cancellation mode, the system first performs initialization operations, loading the default filter coefficients and algorithm parameters. Then, the feedforward microphone and feedback microphone begin to synchronously collect noise signals, with the sampling rate set to 48kHz and the sampling precision to 16 bits. The collected analog signals are processed by the audio codec for anti-aliasing filtering and analog-to-digital conversion, generating digital audio signals that are sent to the DSP for processing.

[0024] The DSP first performs transient noise detection on the external noise signal acquired by the feedforward microphone, dividing the signal into frames of 256 points each, with 128 points of overlap between adjacent frames. A Hamming window is used to window each frame of the signal. Then, the short-time energy, spectral entropy, and zero-crossing rate of each frame of the signal are calculated. The formula for calculating the short-time energy is: E(n)=Σ|x(n,i)|², where x(n,i) is the i-th sampling point of the n-th frame. The calculation of the spectral entropy first performs a Fast Fourier Transform (FFT) on the signal to obtain the spectrum, then calculates the probability distribution of each frequency component, and finally calculates the spectral entropy according to the definition of entropy. The zero-crossing rate is the number of times the signal crosses zero in each frame.

[0025] When three characteristic parameters of a certain frame of signal simultaneously meet the preset judgment conditions, the system determines that transient noise has been detected and immediately triggers the transient noise suppression mechanism. At this time, the DSP increases the step size factor of the filter to speed up the coefficient update speed and compress the step response time to about 2 milliseconds. At the same time, the filter gain in the 1kHz to 8kHz frequency band is increased to enhance the ability to suppress mid-to-high frequency transient noise. When multiple consecutive frames of signal no longer meet the transient noise judgment conditions, the system smoothly restores the filter parameters to the normal state to avoid listening discomfort caused by parameter abrupt changes.

[0026] While performing transient noise detection, the system runs a feedforward-feedback hybrid noise reduction algorithm. The feedforward filter receives the feedforward noise signal and generates a preliminary cancellation signal, while the feedback filter receives the feedback residual noise signal and generates an error correction signal. Both filters adopt a transverse FIR filter structure, and the order is set from 128 to 512 according to hardware performance and noise reduction requirements. The filter coefficients are updated in real time according to the FXLMS algorithm. The iteration formula is: w(n+1)=w(n)+μe(n)x'(n), where w(n) is the current filter coefficient, μ is the step size factor, e(n) is the error signal, and x'(n) is the reference signal after secondary path filtering.

[0027] The system simultaneously analyzes the spectral characteristics and stability of environmental noise in real time. It determines the stability of noise by calculating the spectral variance of multiple consecutive frames of signals. The smaller the variance, the more stable the noise. When the noise is determined to be steady-state, the output weight of the feedforward filter is set to 0.7 and the output weight of the feedback filter is set to 0.3. When the noise is determined to be non-steady-state, the feedforward weight is adjusted to 0.3 and the feedback weight is adjusted to 0.7. Through this dynamic weight allocation, the system can achieve the best noise reduction effect in different noise environments.

[0028] The outputs of the feedforward filter and the feedback filter are weighted and fused according to the above weights to generate the final noise cancellation signal. This signal is converted from digital to analog by the audio codec, and then amplified by the power amplifier to drive the sound unit to play. The cancellation signal and the ambient noise entering the earcup interfere with each other at the ear, thereby achieving active noise cancellation.

[0029] When a user activates Transparency Mode via a physical button or mobile app, the system automatically disables active noise cancellation and switches to the Transparency Mode processing flow. After the external microphone captures ambient sound signals, they are first fed into an equalization compensation filter. The coefficients of this filter are pre-measured experimentally and compensate for frequency response distortion caused by passive noise isolation from the headphones. Specifically, passive noise isolation typically causes a 10dB to 15dB attenuation of high-frequency signals above 8kHz, while simultaneously increasing low-frequency signals below 200Hz by 5dB to 10dB. The equalization compensation filter boosts the high-frequency signals accordingly and attenuates the low-frequency signals accordingly, ensuring that the processed signal's frequency response matches the original sound in a free field.

[0030] After equalization compensation, the signal enters the delay alignment module. The system pre-measures the propagation delay time of the sound wave through the ear cup foam and the shell, which is about 1.5 milliseconds. The delay alignment module delays the digital signal by 1.5 milliseconds, so that it reaches the ear at the same time as the original ambient sound that enters the ear cup through the physical path. This eliminates the time difference caused by the digital processing delay and avoids the auditory sensation of echo or ghosting.

[0031] After delay alignment, the signal passes through a dynamic range control module to limit the peak value of the signal within a safe range, avoiding damage to the user's hearing from sudden high sound pressure levels. Finally, the processed ambient sound signal is mixed with the audio signal played by the headphones and output through the sound unit. The sound heard by the user includes both the content played by the headphones and natural and realistic ambient sounds, enabling clear communication with others or perception of the surrounding environment without removing the headphones.

[0032] When a user makes a call, the system automatically activates the call noise reduction function. The two microphones form a directional beam, with the main lobe pointing towards the user's mouth and the side lobes suppressing noise from other directions. The system uses a combination of beamforming and adaptive noise cancellation algorithms. First, the voice signal is enhanced by beamforming. Then, the ambient noise collected by the feedforward microphone is used as a reference signal, and the residual background noise is canceled by an adaptive filter. The processed clean voice signal is sent to the other party via Bluetooth module, ensuring clear calls even in noisy environments.

[0033] To ensure the noise cancellation effect of the headphones during long-term use, the system also features an adaptive calibration function. The calibration process runs automatically every 30 minutes or can be manually triggered by the user. During calibration, the system plays a low-volume white noise signal and simultaneously collects the response signal inside the earcups through the feedback microphone. It analyzes the spectral characteristics of the response signal and calculates the current noise cancellation effect. If the noise cancellation effect of a certain frequency band is found to be significantly reduced, it indicates that the filter coefficients of that frequency band need to be adjusted. The system automatically retrains the filter coefficients and updates them to the optimal value to compensate for the noise cancellation effect attenuation caused by factors such as earcup foam aging and changes in wearing tightness.

[0034] This implementation scheme fully realizes the adaptive hybrid active noise cancellation control method for headphones proposed in this invention. Through the coordinated work of the above steps, it can provide users with a high-quality noise cancellation experience and transparent listening experience in various complex noise environments. This scheme makes full use of existing hardware resources without adding extra costs, and has high practical value and market competitiveness.

[0035] This implementation scheme achieves optimal suppression of different types of noise by constructing a fully adaptive hybrid active noise cancellation system, especially significantly improving the suppression effect of transient noise. At the same time, it solves the problem of hearing distortion in transparency mode. The integrated call noise reduction and adaptive calibration functions further enhance the overall performance and user experience of the product. This solution is technically mature, easy to implement, and can be directly applied to various over-ear Bluetooth headset products.

[0036] It should also be noted that, based on digital signal processing technology, a closed-loop active noise cancellation system is constructed through the collaborative work of feedforward and feedback microphones. The feedforward microphone detects external ambient noise in advance and generates an initial cancellation signal through a feedforward filter. The feedback microphone monitors residual noise inside the earcups and corrects errors in the cancellation signal through a feedback filter. At the same time, a transient noise detection module identifies sudden noise in real time and dynamically adjusts system parameters to achieve a rapid response. In transparency mode, the system restores the natural ambient sound field through fine signal processing and ensures long-term stable operation of the system through periodic adaptive calibration.

[0037] Compared with existing technologies, this invention achieves fully adaptive control of the noise reduction system. The present invention will be further described in detail below with reference to embodiments: Example

[0038] This embodiment provides the adaptive hybrid active noise cancellation control method for headphones, specifically applied in a mainstream Bluetooth headphone. The headphone uses the QCC3095 Bluetooth chip as the main controller, paired with the AP8220A1 dedicated audio DSP to run the noise cancellation algorithm, and is equipped with a 40mm neodymium magnet dynamic driver unit and a 4-microphone array (2 feedforward microphones and 2 feedback microphones). The headphone has a built-in 500mAh lithium battery, supporting 30 hours of continuous noise cancellation playback.

[0039] In this embodiment, the parameters for transient noise detection are set as follows: frame length 256 points, frame shift 128 points, Hamming window, short-time energy threshold set to 3 times the energy of conventional steady-state noise, spectral entropy threshold set to 0.3, and zero-crossing rate threshold set to 50 times / frame. When a frame signal simultaneously meets these three conditions, it is determined to be transient noise. Experimental tests show that this parameter setting can achieve a transient noise detection accuracy of over 95% and a false detection rate of less than 5%.

[0040] In the hybrid noise reduction model, both the feedforward filter and the feedback filter are 256th order FIR filters. The step size factor of the FXLMS algorithm is set to 0.01 under normal conditions. When transient noise is detected, the step size factor is increased to 0.1, enabling the filter to quickly track noise changes. The threshold for dynamic weight allocation is set as follows: when the spectral variance of 10 consecutive frames of signal is less than 0.1, it is determined to be steady-state noise, with a feedforward weight of 0.7 and a feedback weight of 0.3. When the spectral variance is greater than 0.5, it is determined to be non-steady-state noise, with a feedforward weight of 0.3 and a feedback weight of 0.7. In the transition region, the weights change linearly.

[0041] In transparency mode, the frequency response curve of the equalization compensation filter is obtained by averaging test data from 20 different wearers, which can adapt to the wearing conditions of most users. The delay time for delay alignment is set to 1.5 milliseconds, and the threshold for dynamic range control is set to 85dB. Signals exceeding this threshold will be compressed.

[0042] To verify the noise reduction effect of this embodiment, a comparative test was conducted in a standard anechoic chamber. The test results show that in the low-frequency range of 100Hz to 1000Hz, the average noise reduction depth of this embodiment reaches 38dB, which is better than the 32dB of similar products. In the mid-frequency range of 1000Hz to 4000Hz, the average noise reduction depth reaches 25dB, which is better than the 18dB of similar products. In particular, the suppression effect on transient noise is significant. This embodiment achieves a suppression of 28dB for a door closing sound with a duration of 10 milliseconds, while similar products only achieve 12dB, showing a very obvious advantage.

[0043] In the transparency mode test, 10 subjects with normal hearing were invited to conduct subjective listening evaluation. The results showed that 9 subjects thought that the transparency mode of this embodiment sounded natural and very close to the sound when not wearing headphones. They were able to clearly distinguish sounds from different directions. However, for similar products that use the traditional transparency mode, only 3 subjects thought that the sound was natural. Most subjects reported that the sound was muffled and there was a sense of delay.

[0044] The call noise reduction test was conducted in a simulated office environment with a background noise level of 60dB, consisting of air conditioning and keyboard noise. The test results showed that the call signal-to-noise ratio of this embodiment reached 32dB, and the other party could hear the content clearly. In contrast, the call signal-to-noise ratio of similar products was only 24dB, and the other party reported that the background noise was too loud and they could not hear clearly.

[0045] This embodiment fully verifies the effectiveness and superiority of the control method of the present invention. By adopting the technical solution of the present invention, the headphones have achieved industry-leading levels in terms of noise reduction depth, transient noise suppression, transparency mode listening experience and call quality. At the same time, the solution is implemented entirely on existing hardware platforms without increasing additional production costs, and has a high cost performance. Example

[0046] This embodiment provides an adaptive hybrid active noise cancellation control method for headphones, applied in a high-end professional-grade over-ear noise-canceling headphone. This headphone is designed for professional users with extremely high requirements for noise cancellation performance and sound quality, employing more powerful hardware configurations and more refined algorithm optimizations. In terms of hardware, it uses a 6-microphone array (2 feedforward microphones, 2 feedback microphones, and 2 call microphones) and is equipped with a high-performance dual-core DSP processor, capable of supporting higher-order filters and more complex algorithms.

[0047] Compared with Example 1, this example has been optimized and upgraded in the following aspects: First, in terms of transient noise detection, this embodiment adds a noise direction recognition function. By performing cross-correlation analysis on the signals from two feedforward microphones, the arrival time difference of transient noise is calculated to determine the source direction of the noise. Depending on the direction of the noise, the system adopts different suppression strategies. For transient noise from the front, the feedforward filter is mainly used for fast suppression. For transient noise from the side or rear, the parameters of the feedforward and feedback filters are adjusted simultaneously to achieve all-round transient noise suppression. Experiments show that after adding the direction recognition function, the suppression effect of transient noise from the side and rear is improved by more than 10dB.

[0048] Secondly, regarding the hybrid noise reduction model, this embodiment adopts a dual feedback filter structure. Two feedback microphones are installed at the top and bottom positions inside the earcups, which can more accurately monitor residual noise in different areas inside the earcups. The system is equipped with an independent feedback filter for each feedback microphone. The outputs of the two filters are weighted and fused before being combined with the output of the feedforward filter. This dual feedback structure can effectively improve the problem of uneven sound field inside the earcups, making the noise reduction effect of the entire earcup area more consistent. In particular, for users who wear glasses, it can significantly reduce the sound leakage problem caused by the temples of glasses.

[0049] Third, regarding the transparency mode, this embodiment introduces spatial audio technology, using a 6-microphone array to collect ambient sound signals from all directions, and using beamforming algorithms to recreate a realistic three-dimensional sound field. At the same time, the system has built-in accelerometers and gyroscopes, which can detect the user's head posture in real time and dynamically adjust the gain and phase of ambient sound in different directions according to the direction of head rotation, so that the user can obtain a more natural spatial listening experience. For example, when the user turns their head to the left, the ambient sound from the right side will be enhanced accordingly, which is consistent with the auditory experience of the real world.

[0050] Fourth, in terms of adaptive calibration, this embodiment adopts a more refined personalized calibration function. When used for the first time, the system will guide the user to perform a personalized calibration, play a series of test tones of different frequencies, measure the actual sealing characteristics of the user's earcups through the feedback microphone, and generate a unique equalization compensation curve and filter coefficient. Personalized calibration can fully consider the differences in head shape, ear shape and wearing habits of different users, so that each user can obtain the best noise reduction effect and listening experience.

[0051] Fifth, this embodiment also adds an intelligent scene recognition function. The system can automatically identify the user's scene based on the characteristics of environmental noise, such as office, subway, airplane, outdoors, etc., and automatically switch to the corresponding noise reduction mode and transparency mode parameters. For example, on an airplane, the system will automatically enhance the low-frequency noise reduction depth, and in an office, the system will appropriately reduce the noise reduction depth to retain some human voices for easy communication with colleagues.

[0052] This embodiment demonstrates the application of the adaptive hybrid active noise cancellation control method for headphones in high-end products. Through further hardware upgrades and algorithm optimization, the technical solution of this invention can meet the requirements of professional users for ultimate noise cancellation performance and sound quality. This embodiment achieves industry-leading noise cancellation performance and user experience through in-depth optimization and expansion. The successful application of this high-end headphone shows that the technical solution of this invention is not only applicable to mainstream consumer products, but also meets the stringent requirements of the professional market.

[0053] The present invention has been described in detail above. However, modifications or improvements can be made to it, which will be obvious to those skilled in the art. Therefore, any modifications or improvements that do not depart from the spirit of the present invention are within the scope of protection of the present invention.

Claims

1. A method of adaptive hybrid active noise control for headphones, the method comprising: Includes the following steps: S1. Noise signal acquisition: Simultaneously acquire external environmental noise signals and residual noise signals inside the earcups; S2. Rapid Transient Noise Identification: Multi-dimensional feature extraction and rapid transient noise identification of external noise signals; S3. Hybrid noise reduction model construction: Construct a feedforward-feedback hybrid noise reduction model based on the minimum mean square filtering algorithm, and update the filter coefficients in real time iteratively. S4. Adaptive parameter adjustment: Based on the transient noise identification results and the type of environmental noise, the filter response parameters and noise reduction weight allocation are adaptively adjusted. S5. Cancellation signal generation and playback: Generates a cancellation signal that is out of phase with the target noise, driving the sound unit to achieve active noise suppression.

2. The adaptive hybrid active noise cancellation control method for headphones according to claim 1, characterized in that, The rapid identification step for transient noise is as follows: the acquired external noise signal is processed by framing and windowing, and the short-time energy, spectral entropy and zero-crossing rate characteristics of each frame are calculated. When multiple features simultaneously meet the preset judgment conditions, it is marked as a transient noise event.

3. The adaptive hybrid active noise cancellation control method for headphones according to claim 2, characterized in that, The specific steps of the adaptive adjustment of filter response parameters are as follows: when a transient noise event is detected, the filter step response time is immediately compressed to below a preset threshold, and the noise reduction gain in the mid-to-high frequency band is increased. When the transient noise event ends, the filter is smoothly restored to the normal noise reduction parameters.

4. The adaptive hybrid active noise cancellation control method for headphones according to claim 1, characterized in that, The specific steps for constructing the feedforward-feedback hybrid noise reduction model are as follows: train a feedforward adaptive filter using the feedforward noise signal, train a feedback adaptive filter using the feedback residual noise signal, and dynamically weight and fuse the outputs of the two filters to generate the final noise reduction and cancellation signal.

5. The adaptive hybrid active noise cancellation control method for headphones according to claim 4, characterized in that, The dynamic weighted fusion step specifically involves: analyzing the spectral characteristics and stability of environmental noise in real time; increasing the output weight of the feedforward filter when it is determined to be steady-state noise; and increasing the output weight of the feedback filter when it is determined to be non-steady-state noise.

6. The adaptive hybrid active noise cancellation control method for headphones according to claim 1, characterized in that, It also includes a transparent mode signal processing step: acquiring external ambient sound signals, performing frequency domain equalization compensation, time domain delay alignment and dynamic range control in sequence, and then superimposing them with the original sound field after passive sound insulation for output.

7. The adaptive hybrid active noise cancellation control method for headphones according to claim 6, characterized in that, The frequency domain equalization compensation step specifically involves: pre-measuring the frequency response curve of the passive noise insulation of the headphones, generating an inverse equalization filter, filtering the digitally acquired ambient sound signal, and compensating for the frequency band attenuation caused by the passive noise insulation.

8. The adaptive hybrid active noise cancellation control method for headphones according to claim 6, characterized in that, The time-domain delay alignment step specifically involves: calculating the propagation delay time of sound waves through the physical structure of the earmuffs, delaying the digitally processed ambient sound signal by the same time, and then superimposing the sound fields to eliminate auditory distortion caused by the time difference.

9. The adaptive hybrid active noise cancellation control method for headphones according to claim 1, characterized in that, It also includes a call noise reduction processing step: using a dual-microphone array to collect a mixed signal of speech and noise, and using spatial beamforming and adaptive noise cancellation algorithms to separate and output a clean human voice signal.

10. The adaptive hybrid active noise cancellation control method for headphones according to claim 1, characterized in that, It also includes an adaptive calibration step for noise reduction effect: during the use of the headphones, residual noise signals inside the earcups are periodically collected, the noise reduction effect is analyzed, and the filter coefficients are automatically updated to compensate for changes in earcup sealing and the effects of component aging.

Images