Hybrid noise reduction method and system for Bluetooth headset

Abstract

The invention relates to the technical field of audio noise reduction, in particular to a hybrid noise reduction method and system for a Bluetooth headset. The method comprises the following steps: obtaining multi-dimensional environment noise data based on an environment noise sensor, and carrying out noise component feature analysis to generate environment noise multi-frequency component features; performing intelligent noise reduction intensity and self-adaptive filtering noise reduction suppression on the multi-frequency component characteristics of the environmental noise so as to generate an external environmental noise reduction suppression strategy; a micro audio signal is sent based on an ear canal form sensor, and audio signal feedback information is collected; performing three-dimensional ear canal structure modeling according to the audio signal feedback information so as to construct a three-dimensional ear canal structure model of the user; and obtaining a noise reduction feedback log of the wearer, and carrying out personalized noise reduction demand analysis to obtain the noise reduction demand of the user in the current scene. According to the invention, through comprehensive and finer mixed noise suppression, the audio quality and stability of the Bluetooth earphone are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of audio noise reduction, and in particular to a hybrid noise reduction method and system for a Bluetooth headset. Background Art

[0002] With the continuous development of wireless communication technology and the gradual maturity of Bluetooth technology, Bluetooth headsets are increasingly used in daily life and work. Especially in noisy environments, Bluetooth headsets are increasingly used for functions such as calls, music playback, and voice assistants. However, noise interference is still one of the main problems affecting the user experience of Bluetooth headsets. Especially in noisy environments, traditional Bluetooth headsets often cannot effectively isolate external noise, resulting in poor call quality, unclear sound quality, and other problems, which seriously affects the user experience.

[0003] In order to improve the noise suppression ability of Bluetooth headsets in complex environments, hybrid noise reduction technology has gradually become an effective means to solve this problem. The hybrid noise reduction method combines active noise reduction (ANC) and passive noise reduction (PNC) two noise reduction methods. Active noise reduction uses a built-in microphone to capture external noise in real time and generate reverse sound waves to offset it, while passive noise reduction relies on the physical structure of the headset itself, such as the design of ear pads and earplugs, to achieve noise reduction by isolating external noise. The two noise reduction methods complement each other and can greatly improve the noise reduction effect, especially in more complex noise environments, where the performance is more superior.

[0004] However, with the development of technology, the hybrid noise reduction method of Bluetooth headsets still faces some challenges. First of all, how to adapt to noise changes in real time in a dynamically changing environment and adjust the noise reduction strategy is one of the difficulties in current technology. Secondly, the impact on sound quality during the noise reduction process is also an issue that cannot be ignored. Excessive noise reduction may cause distortion of sound quality or audio delay, affecting the user's auditory experience. Therefore, in order to meet the growing market demand, it is particularly important to develop an efficient, accurate and less impact-on-sound-quality hybrid noise reduction method for Bluetooth headsets. Summary of the invention

[0005] In order to solve the above technical problems, the present invention proposes a hybrid noise reduction method and system for a Bluetooth headset to solve at least one of the above technical problems.

[0006] To achieve the above object, the present invention provides a hybrid noise reduction method for a Bluetooth headset, wherein the Bluetooth headset includes an ambient noise sensor and an ear canal morphology sensor, and comprises the following steps: Step S1: acquiring multi-dimensional environmental noise data based on an environmental noise sensor, and performing noise component feature analysis to generate multi-frequency component features of the environmental noise; Step S2: performing intelligent noise reduction intensity and adaptive filtering noise reduction suppression on the multi-frequency component characteristics of the environmental noise, thereby generating an external environment noise reduction suppression strategy; Step S3: sending a tiny audio signal based on the ear canal morphology sensor and collecting audio signal feedback information; performing three-dimensional ear canal structure modeling according to the audio signal feedback information, thereby constructing a three-dimensional ear canal structure model of the user; Step S4: Obtain the wearer's noise reduction feedback log and perform personalized noise reduction demand analysis to obtain the user's noise reduction demand in the current scene; Step S5: performing different frequency sensitivity analysis on the user's three-dimensional ear canal structure model, and calculating the optimal noise reduction intensity based on the user's noise reduction requirements in the current scene, so as to obtain the optimal noise reduction intensity in the current scene; Step S6: Perform reverse audio noise reduction processing according to the optimal noise reduction strength of the current scene, and perform hybrid collaborative noise reduction optimization according to the external environment noise reduction suppression strategy, thereby building a hybrid noise reduction optimization engine.

[0007] The present invention obtains multi-dimensional environmental noise data through an environmental noise sensor, and can accurately capture the multi-frequency components of external noise. The characteristic analysis of noise components helps to identify the type and intensity of noise, and then provides a more comprehensive and accurate noise model, providing data support for subsequent noise reduction processing. The key to this step is to analyze the noise components in detail, so as to provide high-quality data input for the design of noise reduction strategies. Based on the known noise components, the noise reduction intensity is dynamically adjusted by an intelligent algorithm, and the noise is suppressed by adaptive filtering technology, which can greatly reduce the interference of environmental noise to users. This process can adapt to the changes in environmental noise in real time, ensure that the noise reduction effect is always maintained at the optimal level, and avoid the loss of sound quality or insufficient noise reduction effect caused by excessive noise reduction. The morphology of the ear canal has a great influence on the transmission and perception of sound. By obtaining accurate data of the user's ear canal through the ear canal morphology sensor, and performing three-dimensional modeling through feedback information, the propagation path of the audio signal can be optimized according to the characteristics of the ear canal of each user, ensuring the clarity of the sound and the maximization of the noise reduction effect. This not only helps to formulate personalized noise reduction strategies, but also enhances the sound quality experience. The noise reduction needs of users vary from person to person, and the environment, scene and wearer's preferences may affect the intensity and method of noise reduction. By obtaining the wearer's noise reduction feedback log and analyzing their personalized needs, we can accurately understand the user's needs for noise reduction in different scenarios. This enables the headset to be intelligently adjusted according to the user's actual needs, providing a noise reduction experience that is more in line with personal needs. Each person's ear canal has different sensitivities to sounds of different frequencies, so a detailed sensitivity analysis of each user's ear canal is required. According to the current scenario and the user's noise reduction needs, the noise reduction intensity that best suits the user is calculated. This personalized noise reduction intensity calculation ensures that the noise reduction effect is not only efficient but also retains the real sound quality, avoiding excessive or insufficient noise reduction. Reverse audio noise reduction technology effectively suppresses external noise by generating sound waves that are in the opposite direction of the noise. The noise reduction effect can be further improved by hybrid synergy with the external environment noise reduction suppression strategy, especially in complex noise environments. This hybrid noise reduction optimization engine not only provides more efficient noise reduction performance, but also adapts to changing environments, ensuring an ideal noise reduction experience in various noise scenarios.

[0008] In this specification, a hybrid noise reduction system for a Bluetooth headset is provided, which is used to execute the hybrid noise reduction method for a Bluetooth headset as described above, including: A noise component analysis module is used to obtain multi-dimensional environmental noise data based on the environmental noise sensor and perform noise component feature analysis to generate multi-frequency component features of environmental noise; The filtering and suppression module is used to perform intelligent noise reduction intensity and adaptive filtering noise reduction suppression on the multi-frequency component characteristics of environmental noise, thereby generating an external environment noise reduction suppression strategy; The ear canal structure module is used to send tiny audio signals based on the ear canal morphology sensor and collect audio signal feedback information; perform three-dimensional ear canal structure modeling based on the audio signal feedback information, thereby constructing a three-dimensional ear canal structure model of the user; The noise reduction demand module is used to obtain the wearer's noise reduction feedback log and perform personalized noise reduction demand analysis to obtain the user's noise reduction demand in the current scene; The optimal noise reduction calculation module is used to analyze the sensitivity of the user's three-dimensional ear canal structure model to different frequencies, and calculate the optimal noise reduction intensity based on the user's noise reduction requirements in the current scene to obtain the optimal noise reduction intensity in the current scene; The hybrid collaborative noise reduction module is used to perform reverse audio noise reduction processing according to the optimal noise reduction intensity of the current scene, and to perform hybrid collaborative noise reduction optimization according to the external environment noise reduction suppression strategy, thereby building a hybrid noise reduction optimization engine.

[0009] The present invention can more comprehensively understand the spatial and frequency distribution of noise through multi-dimensional environmental noise data. This process provides accurate data support for subsequent noise classification and dynamic adjustment of noise reduction intensity. By decomposing the frequency components of noise, different types of noise, such as low-frequency vibration noise or high-frequency sharp noise, can be identified, thereby providing targeted processing strategies for noise reduction algorithms. The noise reduction intensity is adjusted according to the changes in real-time environmental noise to ensure that a stronger noise reduction effect is provided when the noise intensity is high, and excessive noise reduction is avoided when the noise is weak, thereby improving the sound quality. Adaptive filtering technology can dynamically adjust the filtering parameters according to the characteristics of different environmental noises, thereby effectively suppressing unwanted noise while maintaining the naturalness and clarity of the sound. By obtaining the three-dimensional structure of the ear canal, the system can more accurately understand the reflection and attenuation characteristics of the wearer's ear canal to sound, thereby customizing noise reduction and audio enhancement strategies to ensure that each wearer can get a good sound experience. Through the modeling of the ear canal structure, the earphone can better adjust the sound output to avoid sound distortion or discomfort caused by differences in the shape of the ear canal. Intelligent learning is carried out according to the wearer's noise reduction feedback log, and the noise reduction strategy is continuously optimized to provide a personalized noise reduction experience. This means that the needs of each wearer will be tracked and accurately met over the long term. The wearer's noise reduction needs will change over time and environment, and the system can intelligently adjust based on historical data and real-time feedback to ensure that the noise reduction effect always meets the wearer's needs. By analyzing the ear canal structure and frequency sensitivity, the system can accurately identify the wearer's sensitivity in a specific frequency band, and then provide more personalized noise reduction processing. According to the individual differences and scene requirements of the wearer, the optimal noise reduction intensity is dynamically calculated to ensure that good noise reduction effects can be provided in various environments, while avoiding excessive noise reduction and loss of sound quality. By combining reverse noise reduction technology with external noise reduction strategies, noise reduction effects can be exerted simultaneously inside the ear canal and in the external environment, providing more comprehensive and refined noise suppression. The hybrid noise reduction optimization engine can make intelligent adjustments based on the wearer's needs and environmental changes to achieve the most optimized noise reduction effect. This collaborative noise reduction mechanism makes the noise reduction processing not only accurate and effective, but also natural and comfortable, avoiding the discomfort caused by excessive noise reduction. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 A schematic diagram of the steps of a hybrid noise reduction method for a Bluetooth headset according to the present invention; Figure 2 Detailed implementation flow chart of step S1; Figure 3 Detailed implementation flow chart of step S2; Figure 4 Detailed implementation flow chart of step S3. DETAILED DESCRIPTION

[0011] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0012] The present application example provides a hybrid noise reduction method and system for a Bluetooth headset. The execution subject of the hybrid noise reduction method and system for a Bluetooth headset includes but is not limited to: mechanical equipment, data processing platform, cloud server node, network upload device, etc. equipped with the system can be regarded as a general computing node of the present application, and the data processing platform includes but is not limited to: at least one of an audio image management system, an information management system, and a cloud data management system.

[0013] See also Figures 1 to 4 The present invention provides a hybrid noise reduction method for a Bluetooth headset, and the hybrid noise reduction method for a Bluetooth headset comprises the following steps: Step S1: acquiring multi-dimensional environmental noise data based on an environmental noise sensor, and performing noise component feature analysis to generate multi-frequency component features of the environmental noise; Step S2: performing intelligent noise reduction intensity and adaptive filtering noise reduction suppression on the multi-frequency component characteristics of the environmental noise, thereby generating an external environment noise reduction suppression strategy; Step S3: sending a tiny audio signal based on the ear canal morphology sensor and collecting audio signal feedback information; performing three-dimensional ear canal structure modeling according to the audio signal feedback information, thereby constructing a three-dimensional ear canal structure model of the user; Step S4: Obtain the wearer's noise reduction feedback log and perform personalized noise reduction demand analysis to obtain the user's noise reduction demand in the current scene; Step S5: performing different frequency sensitivity analysis on the user's three-dimensional ear canal structure model, and calculating the optimal noise reduction intensity based on the user's noise reduction requirements in the current scene, so as to obtain the optimal noise reduction intensity in the current scene; Step S6: Perform reverse audio noise reduction processing according to the optimal noise reduction strength of the current scene, and perform hybrid collaborative noise reduction optimization according to the external environment noise reduction suppression strategy, thereby building a hybrid noise reduction optimization engine.

[0014] The present invention obtains multi-dimensional environmental noise data through an environmental noise sensor, and can accurately capture the multi-frequency components of external noise. The characteristic analysis of noise components helps to identify the type and intensity of noise, and then provides a more comprehensive and accurate noise model, providing data support for subsequent noise reduction processing. The key to this step is to analyze the noise components in detail, so as to provide high-quality data input for the design of noise reduction strategies. Based on the known noise components, the noise reduction intensity is dynamically adjusted by an intelligent algorithm, and the noise is suppressed by adaptive filtering technology, which can greatly reduce the interference of environmental noise to users. This process can adapt to the changes in environmental noise in real time, ensure that the noise reduction effect is always maintained at the optimal level, and avoid the loss of sound quality or insufficient noise reduction effect caused by excessive noise reduction. The morphology of the ear canal has a great influence on the transmission and perception of sound. By obtaining accurate data of the user's ear canal through the ear canal morphology sensor, and performing three-dimensional modeling through feedback information, the propagation path of the audio signal can be optimized according to the characteristics of the ear canal of each user, ensuring the clarity of the sound and the maximization of the noise reduction effect. This not only helps to formulate personalized noise reduction strategies, but also enhances the sound quality experience. The noise reduction needs of users vary from person to person, and the environment, scene and wearer's preferences may affect the intensity and method of noise reduction. By obtaining the wearer's noise reduction feedback log and analyzing their personalized needs, we can accurately understand the user's needs for noise reduction in different scenarios. This enables the headset to be intelligently adjusted according to the user's actual needs, providing a noise reduction experience that is more in line with personal needs. Each person's ear canal has different sensitivities to sounds of different frequencies, so a detailed sensitivity analysis of each user's ear canal is required. According to the current scenario and the user's noise reduction needs, the noise reduction intensity that best suits the user is calculated. This personalized noise reduction intensity calculation ensures that the noise reduction effect is not only efficient but also retains the real sound quality, avoiding excessive or insufficient noise reduction. Reverse audio noise reduction technology effectively suppresses external noise by generating sound waves that are in the opposite direction of the noise. The noise reduction effect can be further improved by hybrid synergy with the external environment noise reduction suppression strategy, especially in complex noise environments. This hybrid noise reduction optimization engine not only provides more efficient noise reduction performance, but also adapts to changing environments, ensuring an ideal noise reduction experience in various noise scenarios.

[0015] In the embodiment of the present invention, refer to Figure 1 , is a schematic flow chart of the steps of a hybrid noise reduction method for a Bluetooth headset of the present invention. In this example, the steps of the hybrid noise reduction method for a Bluetooth headset include: Step S1: acquiring multi-dimensional environmental noise data based on an environmental noise sensor, and performing noise component feature analysis to generate multi-frequency component features of the environmental noise; In this embodiment, the built-in environmental noise sensor based on the Bluetooth headset ensures that its frequency response range can cover the sound frequency that can be perceived by the human ear, usually 20 Hz to 20 kHz. The sensor should have a low self-noise level to improve the accuracy of environmental noise data collection. The sensor is configured to collect data in a variety of environments, such as city streets, quiet indoor environments, industrial areas, and areas with heavy traffic. By adjusting the gain setting of the sensor, it is ensured that the signal can be effectively captured under different noise intensities. The gain is usually set between 20-40 dB to adapt to environmental changes. In the set environment, continuous noise data collection is performed. It is recommended to record data for at least 30 minutes to 1 hour in each environment to capture the changes and characteristics of the noise. The data sampling frequency should be set to a sampling rate of 44.1 kHz to ensure high-fidelity sound quality and details. At the same time, record the meteorological parameters of the environment (such as temperature, humidity, wind speed, etc.) so that the impact of these factors on noise characteristics can be considered in subsequent analysis. This information will help understand the changes in noise and its sources. The collected data needs to be preprocessed, including removing silent segments and non-correlated signals, and using digital filters (such as low-pass, high-pass or band-pass filters) to remove redundant frequency components. Window functions (such as Hamming windows) can be used to window the signal to reduce edge effects. Fast Fourier transform (FFT) is used to convert the time domain signal into a frequency domain signal, and the amplitude and phase information of each frequency component are analyzed. By analyzing the frequency domain data, the noise characteristics of different frequency bands can be identified. The power spectral density (PSD) of each frequency band is calculated to determine the energy distribution of the environmental noise. The frequency is divided into several standard frequency bands (such as low frequency: 20 Hz-250 Hz, medium frequency: 250 Hz-2000 Hz, high frequency: 2000 Hz-20 kHz), and the total energy of each frequency band is calculated. The generated multi-frequency component feature data is sorted, including the center frequency, bandwidth, power value and relative intensity of each frequency band. A database is established to store these features for subsequent analysis and comparison.

[0016] Step S2: performing intelligent noise reduction intensity and adaptive filtering noise reduction suppression on the multi-frequency component characteristics of the environmental noise, thereby generating an external environment noise reduction suppression strategy; In this embodiment, a noise reduction intensity calculation model is constructed based on the multi-frequency component characteristics of the ambient noise generated in the previous step. The model should take into account the noise energy distribution of each frequency band and the user's comfort requirements. Use machine learning algorithms such as linear regression or random forest to analyze the relationship between noise intensity and noise reduction requirements in different environments. During the model training process, the historical data set, including user feedback information, noise intensity and frequency component characteristics in different environments, is used to optimize the model parameters. Ensure that the model can accurately predict the required noise reduction intensity in different environments. Use the established noise reduction intensity calculation model to perform real-time analysis on the multi-frequency component characteristics of the current environment to generate an intelligent noise reduction intensity value. This value should be able to reflect the noise reduction capability required by the headset in a specific environment. Generally, the unit of noise reduction intensity can be expressed in decibels (dB). The set noise reduction intensity should be based on the user's subjective comfort and the objective intensity of the ambient noise to ensure that effective noise reduction effects can be provided in different noise environments. Use adaptive filtering technology (such as LMS or NLMS algorithm) to design a noise reduction filter. The input signal of the adaptive filter is the real-time collected data of the ambient noise, and the output signal is the audio signal after noise reduction processing. Set the learning rate and filter order of the filter. The learning rate is usually selected between 0.01 and 0.1 to ensure the convergence speed and stability of the filter. The selection of the filter order needs to balance the processing delay and the noise reduction effect. The order is usually selected between 32 and 64. Process the environmental noise signal through the adaptive filter to generate a noise-reduced output signal. Record the noise reduction effect of each frequency band, including the signal waveform, spectrum and signal-to-noise ratio (SNR) before and after noise reduction, to evaluate the effectiveness of the noise reduction strategy. Organize the parameter settings and results of the noise reduction process, generate an external environment noise reduction suppression strategy document, and record the noise reduction intensity, filter settings and final noise reduction effect in different environments. Ensure that the strategy can adapt to the needs of users in different scenarios.

[0017] Step S3: sending a tiny audio signal based on the ear canal morphology sensor and collecting audio signal feedback information; performing three-dimensional ear canal structure modeling according to the audio signal feedback information, thereby constructing a three-dimensional ear canal structure model of the user; In this embodiment, based on the ear canal morphology sensor inside the Bluetooth headset, the sensor can send a tiny audio signal and capture the reflected signal inside the ear canal. Common sensor types include ultrasonic sensors and miniature microphones, which can work in the ear canal without causing any discomfort. The sensor is configured to adapt to the shape of the user's ear canal to ensure good contact between the sensor and the ear canal. The frequency of the transmitted audio signal is set between 1 kHz and 4 kHz. This frequency range can effectively penetrate the ear canal and provide clear feedback information. The sampling rate of the sensor should be set to 44.1 kHz to ensure that sufficiently detailed signal changes are captured. In the user's ear canal, the ear canal morphology sensor is activated to send a tiny audio signal. Generally, the intensity of the transmitted signal should be kept below 60 dB SPL to avoid interference or discomfort to the user. The signal can be a simple sine wave or a pulse signal. During the propagation of the audio signal, the sensor simultaneously collects the reflected audio signal. This feedback information contains the geometric characteristics of the ear canal and the propagation characteristics of the sound wave in the ear canal. The timestamp, signal strength and frequency characteristics during the acquisition process are recorded for subsequent processing. The collected audio signal feedback information is digitally processed. First, a filter is used to remove background noise and interference signals. A bandpass filter can be used to ensure that only signals in the range of 1 kHz to 4 kHz are retained to improve the clarity of the data. Then, the fast Fourier transform (FFT) is applied to convert the time domain signal into a frequency domain signal to extract the amplitude and phase information of each frequency component. This process can help identify the sound wave reflection pattern in the ear canal and then analyze the geometric characteristics of the ear canal. According to the processed audio signal feedback information, the three-dimensional ear canal structure modeling is performed using an acoustic model and a geometric reconstruction algorithm. An inversion algorithm can be used to deduce the geometry of the ear canal through the propagation characteristics of sound waves, including the length, width, curvature and interface characteristics of the ear canal. Commonly used modeling methods include the finite element method (FEM) or the boundary element method (BEM), which can handle complex geometric structures and provide accurate acoustic characteristics. Finally, the generated three-dimensional ear canal structure model should contain detailed geometric information and acoustic characteristics of the ear canal. The model is visualized using computer-aided design (CAD) software for subsequent analysis and application. The constructed ear canal structure model is stored in the database to ensure data integrity and traceability. The key parameters, experimental conditions and modeling results in the model construction process are recorded to facilitate subsequent verification and optimization.

[0018] Step S4: Obtain the wearer's noise reduction feedback log and perform personalized noise reduction demand analysis to obtain the user's noise reduction demand in the current scene; In this embodiment, it is first necessary to integrate a noise reduction feedback log system into the Bluetooth headset. The system can record the wearer's subjective feedback on the noise reduction effect during the use of the headset in real time, including information such as satisfaction rating, noise reduction intensity adjustment and scene change. Design a user-friendly interface to allow users to submit feedback conveniently during use. For example, a concise feedback option can be provided through the touch panel on the headset or the supporting application, allowing users to choose their noise reduction experience (such as "satisfied", "average", "unsatisfied") when listening to music or talking on the phone. At each feedback submission, the system should record relevant environmental parameters (such as noise level, environment type, usage time, etc.) to ensure the comprehensiveness and accuracy of the data. It is recommended that each user use the headset in different environments for at least 1 week to accumulate sufficient feedback data. All noise reduction feedback logs obtained should be stored in the cloud or local database in real time to ensure data security and accessibility. Record the timestamp, environmental status and noise reduction intensity set by the user for each feedback for subsequent analysis. During the data sorting process, remove duplicate feedback records and obvious outliers (for example, users submit multiple feedbacks in a very short period of time) to ensure data quality. Data cleaning techniques such as range limiting and mean smoothing can be used to improve the reliability of the data. Use statistical analysis methods to analyze the collected noise reduction feedback data to identify users' noise reduction needs and preferences. Descriptive statistical analysis can be used to calculate the average, standard deviation and distribution of user satisfaction in different environments to understand the overall trend. For different scenarios (such as indoors, outdoors, and traffic), compare the noise reduction intensity and satisfaction of user feedback. Identify the characteristics of noise reduction requirements in each scenario, and establish a model to predict users' noise reduction needs in specific environments. Cluster analysis can be used to group users and identify the noise reduction preferences of different user groups. Based on the analysis results, generate a personalized noise reduction demand report that describes in detail the noise reduction needs of each user in different scenarios. This report should include the noise reduction intensity preferred by the user, common types of environmental noise, and the user's subjective satisfaction score.

[0019] Step S5: performing different frequency sensitivity analysis on the user's three-dimensional ear canal structure model, and calculating the optimal noise reduction intensity based on the user's noise reduction requirements in the current scene, so as to obtain the optimal noise reduction intensity in the current scene; In this embodiment, according to the user's three-dimensional ear canal structure model obtained in the previous step, detailed parameterization is performed using computer-aided design (CAD) software. This model should include the length, width, curvature, and acoustic properties (such as impedance and attenuation rate) of the ear canal. By simulating the propagation of sound waves in the ear canal, the resonance characteristics and frequency response of the ear canal are analyzed. The finite element analysis (FEA) method is used to calculate the propagation characteristics of sound waves in the ear canal and evaluate the sensitivity at different frequencies. This process needs to consider the complex geometry of the ear canal and the incident angle of the sound wave. In the established three-dimensional ear canal model, acoustic simulation is performed for different frequencies (for example, 250 Hz, 500 Hz, 1000 Hz, 2000 Hz, and 4000 Hz). By inputting a sine wave signal of the corresponding frequency, observing its propagation and reflection in the ear canal, and recording the sound pressure level (SPL) and corresponding resonance characteristics of each frequency. The gain and attenuation of each frequency in the ear canal are calculated to generate a frequency response curve. These curves can help identify the sensitivity of the user's ear canal to different frequencies and determine which frequencies have the greatest impact on the user's hearing. Combine the noise reduction demand information obtained from the user feedback log with the frequency sensitivity analysis of the ear canal. Analyze the user's noise reduction needs in specific environments and compare them with the frequency response data to determine the frequency components that need to be suppressed most in different scenarios. If the user feedback needs more noise reduction in a noisy urban environment, analyze the main noise frequency components of the environment (such as the low-frequency part of traffic noise) to focus on optimizing the noise reduction effect of these frequencies. Based on the above analysis, calculate the optimal noise reduction intensity in the current scenario. The weighted average method can be used to combine the sensitivity of different frequencies with the user's noise reduction needs to obtain a comprehensive noise reduction intensity value. Apply the calculated optimal noise reduction intensity for the current scenario to the noise reduction algorithm of the headset for actual testing. After the user wears the headset, observe its usage feedback in the selected scenario to evaluate whether the noise reduction effect meets expectations.

[0020] Step S6: Perform reverse audio noise reduction processing according to the optimal noise reduction strength of the current scene, and perform hybrid collaborative noise reduction optimization according to the external environment noise reduction suppression strategy, thereby building a hybrid noise reduction optimization engine.

[0021] In this embodiment, the basic principle of reverse audio noise reduction is to eliminate external noise by generating an audio signal with a phase opposite to that of the external noise. During the implementation process, it is necessary to set the microphone of the headset to capture the noise signal of the surrounding environment and record its amplitude and phase characteristics. A digital signal processing algorithm is used to generate an audio signal opposite to the captured noise signal. This process usually uses phase reversal technology, and the specific method is to analyze the spectral characteristics of the external noise signal through fast Fourier transform (FFT) and perform inversion processing on each frequency component. The generated reverse audio signal must be properly filtered to eliminate high-frequency noise and nonlinear distortion. A digital filter can be used to ensure that the quality of the output signal meets the requirements of user experience. The cutoff frequency of the filter is set within the user's sensitive frequency range, usually 1 kHz to 4 kHz, to ensure the effectiveness of the noise reduction effect. While performing reverse audio noise reduction, it is necessary to perform hybrid collaborative noise reduction optimization according to the external environment noise reduction suppression strategy. The external environment noise reduction suppression strategy generated in the previous step is used, combined with the noise characteristics of the current scene, to dynamically adjust the noise reduction processing parameters. An adaptive filtering algorithm (such as LMS or NLMS) is used to adjust the noise reduction intensity in real time to cope with changes in environmental noise. During this process, the headphones need to continuously monitor the noise level of the surrounding environment, and adjust the strength and phase of the reverse audio signal according to the changes, combining the reverse audio noise reduction processing with the environmental noise suppression strategy to build a hybrid noise reduction optimization engine. The engine should be able to process audio signals in different environments in real time and automatically select the optimal noise reduction strategy. The feedback mechanism of the engine is designed so that it can self-adjust according to the actual user's usage and changes in environmental noise. This mechanism can be implemented through machine learning algorithms to analyze user feedback and environmental data and optimize noise reduction processing strategies.

[0022] In this embodiment, refer to Figure 2 , is a flowchart of detailed implementation steps of step S1. In this embodiment, the detailed implementation steps of step S1 include: Step S11: acquiring multi-dimensional environmental noise data based on the environmental noise sensor; Step S12: performing multi-scale time-frequency decomposition on the multi-dimensional environmental noise data to generate multiple environmental noise frequency sub-bands; Step S13: Calculating the frequency components of the multiple ambient noise frequency sub-bands one by one to generate a sub-frequency component of each sub-band; Step S14: Calculate the power spectrum density of multiple ambient noise frequency sub-bands to generate power spectrum density values; Step S15: performing noise component feature analysis according to the power spectrum density value and the sub-frequency components of each sub-band, thereby generating multi-frequency component features of the environmental noise.

[0023] In this embodiment, based on the multi-dimensional environmental noise sensors built into the Bluetooth headset, these sensors should support the measurement of multiple parameters, such as noise intensity, frequency spectrum, sound pressure level (SPL), etc. It is recommended to use sensors with high sensitivity that can capture lower noise levels. Configure the working parameters of the sensor, including the sampling frequency (recommended to be 44.1 kHz or above) and the measurement time length to ensure the continuity and accuracy of the data. Collect data under different environmental conditions (such as cities, industrial areas, and quiet natural environments) to ensure coverage of multiple noise sources and environmental changes. Record the time, location, and environmental conditions of each collection. Use an automated data recording system to transmit the raw noise data obtained by the sensor to the data processing center in real time to ensure the integrity and timeliness of the data. Store the collected multi-dimensional environmental noise data in a database to ensure structured storage of the data, including information such as timestamps, noise intensity, and frequency. The data should be classified by location and time for subsequent analysis. Perform a preliminary analysis to calculate the average noise level and its fluctuation range at each collection point, identify potential noise hotspots, and provide a reference for subsequent multi-scale analysis. Select a suitable time-frequency analysis method, such as wavelet transform or short-time Fourier transform (STFT). These methods can effectively capture the changing characteristics of the signal in time and frequency. Determine the scale and parameter settings of the decomposition, such as wavelet basis (such as Daubechies wavelet) and scale factor, to generate multiple frequency sub-bands of environmental noise. Input the collected multi-dimensional environmental noise data into the selected time-frequency analysis algorithm for multi-scale time-frequency decomposition. Record the frequency range and corresponding time window of each sub-band. Generate multiple frequency sub-bands, each of which represents the characteristics of environmental noise in a different frequency range (such as low frequency, medium frequency and high frequency) for subsequent analysis. Store the generated frequency sub-band data in a database to ensure the integrity of the frequency characteristics and time information of each sub-band. Record the parameter settings and output results during the decomposition process for subsequent verification and analysis. Generate a visual chart to show the time-frequency characteristics of different sub-bands to help understand the noise distribution in each frequency band. Select a suitable frequency component calculation method, such as Fourier transform, which can decompose each frequency sub-band into multiple frequency components for analyzing its components. Determine the calculation parameters, such as frequency resolution and window function type, to ensure the accuracy of the calculation results. Perform Fourier transform on each frequency subband and calculate its frequency components. Record the amplitude and phase information of each frequency component to ensure data integrity. Process each subband one by one, generate sub-frequency components for each subband, and record the characteristics of each frequency component for subsequent analysis. Store the calculated frequency component information in a database to ensure that the frequency characteristics and component information of each subband are complete and traceable. Record the parameter settings and results during the calculation process to facilitate subsequent analysis and optimization. Perform comparative analysis of frequency components, identify the differences in frequency components between different subbands, and explore their relationship with noise sources.Select a suitable power spectral density calculation method, such as the Welch method or the autospectral estimation method, which can effectively estimate the power distribution of the signal. Determine the calculation parameters, including segment length and overlap ratio, to optimize the estimation of power spectral density. Calculate the power spectral density for each frequency subband and record the power value of each frequency component to ensure the accuracy and reliability of the calculation results. Generate a power spectral density plot for each subband to show the power distribution of different frequency bands to facilitate the analysis of the characteristics of the noise source. Store the calculated power spectral density values ​​in a database to ensure the integrity and traceability of the data. Record the parameter settings and results during the calculation process for subsequent analysis. Select a suitable feature analysis method, such as principal component analysis (PCA) or factor analysis, to extract the main noise component characteristics from the power spectral density values ​​and frequency components. Determine the input parameters of the analysis, including power spectral density values, frequency components and their corresponding subbands, to facilitate the generation of comprehensive features. Input the power spectral density values ​​and the sub-frequency components of the frequency subbands into the selected feature analysis model for comprehensive analysis to identify the main component characteristics of the environmental noise. Record the characteristic values ​​of each noise component, including frequency range, power value and component type, for subsequent analysis and application. Store the multi-frequency component characteristics of environmental noise obtained by analysis in the database to ensure data integrity and traceability. Record the parameter settings and results during the analysis process for subsequent verification and optimization. Generate a noise component characteristic analysis report to display the main components and characteristics of environmental noise, helping to understand the impact of noise on the environment and its management countermeasures.

[0024] In this embodiment, refer to Figure 3 , is a flowchart of detailed implementation steps of step S2. In this embodiment, the detailed implementation steps of step S2 include: Step S21: performing noise type cluster analysis on the multi-frequency component characteristics of the environmental noise, thereby generating characteristic types of noise in the environment; Step S22: identifying dynamic noise type changes of noise feature types in the environment to obtain dynamic type change data of environmental noise; Step S23: performing intelligent noise reduction intensity calculation based on the dynamic type change data of the environmental noise to generate dynamic noise reduction intensity parameters; Step S24: Adaptively filter and suppress noise reduction according to the dynamic noise reduction strength parameter, thereby generating an external environment noise reduction suppression strategy.

[0025] In this embodiment, multi-frequency component feature data of environmental noise are collected, including power spectrum density, frequency components and corresponding time features. Ensure the integrity of the data and remove outliers to improve the accuracy of cluster analysis. Determine the features used for cluster analysis, including frequency center, power value, bandwidth, etc., so that different noise types can be effectively distinguished. It can be considered to use standardization processing to make the dimensions of different features consistent. Select a suitable clustering algorithm, such as K-means clustering, hierarchical clustering or DBSCAN. These algorithms can classify environmental noise into different types based on the similarity between features. Determine clustering parameters, such as the number of clusters (K value) or distance measurement methods (such as Euclidean distance, Manhattan distance) to improve the clustering effect. Perform cluster analysis on the multi-frequency component feature data of environmental noise, and record the center and distribution of each cluster. According to the similarity of the features, similar noise features are classified into the same category. Generate a noise feature type report to display the features of each noise type and its corresponding sample number to help identify the main noise sources in the environment. Store the results of the cluster analysis in a database, including a description of each noise feature type, the number of samples and its corresponding frequency features. Ensure data integrity and traceability. Generate visual charts to show the distribution of noise feature types, such as using scatter plots or radar charts, to help intuitively understand the characteristics of different noise types. Monitor environmental noise feature types in different time periods (such as morning, daytime and night) and collect dynamic change data. Ensure monitoring under different environmental conditions to capture comprehensive change characteristics. Record the noise feature types and their corresponding changes in each time period, including the frequency of occurrence and change trend of the feature types. Select appropriate dynamic change identification methods, such as time series analysis-based methods (such as sliding window technology, ARIMA model) or change point detection-based methods (such as CUSUM algorithm) to identify changes in noise feature types. Determine the time window and parameter settings for analysis, such as window size and significance level, to improve the accuracy of change identification. Analyze the collected dynamic change data to identify changes in noise feature types. Record the change trend and frequency of occurrence of each noise type for subsequent analysis. Select a suitable noise reduction intensity calculation model, which can be an experience-based model (such as equivalent continuous sound level calculation) or a regression model based on machine learning to accurately estimate the noise reduction intensity. Determine the input characteristics of the model, including dynamic noise type change data, current environmental noise level, and target noise reduction level. Input the dynamic noise type change data into the selected calculation model, calculate the noise reduction intensity, and generate dynamic noise reduction intensity parameters. Record the input characteristics and output results of each calculation to ensure data traceability. Generate a noise reduction intensity report to display the noise reduction intensity parameters under different noise types to help understand the required noise reduction intensity under different environmental conditions.Store the calculated dynamic noise reduction strength parameters in a database to ensure data integrity and traceability. Record the parameter settings and results during the calculation process for subsequent optimization and verification. Select a suitable adaptive filtering algorithm, such as the LMS (least mean square) algorithm or the RLS (recursive least squares) algorithm, to perform real-time noise reduction based on the dynamic noise reduction strength parameters. Determine the parameter settings of the filter, including the filter order, learning rate, and initial weight, and adjust the settings of the adaptive filter according to the dynamic noise reduction strength parameters to process the noise signals in the environment in real time. Record the filtering effect for each time period, including the signal-to-noise ratio (SNR) and noise reduction effect evaluation. Generate a noise reduction effect report to show the signal comparison before and after noise reduction to help identify the effectiveness of the noise reduction strategy.

[0026] In this embodiment, the specific steps of step S21 are: Calculate the noise intensity at multiple time points based on the multi-frequency component characteristics of environmental noise and extract the noise intensity values ​​at different time points; Perform real-time noise intensity change analysis based on noise intensity values ​​at different time points to generate real-time noise intensity change data; Perform time series discrete change fitting on real-time noise intensity change data to construct a time series noise intensity change curve; Quantify the frequency fluctuation range of the multi-frequency component characteristics of the environmental noise to obtain the frequency fluctuation range of the environmental noise; Perform noise duration statistics on the multi-frequency component characteristics of the environmental noise to generate the noise duration; Noise type clustering analysis is performed on the noise duration, ambient noise frequency fluctuation range and time-series noise intensity change curve to generate noise feature types in the environment, which include continuous noise, impact noise and dragging noise.

[0027] In this embodiment, multi-frequency component characteristic data of environmental noise is collected to ensure that the data covers the noise characteristics at multiple time points. The data should include the frequency component, power spectrum density and noise intensity at each time point. The method for calculating the noise intensity is usually to use the sound pressure level (SPL) calculation formula: ,in, is the measured sound pressure, The reference sound pressure (usually 20 µPa) is used to calculate the noise intensity value at each time point based on the collected frequency component characteristics. To ensure accuracy, it is recommended to use a high sampling rate (such as 44.1 kHz) for data acquisition. Use standardized methods to process noise intensity data to obtain more accurate and comparable results. Record the noise intensity value at each time point and its corresponding timestamp. Store the calculated noise intensity values ​​at different time points in the database to ensure the integrity and traceability of the data. Record the source and calculation process of each noise intensity value for subsequent analysis. Extract data at different time points from the stored noise intensity values ​​to ensure that the time intervals of the data are uniform for subsequent real-time change analysis. Determine the method for real-time noise intensity change analysis, usually using time series analysis techniques such as moving average or exponential smoothing to smooth the noise intensity data. Perform real-time change analysis on the extracted noise intensity values ​​and calculate the change amplitude and trend at each time point. Record the rate of change, which can usually be calculated using the formula ΔL=L(t)−L(t−1). Generate real-time noise intensity change data and record the changes at each time point, including rising, falling and stable states, for subsequent analysis. Select a suitable time series discrete change fitting model, such as polynomial regression or spline regression, to fit the real-time noise intensity change data. Determine the parameters and order of the fitting model, such as selecting a quadratic or cubic polynomial to better capture the change trend of the data. Input the real-time noise intensity change data into the selected fitting model to fit the time series discrete change and generate a noise intensity change curve. Record the key parameters in the fitting process, including the goodness of fit (such as R² value) and residual analysis to evaluate the fitting effect of the model. Store the fitted time series noise intensity change curve in the database to ensure the integrity and traceability of the data. Record the parameters and results used in the fitting process. Generate a visual chart to display the noise intensity change curve after fitting to help analyze the pattern and trend of noise intensity change. Select a suitable frequency fluctuation range calculation method, usually using standard deviation or range (difference between maximum and minimum values) to quantify the frequency fluctuation. Determine the input parameters of the calculation, including the upper and lower limits of the frequency component, to accurately quantify the frequency fluctuation range. Calculate the fluctuation range of the multi-frequency component characteristics of environmental noise and record the fluctuation of each frequency component. You can divide it by frequency band and calculate the fluctuation range of low frequency, medium frequency and high frequency separately. Generate a frequency fluctuation range report to show the fluctuation characteristics of different frequency bands to help understand the changes in environmental noise. Determine the calculation method for the duration of noise, usually using the time period when the sound pressure level exceeds a certain threshold for statistics. The threshold can be set to the average level of environmental noise plus 1-3 dB. Set the calculation time window to obtain accurate duration data.Analyze the multi-frequency component characteristics of environmental noise, record the time period when the noise exceeds the set threshold, and count the duration of each noise. Generate a noise duration statistics report to display the duration of different types of noise to help understand its impact on the environment. Determine the input characteristics of clustering, which usually include the characteristic values ​​of duration, frequency fluctuation range, and time series noise intensity change curve. Input the above characteristic data into the selected clustering algorithm to perform cluster analysis of noise types. Record the center and distribution of each cluster, and classify similar noise characteristics into the same category. Generate a noise characteristic type report to display the cluster analysis results, including the characteristics of continuous noise, impact noise, and drag noise. Store the cluster analysis results in the database to ensure that the description and sample quantity of each noise characteristic type are complete and traceable. Record the parameter settings and results during the analysis process. Generate a visual chart to display the clustering results of different noise types to help quickly understand the distribution and characteristics of different noise characteristics.

[0028] In this embodiment, refer to Figure 4 , is a flowchart of detailed implementation steps of step S3. In this embodiment, the detailed implementation steps of step S3 include: Step S31: sending a tiny audio signal based on the ear canal morphology sensor and collecting audio signal feedback information; Step S32: calculating the ear canal reflection intensity of the audio signal feedback information to obtain the audio reflection intensity; Step S33: performing signal attenuation rate analysis on the audio signal feedback information to generate a signal attenuation rate; Step S34: Calculate the reflection time according to the audio signal feedback information to obtain the audio reflection time; Step S35: performing audio reflection feature evolution on the audio reflection intensity, the signal attenuation rate and the audio reflection time, thereby generating an audio reflection evolution feature; Step S36: Perform three-dimensional ear canal structure modeling on the audio reflection evolution characteristics, thereby constructing a three-dimensional ear canal structure model of the user.

[0029] In this embodiment, the Bluetooth headset has an ear canal morphology sensor built in, and a generator of a tiny audio signal is configured on the ear canal morphology sensor to ensure that an audio signal of a specific frequency and amplitude can be generated. A signal with a frequency range of 1kHz to 4 kHz is usually selected because this range can effectively penetrate the ear canal and be reflected. The amplitude of the audio signal is set, usually around 70 dB SPL (sound pressure level) to ensure that the signal is neither too strong to damage hearing nor effectively reflected by the ear canal. The ear canal morphology sensor is started, and a tiny audio signal is sent to the wearer's ear canal through its transmitting device. Ensure that the duration of signal transmission is moderate, usually set between 200 milliseconds and 500 milliseconds, so that the ear canal has sufficient time for reflection. At the same time as the signal is sent, the feedback information collection system is started to record the audio signal reflection in the ear canal in real time. During the recording process, the ambient noise should be ensured to be lower than the background noise threshold (such as 30 dB SPL) to reduce interference. The feedback audio signal in the ear canal is collected, and the data should include the timestamp, frequency component and amplitude information of the signal. The collected feedback signal is stored in a high-capacity database to ensure the integrity and traceability of the information. Determine the input parameters for the calculation, including the peak sound pressure and time window of the feedback signal, to ensure that the reflection intensity is accurately calculated. Analyze the collected feedback audio signal, extract the peak sound pressure of the feedback signal, and use the above formula to calculate the reflection intensity. Ensure that high-accuracy signal processing tools are used in the calculation process to reduce errors. Record the reflection intensity value calculated each time to ensure detailed analysis of the reflection intensity at different time points and environmental conditions. Analyze the audio signal sent each time and the corresponding feedback signal to calculate the signal attenuation rate. Record the sound pressure values ​​of the input signal and the feedback signal to ensure the accuracy of the calculation. Generate a signal attenuation rate statistical report to show the attenuation at different time points to help identify the relationship between the acoustic characteristics of the ear canal and environmental factors. Store the calculated signal attenuation rate in the database to ensure data integrity and traceability. Record the parameters and results used in the calculation process for subsequent analysis and optimization. Generate a visual chart to show the changes in the signal attenuation rate to help quickly understand the acoustic characteristics and changing trends of the ear canal. The reflection time is calculated based on the basic physical principle of sound wave propagation in the ear canal and the speed of sound wave propagation (approximately 343 m / s). The reflection time is calculated as follows: T = 2D / V, where D is the ear canal length and V is the speed of sound wave propagation. Determine the average length of the ear canal, which is usually between 2.5 cm and 3.5 cm. The specific length can be measured based on the ear canal morphology of the participant. Calculate the reflection time based on the length of the ear canal and the speed of sound wave propagation. Compare the calculated results with the collected audio feedback information to verify the accuracy of the calculation. Record the reflection time value for each calculation, including the settings of the ear canal length and sound wave propagation speed, for subsequent analysis. Store the calculated reflection time in a database to ensure data integrity and traceability.Record the key parameter settings and results during the calculation process for subsequent verification and optimization. Generate a reflection time statistical report to show the changes in reflection time under different ear canal morphologies, providing a reference for the subsequent ear canal model construction. Select a suitable feature evolution analysis method, such as principal component analysis (PCA) or factor analysis (FA), to extract the main features from reflection intensity, signal decay rate and reflection time. Determine the input features of the analysis, including the values ​​of reflection intensity, decay rate and reflection time, to effectively extract audio reflection features. Input the reflection intensity, signal decay rate and reflection time data into the selected analysis model to perform evolution analysis of audio reflection features. Record the contribution and change trend of each feature. Generate an audio reflection feature evolution report to show the evolution process of reflection features to help understand the acoustic behavior and changes of the ear canal. Select a suitable three-dimensional modeling method, which can use computer-aided design (CAD) software or three-dimensional scanning technology to accurately construct the three-dimensional structure of the ear canal. Determine the input parameters of the modeling, including the average length, width and shape characteristics of the ear canal, to ensure the accuracy of the modeling. Construct a 3D ear canal model of the user based on the collected audio reflection evolution characteristics and physical parameters of the ear canal. Ensure that the model can reflect the actual morphology and acoustic characteristics of the ear canal. Record key parameter settings and results during the modeling process for subsequent verification and optimization. Store the constructed 3D ear canal structure model in a database to ensure data integrity and traceability. Record the parameters and results used in the modeling process for subsequent analysis and application. Generate a visualization of the 3D ear canal model to display the internal structure and acoustic characteristics of the ear canal, helping to improve understanding of the acoustic behavior of the ear canal.

[0030] In this embodiment, step S4 includes the following steps: Step S41: Obtaining the wearer's noise reduction feedback log; Step S42: identifying multiple historical headphone noise reduction scenarios based on the wearer's noise reduction feedback log; Step S43: performing noise reduction feedback analysis on multiple historical headphone noise reduction scenarios to extract noise reduction feedback information for each scenario; Step S44: performing scene feature analysis on the multi-frequency component features of the ambient noise to obtain the real-time scene of the user; Step S45: Perform personalized noise reduction demand analysis on the scene where the user is in real time according to the noise reduction feedback information of each scene, so as to obtain the noise reduction demand of the user in the current scene.

[0031] In this embodiment, the classification criteria for noise reduction scenes are set, which usually include traffic noise, indoor noise, outdoor noise, industrial noise, etc. According to the feedback content of the wearer, a suitable scene label is selected. A classification algorithm is determined, such as a rule-based classification or a machine learning-based classification model, and a decision tree or a support vector machine (SVM) can be used for scene recognition. The collected noise reduction feedback logs are analyzed, the keywords and scene descriptions in the feedback are extracted, and the scene recognition is performed for each feedback using the set classification criteria. The scene type corresponding to each feedback is recorded. A scene recognition report is generated to display the historical headphone noise reduction scenes identified and the corresponding number of feedbacks to help understand the user's noise reduction needs. The identified historical headphone noise reduction scenes are stored in a database to ensure the integrity and traceability of the data. The parameter settings and results in the recognition process are recorded for subsequent verification and optimization. Scene analysis is performed to identify the user feedback characteristics of different noise reduction scenes to help provide a basis for subsequent personalized noise reduction demand analysis. A suitable feedback information parsing method is selected, such as text mining technology and sentiment analysis, to extract noise reduction feedback information for each scene. Ensure that satisfaction and specific problems can be extracted from user feedback. Determine the input parameters for analysis, including the type of noise reduction scenario and user feedback content, so as to extract effective information. Perform noise reduction feedback analysis on the identified historical headphone noise reduction scenarios, and extract the satisfaction score, specific opinions and suggestions of each scenario. Use sentiment analysis tools to evaluate the sentiment tendency (positive, negative or neutral) of each feedback. Record the feedback information of each scenario and generate a feedback analysis report to show the satisfaction and main problems of each noise reduction scenario. Store the noise reduction feedback information obtained from the analysis in the database to ensure the integrity and traceability of the data. Record the key parameter settings and results during the analysis process for subsequent analysis and optimization. Determine the analysis criteria for the multi-frequency component characteristics of environmental noise, including frequency distribution, noise intensity and fluctuation range. Ensure that the feature definition can reflect the audio characteristics of different scenarios. Collect environmental noise data related to the scenario to ensure that the data covers the real noise characteristics in various scenarios. Perform multi-frequency component feature analysis on the collected environmental noise data to extract the noise frequency characteristics and intensity distribution of various scenarios. Use the power spectral density (PSD) analysis method to evaluate the changes in frequency components. Generate a scene feature analysis report to display the frequency components and noise intensity characteristics in different scenes to help identify the scene the user is in in real time. Store the scene feature analysis results in the database to ensure data integrity and traceability. Record the parameter settings and results during the analysis process to facilitate subsequent verification and optimization. Generate a visual chart to display the noise characteristics of various scenes to help quickly identify the user's current environmental noise conditions. Select a suitable personalized noise reduction demand analysis model, which can use a rule-based system or machine learning method (such as random forest or neural network) to extract user needs from feedback information.Determine the input features for analysis, including noise reduction scenarios, feedback satisfaction, and environmental noise characteristics, so as to accurately assess the user's noise reduction needs. Input the noise reduction feedback information of each scenario into the selected demand analysis model, analyze the user's noise reduction needs in different scenarios, and identify the user's preferences for noise reduction intensity and type. Record the personalized noise reduction needs of each scenario, generate a demand analysis report, and display the user's noise reduction demand characteristics in the current scenario.

[0032] In this embodiment, step S5 includes the following steps: Step S51: defining a multi-frequency audio signal; Step S52: performing multi-frequency propagation simulation on the three-dimensional ear canal structure model of the user according to the multi-frequency audio signal to generate multi-frequency audio simulation data; Step S53: performing different frequency sensitivity analysis on the multi-band audio simulation data to obtain different frequency sensitivities of the user's ear canal; Step S54: Calculate the optimal noise reduction intensity based on the different frequency sensitivities of the user's ear canal and the user's noise reduction requirements in the current scene to obtain the optimal noise reduction intensity for the current scene.

[0033] In this embodiment, the parameters of the multi-frequency audio signal are defined, including the frequency range, amplitude and phase. The selected frequency range should generally cover the frequency range to which the human ear is sensitive, generally 20 Hz to 20 kHz, with a focus on the mid-frequency band (about 300 Hz to 3 kHz) because this is the frequency range to which the human ear is most sensitive. The amplitude of the signal is determined, usually set at 70 dBSPL to ensure that the signal does not cause hearing damage and can effectively simulate the propagation in the ear canal. The required multi-frequency audio signal is generated using synthesis software (such as an audio processing library in MATLAB or Python). This can be achieved by superimposing multiple sine wave signals to ensure that the amplitude and phase of each frequency can be set independently. The generated signal should be a continuous audio sample with a set length of 1 second to 5 seconds to facilitate subsequent propagation simulation. The generated multi-frequency audio signal is stored in WAV or other audio formats to ensure that it can be easily read and used in subsequent simulations. The user's three-dimensional ear canal structure model constructed in the previous steps is used to ensure that the model accurately reflects the physical properties of the user's ear canal. The model should include the length, width, curvature of the ear canal and its connection with the outer ear and middle ear. Determine the material properties of the ear canal model, usually using materials with good acoustic properties, so as to accurately simulate the propagation of sound waves in the ear canal. Select an appropriate sound wave propagation simulation method, such as the finite element method (FEM) or the boundary element method (BEM). These methods can effectively simulate the propagation behavior of sound waves in complex geometric structures. Set simulation parameters, including frequency range, time step, and spatial grid size, to ensure the accuracy and stability of the simulation. Input the multi-frequency audio signal into the simulation software, run the sound wave propagation simulation, and record the propagation characteristics of the sound wave in the ear canal, including reflection, attenuation, and interference. Generate multi-frequency audio simulation data and record the sound pressure level and phase information of each frequency in the ear canal for subsequent analysis. Select an appropriate frequency sensitivity analysis method, such as energy ratio or gain analysis in audio signal processing, to evaluate the sensitivity of the user's ear canal to different frequencies. Determine the input parameters of the analysis, including the sound pressure level and phase information of the multi-band audio simulation data. Perform sensitivity analysis on the simulated multi-band audio data to calculate the degree of response of the user's ear canal to each frequency component, usually based on the change in sound pressure level. Record the sensitivity value of each frequency and generate a sensitivity analysis report to show the response characteristics of the user's ear canal at different frequencies. Select a suitable noise reduction strength calculation model, usually a linear regression model or a machine learning method (such as a support vector machine) to derive the optimal noise reduction strength from the ear canal sensitivity and scene requirements. Determine the input characteristics of the model, including the sensitivity values ​​of different frequencies and the noise characteristics of the current scene. Input the frequency sensitivity data of the user's ear canal and the noise reduction requirements of the current scene into the selected calculation model to calculate the optimal noise reduction strength.Record the calculation results and generate an optimal noise reduction intensity report, showing the noise reduction intensity and frequency range that need to be applied in the current scenario.

[0034] In this embodiment, step S6 includes the following steps: Step S61: performing dynamic noise frequency waveform fitting on the multi-dimensional environmental noise data, thereby generating a dynamic noise frequency waveform; Step S62: performing reverse audio noise reduction calculation based on the dynamic noise frequency waveform thus generated and the optimal noise reduction strength of the current scene to generate a reverse noise reduction waveform; Step S63: performing internal noise reduction processing in the ear canal according to the reverse noise reduction waveform to generate an internal noise reduction strategy; Step S64: Perform hybrid collaborative noise reduction optimization on the internal noise reduction strategy and the external environment noise reduction suppression strategy, thereby constructing a hybrid noise reduction optimization engine.

[0035] In this embodiment, multi-dimensional environmental noise data is collected, and these data should include noise intensity, frequency components and their changes in different time periods. To ensure the integrity and accuracy of the data, it is generally recommended to use a high sampling rate (such as 44.1 kHz) for data collection. The collected noise data is preprocessed, including removing noise and outliers, to facilitate subsequent waveform fitting. A filter (such as a low-pass filter) can be used to process the data to eliminate the interference of high-frequency noise. Select a suitable waveform fitting algorithm, usually using frequency domain analysis methods such as Fourier transform (FFT) or wavelet transform (Wavelet Transform). These methods can effectively capture the frequency components of dynamic noise and generate corresponding frequency waveforms. Determine the fitting parameters, including time window, frequency range and fitting accuracy, to ensure that the generated waveform can truly reflect the characteristics of dynamic noise. The preprocessed multi-dimensional environmental noise data is input into the selected fitting algorithm to fit the dynamic noise frequency waveform. Record the frequency component and amplitude information of each frequency band. Generate a dynamic noise frequency waveform and record the waveform characteristics of each time period for subsequent analysis and processing. Understand the basic principle of inverse noise reduction, which is to cancel noise by generating a sound wave (inverse waveform) opposite to the ambient noise waveform. The amplitude of the inverse waveform should be equal to the noise amplitude, but the phase is opposite. Determine the optimal noise reduction strength for the current scene, usually based on the results of the previous calculation, and this strength will be used to generate the inverse waveform. For the generated dynamic noise frequency waveform, use its amplitude and phase information to calculate the inverse audio noise reduction waveform. The specific method is: ,in, is a dynamic noise waveform. Record the amplitude and phase characteristics of the generated reverse waveform to ensure that the original noise can be effectively covered when implementing noise reduction. Compare the generated reverse waveform with the dynamic noise frequency waveform to confirm the effectiveness and noise reduction ability of the reverse waveform. The noise reduction effect can be evaluated using indicators such as signal-to-noise ratio (SNR). Generate a reverse noise reduction waveform and record the waveform characteristics of each frequency band to ensure that it can be effectively applied in the subsequent noise reduction processing inside the ear canal. Understand the basic principle of noise reduction processing inside the ear canal, that is, the generated reverse noise reduction waveform is transmitted through headphones or ear canals to form sound wave interference against environmental noise. Determine the acoustic characteristics of the ear canal, including the length, curvature and material properties of the ear canal, in order to optimize the noise reduction effect. Select a suitable noise reduction processing method, usually adaptive filtering technology (such as LMS algorithm) or digital signal processing technology, to process the reverse noise reduction waveform in real time. Determine the processing parameters, including the order of the filter, learning rate and delay time, to ensure the real-time and accuracy of the noise reduction processing. Input the generated reverse noise reduction waveform into the noise reduction processing system inside the ear canal to generate a noise reduction waveform for environmental noise in real time. Ensure that the processing system can respond to the user's ear canal conditions and environmental changes in real time. Record key parameter settings and results during the processing process for subsequent verification and optimization. Store the results of the internal ear canal noise reduction processing in a database to ensure data integrity and traceability. Record the evaluation data of the noise reduction processing effect for subsequent analysis. Generate an evaluation report of the noise reduction effect, showing the audio signal characteristics before and after noise reduction, to help understand the effectiveness of the internal ear canal noise reduction strategy. Understand the basic principles of hybrid collaborative noise reduction, that is, combining the internal ear canal noise reduction strategy with the external environment noise reduction suppression strategy to achieve enhanced noise reduction effect. Internal noise reduction is mainly aimed at the noise in the ear canal, while external noise reduction is aimed at the noise in the external environment. Determine the optimization goals, including maximizing the noise reduction effect, improving user comfort, and ensuring the clarity of the audio signal. Select a suitable collaborative optimization algorithm, such as genetic algorithm or particle swarm optimization (PSO), to find a balance between internal and external noise reduction. Determine the input parameters for optimization, including the current effect of internal and external noise reduction, user preferences, and environmental noise characteristics. Input the effects of the internal and external noise reduction strategies into the selected optimization algorithm for collaborative optimization calculation. Record key parameter settings and results during the optimization process to ensure that the final strategy can meet the user's noise reduction needs. Generate the output of the hybrid noise reduction optimization engine to determine the combination of noise reduction settings and strategies. Store the optimization results in the database to ensure data integrity and traceability. Record the implementation effect of the hybrid noise reduction strategy for subsequent verification and optimization. Provide a feedback mechanism to allow users to adjust the settings of the headphones based on the optimization results, thereby optimizing the overall noise reduction effect and improving the user's listening experience.

[0036] In this embodiment, a hybrid noise reduction system for a Bluetooth headset is provided, which is used to execute the hybrid noise reduction method for a Bluetooth headset as described above, including: A noise component analysis module is used to obtain multi-dimensional environmental noise data based on the environmental noise sensor and perform noise component feature analysis to generate multi-frequency component features of environmental noise; The filtering and suppression module is used to perform intelligent noise reduction intensity and adaptive filtering noise reduction suppression on the multi-frequency component characteristics of environmental noise, thereby generating an external environment noise reduction suppression strategy; The ear canal structure module is used to send tiny audio signals based on the ear canal morphology sensor and collect audio signal feedback information; perform three-dimensional ear canal structure modeling based on the audio signal feedback information, thereby constructing a three-dimensional ear canal structure model of the user; The noise reduction demand module is used to obtain the wearer's noise reduction feedback log and perform personalized noise reduction demand analysis to obtain the user's noise reduction demand in the current scene; The optimal noise reduction calculation module is used to analyze the sensitivity of the user's three-dimensional ear canal structure model to different frequencies, and calculate the optimal noise reduction intensity based on the user's noise reduction requirements in the current scene to obtain the optimal noise reduction intensity in the current scene; The hybrid collaborative noise reduction module is used to perform reverse audio noise reduction processing according to the optimal noise reduction intensity of the current scene, and to perform hybrid collaborative noise reduction optimization according to the external environment noise reduction suppression strategy, thereby building a hybrid noise reduction optimization engine.

[0037] The present invention can more comprehensively understand the spatial and frequency distribution of noise through multi-dimensional environmental noise data. This process provides accurate data support for subsequent noise classification and dynamic adjustment of noise reduction intensity. By decomposing the frequency components of noise, different types of noise, such as low-frequency vibration noise or high-frequency sharp noise, can be identified, thereby providing targeted processing strategies for noise reduction algorithms. The noise reduction intensity is adjusted according to the changes in real-time environmental noise to ensure that a stronger noise reduction effect is provided when the noise intensity is high, and excessive noise reduction is avoided when the noise is weak, thereby improving the sound quality. Adaptive filtering technology can dynamically adjust the filtering parameters according to the characteristics of different environmental noises, thereby effectively suppressing unwanted noise while maintaining the naturalness and clarity of the sound. By obtaining the three-dimensional structure of the ear canal, the system can more accurately understand the reflection and attenuation characteristics of the wearer's ear canal to sound, thereby customizing noise reduction and audio enhancement strategies to ensure that each wearer can get a good sound experience. Through the modeling of the ear canal structure, the earphone can better adjust the sound output to avoid sound distortion or discomfort caused by differences in the shape of the ear canal. Intelligent learning is carried out according to the wearer's noise reduction feedback log, and the noise reduction strategy is continuously optimized to provide a personalized noise reduction experience. This means that the needs of each wearer will be tracked and accurately met over the long term. The wearer's noise reduction needs will change over time and environment, and the system can intelligently adjust based on historical data and real-time feedback to ensure that the noise reduction effect always meets the wearer's needs. By analyzing the ear canal structure and frequency sensitivity, the system can accurately identify the wearer's sensitivity in a specific frequency band, and then provide more personalized noise reduction processing. According to the individual differences and scene requirements of the wearer, the optimal noise reduction intensity is dynamically calculated to ensure that good noise reduction effects can be provided in various environments, while avoiding excessive noise reduction and loss of sound quality. By combining reverse noise reduction technology with external noise reduction strategies, noise reduction effects can be exerted simultaneously inside the ear canal and in the external environment, providing more comprehensive and refined noise suppression. The hybrid noise reduction optimization engine can make intelligent adjustments based on the wearer's needs and environmental changes to achieve the most optimized noise reduction effect. This collaborative noise reduction mechanism makes the noise reduction processing not only accurate and effective, but also natural and comfortable, avoiding the discomfort caused by excessive noise reduction.

[0038] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0039] The above is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A hybrid noise reduction method for a Bluetooth headset, characterized in that: The Bluetooth headset includes an environmental noise sensor and an ear canal morphology sensor, and includes the following steps: Step S1: acquiring multi-dimensional environmental noise data based on an environmental noise sensor, and performing noise component feature analysis to generate multi-frequency component features of the environmental noise; Step S2: performing intelligent noise reduction intensity and adaptive filtering noise reduction suppression on the multi-frequency component characteristics of the environmental noise, thereby generating an external environment noise reduction suppression strategy; Step S3: sending a tiny audio signal based on the ear canal morphology sensor and collecting audio signal feedback information; performing three-dimensional ear canal structure modeling according to the audio signal feedback information, thereby constructing a three-dimensional ear canal structure model of the user; Step S4: Obtain the wearer's noise reduction feedback log and perform personalized noise reduction demand analysis to obtain the user's noise reduction demand in the current scene; Step S5: Analyze the sensitivity of different frequencies to the user's three-dimensional ear canal structure model, and calculate the optimal noise reduction strength based on the user's noise reduction requirements in the current scene to obtain the optimal noise reduction strength in the current scene; Step S6: Perform reverse audio noise reduction processing according to the optimal noise reduction strength of the current scene, and perform hybrid collaborative noise reduction optimization according to the external environment noise reduction suppression strategy, thereby building a hybrid noise reduction optimization engine.

2. The hybrid noise reduction method for Bluetooth headset according to claim 1, characterized in that: The specific steps of step S1 are: Step S11: acquiring multi-dimensional environmental noise data based on the environmental noise sensor; Step S12: performing multi-scale time-frequency decomposition on the multi-dimensional environmental noise data to generate multiple environmental noise frequency sub-bands; Step S13: Calculating the frequency components of the multiple ambient noise frequency sub-bands one by one to generate a sub-frequency component of each sub-band; Step S14: Calculate the power spectrum density of multiple ambient noise frequency sub-bands to generate power spectrum density values; Step S15: performing noise component feature analysis according to the power spectrum density value and the sub-frequency components of each sub-band, thereby generating multi-frequency component features of the environmental noise.

3. The hybrid noise reduction method for Bluetooth headset according to claim 1, characterized in that: The specific steps of step S2 are: Step S21: performing noise type cluster analysis on the multi-frequency component characteristics of the environmental noise, thereby generating characteristic types of noise in the environment; Step S22: identifying dynamic noise type changes of noise feature types in the environment to obtain dynamic type change data of environmental noise; Step S23: performing intelligent noise reduction intensity calculation based on the dynamic type change data of the environmental noise to generate dynamic noise reduction intensity parameters; Step S24: Adaptively filter and suppress noise reduction according to the dynamic noise reduction strength parameter, thereby generating an external environment noise reduction suppression strategy.

4. The hybrid noise reduction method for Bluetooth headset according to claim 3, characterized in that: The specific steps of step S21 are: Calculate the noise intensity at multiple time points based on the multi-frequency component characteristics of environmental noise and extract the noise intensity values ​​at different time points; Perform real-time noise intensity change analysis based on noise intensity values ​​at different time points to generate real-time noise intensity change data; Perform time series discrete change fitting on real-time noise intensity change data to construct a time series noise intensity change curve; The frequency fluctuation range of the multi-frequency component characteristics of the environmental noise is quantified to obtain the frequency fluctuation range of the environmental noise; Perform noise duration statistics on the multi-frequency component characteristics of the environmental noise to generate the noise duration; Noise type clustering analysis is performed on the noise duration, ambient noise frequency fluctuation range and time-series noise intensity change curve to generate noise feature types in the environment, which include continuous noise, impact noise and dragging noise.

5. The hybrid noise reduction method for Bluetooth headset according to claim 1, characterized in that: The specific steps of step S3 are: Step S31: sending a tiny audio signal based on the ear canal morphology sensor and collecting audio signal feedback information; Step S32: calculating the ear canal reflection intensity of the audio signal feedback information to obtain the audio reflection intensity; Step S33: performing signal attenuation rate analysis on the audio signal feedback information to generate a signal attenuation rate; Step S34: Calculate the reflection time according to the audio signal feedback information to obtain the audio reflection time; Step S35: performing audio reflection feature evolution on the audio reflection intensity, the signal attenuation rate and the audio reflection time, thereby generating an audio reflection evolution feature; Step S36: Perform three-dimensional ear canal structure modeling on the audio reflection evolution characteristics, thereby constructing a three-dimensional ear canal structure model of the user.

6. The method according to claim 1, characterized in that The specific steps of step S4 are: Step S41: Obtaining the wearer's noise reduction feedback log; Step S42: identifying multiple historical headphone noise reduction scenarios based on the wearer's noise reduction feedback log; Step S43: performing noise reduction feedback analysis on multiple historical headphone noise reduction scenarios to extract noise reduction feedback information for each scenario; Step S44: performing scene feature analysis on the multi-frequency component features of the ambient noise to obtain the real-time scene of the user; Step S45: Perform personalized noise reduction demand analysis on the scene where the user is in real time according to the noise reduction feedback information of each scene, so as to obtain the noise reduction demand of the user in the current scene.

7. The hybrid noise reduction method for Bluetooth headset according to claim 1, characterized in that: The specific steps of step S5 are: Step S51: defining a multi-frequency audio signal; Step S52: performing multi-frequency propagation simulation on the three-dimensional ear canal structure model of the user according to the multi-frequency audio signal to generate multi-frequency audio simulation data; Step S53: performing different frequency sensitivity analysis on the multi-band audio simulation data to obtain different frequency sensitivities of the user's ear canal; Step S54: Calculate the optimal noise reduction intensity based on the different frequency sensitivities of the user's ear canal and the user's noise reduction requirements in the current scene to obtain the optimal noise reduction intensity for the current scene.

8. The hybrid noise reduction method for Bluetooth headset according to claim 1, characterized in that: The specific steps of step S6 are: Step S61: performing dynamic noise frequency waveform fitting on the multi-dimensional environmental noise data, thereby generating a dynamic noise frequency waveform; Step S62: performing reverse audio noise reduction calculation based on the dynamic noise frequency waveform thus generated and the optimal noise reduction strength of the current scene to generate a reverse noise reduction waveform; Step S63: performing internal noise reduction processing in the ear canal according to the reverse noise reduction waveform to generate an internal noise reduction strategy; Step S64: Perform hybrid collaborative noise reduction optimization on the internal noise reduction strategy and the external environment noise reduction suppression strategy, thereby constructing a hybrid noise reduction optimization engine.

9. A hybrid noise reduction system for a Bluetooth headset, characterized in that: The hybrid noise reduction method for executing the Bluetooth headset according to claim 1 comprises: A noise component analysis module is used to obtain multi-dimensional environmental noise data based on the environmental noise sensor and perform noise component feature analysis to generate multi-frequency component features of environmental noise; The filtering and suppression module is used to perform intelligent noise reduction intensity and adaptive filtering noise reduction suppression on the multi-frequency component characteristics of environmental noise, thereby generating an external environment noise reduction suppression strategy; The ear canal structure module is used to send tiny audio signals based on the ear canal morphology sensor and collect audio signal feedback information; perform three-dimensional ear canal structure modeling based on the audio signal feedback information, thereby constructing a three-dimensional ear canal structure model of the user; The noise reduction demand module is used to obtain the wearer's noise reduction feedback log and perform personalized noise reduction demand analysis to obtain the user's noise reduction demand in the current scene; The optimal noise reduction calculation module is used to analyze the sensitivity of the user's three-dimensional ear canal structure model to different frequencies, and calculate the optimal noise reduction intensity based on the user's noise reduction requirements in the current scene to obtain the optimal noise reduction intensity in the current scene; The hybrid collaborative noise reduction module is used to perform reverse audio noise reduction processing according to the optimal noise reduction intensity of the current scene, and to perform hybrid collaborative noise reduction optimization according to the external environment noise reduction suppression strategy, thereby building a hybrid noise reduction optimization engine.

Images