A conduction device fitting management system and a conduction device

Abstract

The application discloses a kind of conduction device fitting management system and conduction device, belong to data analysis technical field, it specifically includes: acquisition multi-source fitting data and real-time fusion, combine the regional fitting strategy defined based on geographic location or environment type, construct scene-physiology-wear three-dimensional adaptive model;According to real-time fusion data and model data, dynamically allocate three types of data priority and generate initial compensation parameter, input initial compensation parameter into three-order LSTM acoustic compensation engine containing attention mechanism, combine the hearing audio signal collected by real-time hearing test strategy, output acoustic compensation signal;Acoustic compensation signal is converted into hearing aid DSP executable instruction, drives hearing aid compensation and outputs gain curve, while collecting user feedback;The system can adaptively adjust compensation parameters according to user real-time state and environmental changes, improve hearing aid compensation effect.

Description

Technical Field

[0001] This invention belongs to the field of data analysis technology, specifically a transmission device fitting and management system and a transmission device. Background Technology

[0002] With the continuous development of technology, acoustic devices are playing an increasingly important role in people's lives, especially in fields such as hearing aids and communication. The main function of acoustic devices is to convert sound signals into electrical signals and transmit them to the user through vibration, enabling the user to perceive sound. However, in practical applications, the fitting and acoustic compensation effects of acoustic devices directly affect the user experience.

[0003] Traditional methods of fitting conductive devices typically rely on professionals testing and adjusting the device in a specific environment. This approach has several limitations. First, the testing environment is often relatively simple and cannot simulate the various complex scenarios that users may encounter in real life, such as noisy public places or quiet indoor environments. Therefore, fitting based on the results of testing in a single environment may not meet the user's needs in different scenarios during actual use, resulting in significant differences in the sound perceived by the user in different environments.

[0004] Secondly, traditional fitting methods primarily focus on physiological indicators such as the degree of hearing loss, neglecting changes in the user's physiological state and the wear and tear on the acoustic device itself. A user's physiological state changes over time and with various factors, such as fatigue and emotional fluctuations, which can affect their perception of sound. Simultaneously, acoustic devices gradually experience component wear and aging during use, impacting performance and acoustic compensation effectiveness. Traditional fitting methods cannot adapt to these changes in a timely manner, limiting the accuracy and effectiveness of fitting results. Therefore, developing a acoustic device fitting management system that comprehensively considers multi-source fitting data and enables dynamic compensation parameter generation and real-time feedback is of significant practical importance. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention proposes a hearing aid fitting management system and a hearing aid device. It collects and fuses multi-source fitting data in real time, and combines this with a regionalized fitting strategy defined by geographical location or environmental type to construct a three-dimensional adaptive model of scene-physiology-loss. Based on the real-time fused data and model data, it dynamically allocates the priorities of three types of data and generates initial compensation parameters. These initial compensation parameters are input into a third-order LSTM acoustic compensation engine with an attention mechanism, and combined with hearing audio signals collected by a real-time hearing test strategy, outputting an acoustic compensation signal. The acoustic compensation signal is converted into hearing aid DSP executable instructions to drive the hearing aid compensation and output a gain curve, while simultaneously collecting user feedback. This system can adaptively adjust the compensation parameters according to the user's real-time state and environmental changes, improving the hearing aid compensation effect.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A transmission device fitting and management system includes: a data acquisition module, a three-dimensional modeling module, a compensation parameter generation module, an acoustic compensation module, and an execution and feedback module;

[0008] The 3D modeling module performs real-time fusion of multi-source fitting data collected by the data acquisition module, and combines it with a predefined regional fitting strategy to construct a scene-physiology-weariness 3D adaptive model.

[0009] The compensation parameter generation module dynamically allocates the priority of the three types of data based on the real-time fused multi-source fitting data and the scenario-physiology-loss three-dimensional adaptive model data, and generates initial compensation parameters.

[0010] The initial compensation parameters are input into the third-order LSTM acoustic compensation engine configured in the acoustic compensation module. Combined with the audio signal collected according to the predefined real-time hearing test strategy, the acoustic compensation signal is output and transmitted to the execution and feedback module.

[0011] The execution and feedback module converts the acoustic compensation signal into executable instructions for the main control chip of the conduction device, drives the conduction device to perform acoustic compensation, and outputs the gain curve in real time. At the same time, it collects user feedback information on the current acoustic compensation effect.

[0012] Specifically, the 3D modeling module performs real-time fusion of multi-source fitting data collected by the data acquisition module, and combines it with a predefined regional fitting strategy to construct a scene-physiology-weariness 3D adaptive model, including:

[0013] A1: Collect multi-source fitting data; the multi-source fitting data includes environmental acoustic characteristic data, physiological index data, and motion data;

[0014] A2: Define a regional fitting strategy; the defined regional fitting strategy includes defining a set of environmental scene categories, a set of physiological state categories, and a set of loss characteristics;

[0015] A3: A three-dimensional strategy matrix is ​​established using environmental scene category, physiological state category, and loss characteristics as three dimensions; each element in the three-dimensional strategy matrix corresponds to a scene-physiology-loss combination, which is used to store fitting strategy information under the combination; the fitting strategy information includes gain adjustment parameters and noise reduction parameters.

[0016] A4: Assign initial weights to fitting strategies under different scenarios, physiological conditions, and wear-out combinations;

[0017] A5: For environmental acoustic feature data, physiological index data and motion data, Kalman filter models are established respectively, and the Kalman filter prediction results from different data sources are weighted and fused to obtain fused time series data;

[0018] A6: For the questionable portion in the fused time series data, a basic probability assignment function is defined for each fitting data source to map the fitting data to different propositions and assign corresponding probability values. At the same time, the basic probability assignment functions of different fitting data sources are synthesized using the synthesis rules of DS evidence theory to obtain high-confidence fused data. The questionable portion refers to measurement errors and individual differences in physiological indicator data.

[0019] A7: Feature extraction is performed on the high-confidence fusion data to obtain feature vectors of different feature dimensions. The importance weight of each feature dimension is calculated through the attention mechanism, and the final attention fusion fitting data is obtained after weighted fusion.

[0020] Specifically, the construction of the scenario-physiology-loss three-dimensional adaptive model also includes:

[0021] A8: Using machine learning algorithms, an environmental scene recognition sub-model is constructed by taking environmental acoustic feature data and motion data from the attention fusion fitting data as input, and a physiological state monitoring sub-model is constructed by taking physiological indicator data and environmental acoustic feature data from the attention fusion fitting data as input.

[0022] A9: By analyzing equipment usage data and periodic calibration results, and combining them with current input data, a loss characteristic analysis sub-model is established;

[0023] A10: The output results of the environmental scene recognition sub-model, the physiological state monitoring sub-model, and the loss feature analysis sub-model are mapped to a three-dimensional feature space to form a scene-physiology-loss three-dimensional adaptive model; in the three-dimensional feature space, each point represents a specific scene-physiology-loss combination.

[0024] Specifically, the compensation parameter generation module dynamically allocates the priorities of the three types of data based on the real-time fused multi-source fitting data and the scene-physiology-loss three-dimensional adaptive model data, and generates initial compensation parameters, including:

[0025] B1: Preprocess the multi-source fitting data after real-time fusion; the preprocessing includes data normalization, outlier removal, and missing value imputation.

[0026] B2: Based on the three-dimensional adaptive model of scenario-physiology-loss, determine the reference range of gain adjustment parameters and noise reduction parameters in the fitting strategy information corresponding to the current scenario-physiology-loss combination;

[0027] B3: Assign initial priority weights to the environmental acoustic feature data based on the noise type, noise intensity, and output results of the environmental scene recognition sub-model in the environmental acoustic feature data;

[0028] B4: Assign initial priority weights to the physiological indicator data based on the degree of hearing loss, the trend of hearing change, and the output results of the physiological state monitoring sub-model in the physiological indicator data;

[0029] B5: Combining the output of the sub-model with the analysis results of motion intensity, motion type and loss characteristics in the motion data, assign initial priority weights to the motion data;

[0030] B6: Construct a comprehensive priority evaluation model based on fuzzy logic. Input the initial priority weights of environmental acoustic feature data, physiological index data, and motion data into the comprehensive priority evaluation model. Take into account the mutual influence between the three types of data to obtain the final dynamic priority weights.

[0031] B7: Based on the final dynamic priority weight, the multi-source fitting data after real-time fusion is weighted to obtain weighted fusion data;

[0032] B8: Using weighted fused data as input, and combining the reference ranges of the corresponding gain adjustment parameters and noise reduction parameters in the scene-physiology-loss three-dimensional adaptive model, a genetic algorithm is used for optimization search to generate initial compensation parameters; the initial compensation parameters include initial gain parameters and initial noise reduction parameters.

[0033] Specifically, when assigning initial priority weights to environmental acoustic feature data in B3, the following is included:

[0034] B3.1: Establish a mapping table between noise type and priority weight, with different initial priority weight adjustment values ​​corresponding to different noise types;

[0035] B3.2: The initial priority weights are dynamically adjusted based on the comparison between the noise intensity and the preset noise threshold.

[0036] When the noise intensity exceeds the preset noise threshold, the priority weight of the environmental acoustic feature data is increased according to the preset proportional coefficient.

[0037] B3.3: Based on the output of the environmental scene recognition sub-model, if the identified environmental scene is an environmental sound interest scene, the priority weight of the environmental acoustic feature data is increased by a dynamic scaling factor; the environmental sound interest scene refers to a scene where environmental sound needs to be considered.

[0038] Specifically, the initial compensation parameters are input into the third-order LSTM acoustic compensation engine configured within the acoustic compensation module, and combined with the audio signals acquired according to a predefined real-time hearing test strategy, an acoustic compensation signal is output, including:

[0039] C1: Audio signals are acquired under different scenarios, physiological conditions, and hearing loss combinations according to a predefined real-time hearing test strategy; the audio signals include the original audio signals and the audio signals after initial compensation.

[0040] The predefined real-time hearing test strategy is as follows: every 500 milliseconds, a set of test audio signals containing different frequencies and intensities is input into the conduction device, and the output audio signal processed by the conduction device is collected at the same time. The difference between the input and output audio signals is compared to obtain the current acoustic performance data of the conduction device.

[0041] C2: Preprocess the acquired audio signal;

[0042] C3: Input the preprocessed audio signal into the third-order LSTM acoustic compensation engine; the third-order LSTM acoustic compensation engine includes three LSTM layers, each of which is used to extract features of the audio signal at different time scales;

[0043] C4: In the third-order LSTM acoustic compensation engine, the initial compensation parameters are used as input information and fused with the audio signal features to obtain the fused audio signal features.

[0044] C5: The third-order LSTM acoustic compensation engine calculates the preliminary results of the acoustic compensation signal through forward propagation based on the characteristics of the input fused audio signal.

[0045] C6: Post-processing the preliminary results of the acoustic compensation signal; the post-processing includes amplitude limiting and smoothing operations;

[0046] C7: Outputs the post-processed acoustic compensation signal.

[0047] Specifically, the parameter settings for each LSTM layer of the third-order LSTM acoustic compensation engine in C3 are as follows:

[0048] C3.1: The first layer of the LSTM is used to extract short-term features of the audio signal, and the number of hidden layer nodes is set to a;

[0049] C3.2: The second layer of the LSTM is used to extract the mid-term features of the audio signal, and the number of hidden layer nodes is set to b;

[0050] C3.3: The third layer of the LSTM is used to extract long-term features of the audio signal. The number of hidden layer nodes is set to c, and the following conditions are met: ;

[0051] C3.4: Each LSTM layer uses dropout technology to randomly discard a preset number of neurons.

[0052] Specifically, the execution and feedback module converts the acoustic compensation signal into executable instructions for the main control chip of the conduction device, drives the conduction device to perform acoustic compensation, and outputs the gain curve in real time. Simultaneously, it collects user feedback on the current acoustic compensation effect, including:

[0053] D1: Based on the instruction format requirements of the main control chip of the conduction device, convert the acoustic compensation signal into executable instructions for the main control chip of the conduction device;

[0054] The process of converting the acoustic compensation signal into executable instructions for the main control chip of the conduction device is as follows: the digital acoustic compensation signal is converted into an analog signal through a digital-to-analog converter chip, and then the analog signal is encoded into digital code that conforms to the instruction format of the main control chip using a communication protocol and transmitted to the main control chip.

[0055] D2: Send the executable instructions from the main control chip to the main control chip of the conduction device to drive the acoustic compensation module of the conduction device to perform acoustic compensation operation;

[0056] D3: During the acoustic compensation process, the output signal of the conduction device is acquired in real time, and the gain curve is calculated based on the output signal;

[0057] D4: Display the current acoustic compensation effect to the user through a user interface or wireless communication, and collect user feedback on the acoustic compensation effect; the feedback includes user evaluations of sound clarity, comfort, and volume.

[0058] D5: Quantify user feedback information and convert it into numerical indicators.

[0059] Specifically, the calculation of the gain curve in D3 includes:

[0060] D3.1: Determine the frequency range and number of sampling points for the gain curve; the frequency range covers the range of audio frequencies that the conduction device can process;

[0061] D3.2: At each sampling point, calculate the amplitude ratio of the output signal to the input signal of the conduction device to obtain the gain value at the sound wave frequency corresponding to the current sampling point;

[0062] D3.3: Connect the gain values ​​of all sampling points to form a gain curve, and then smooth the gain curve.

[0063] A conductive device includes a conductive device body, a data acquisition sensor group, a main control chip, an acoustic compensation module, and a user interface;

[0064] The main body of the conductive device is the main structure of the device, used for installing and fixing components;

[0065] The data acquisition sensor group is installed on the main body of the transmission device and is used to collect environmental acoustic feature data of the user's environment, physiological index data of the user, and motion data of the user, and transmit the collected data to the main control chip.

[0066] The main control chip receives data transmitted by the data acquisition sensor group and runs the conduction device fitting management system. Based on the real-time fused multi-source fitting data and the scene-physiology-loss three-dimensional adaptive model data, it dynamically generates compensation parameters and controls the acoustic compensation module to perform acoustic compensation.

[0067] The acoustic compensation module receives compensation parameters transmitted by the main control chip, performs compensation processing on the received audio signal, and outputs the compensated audio signal.

[0068] The user interface is located on the main body of the transmission device and is used to display the current acoustic compensation effect, receive user feedback on the acoustic compensation effect, and transmit the feedback information to the main control chip.

[0069] Compared with the prior art, the beneficial effects of the present invention are:

[0070] 1. This invention proposes a transmission device fitting and management system, and optimizes and improves its architecture, operation steps and processes. The system has the advantages of simple process, low investment and operating costs and low production and working costs.

[0071] 2. This invention proposes a sound conduction device fitting management system. A 3D modeling module fuses multi-source fitting data in real time and combines it with a regionalized fitting strategy to construct a scene-physiological-loss 3D adaptive model. A compensation parameter generation module dynamically allocates the priorities of three types of data based on this model and the real-time fused data, generating initial compensation parameters. This collaborative mechanism enables the sound conduction device to accurately adapt to different scenes, physiological states, and equipment loss conditions, breaking through the limitations of traditional fixed compensation modes and achieving personalized acoustic compensation. This effectively improves the accuracy and adaptability of acoustic compensation, ensuring that users can obtain a more tailored auditory experience in different environments.

[0072] 3. This invention proposes a sounding device fitting management system. Initial compensation parameters are input into a third-order LSTM acoustic compensation engine, which outputs an acoustic compensation signal by combining the audio signal acquired through a real-time hearing test strategy. The execution and feedback module then converts this signal into executable instructions to drive the sounding device for compensation. The system also outputs gain curves in real time and collects user feedback. This process not only achieves efficient generation and accurate execution of the acoustic compensation signal, but also allows for timely understanding of the user's perception of the compensation effect through a real-time feedback mechanism. This helps to continuously optimize compensation parameters and algorithms, further improving the acoustic compensation effect. Attached Figure Description

[0073] Figure 1 This is an architecture diagram of a transmission device fitting management system according to the present invention;

[0074] Figure 2 This is a flowchart illustrating the construction of a three-dimensional adaptive model of scenario-physiology-loss in a conduction device fitting and management system according to the present invention.

[0075] Figure 3 This is a flowchart illustrating the implementation of initial compensation parameters in a transmission device fitting management system according to the present invention. Detailed Implementation

[0076] Example 1

[0077] Please see Figure 1 The present invention provides an embodiment of a conductive device fitting management system, which is applicable to the field of hearing aid application. In the field of hearing aid application, loss refers to hearing loss and conductive device refers to hearing aid. The system includes: a data acquisition module, a three-dimensional modeling module, a compensation parameter generation module, an acoustic compensation module, and an execution and feedback module.

[0078] The 3D modeling module performs real-time fusion of the multi-source fitting data collected by the data acquisition module, and constructs a scene-physiology-weariness 3D adaptive model by combining it with a predefined regional fitting strategy; the multi-source fitting data includes environmental acoustic feature data, physiological index data, and motion data.

[0079] Among them, environmental acoustic characteristic data: using a high-precision microphone array, environmental sound signals are collected from multiple directions. The microphone array can acquire information such as sound intensity, frequency distribution, and phase. Through beamforming technology, the direction of the main noise source is determined, and the sound pressure level at different frequencies is calculated. For example, in an office setting, there are noise sources such as computer fans and printers. The microphone array can accurately measure the sound pressure level and frequency characteristics of these noises.

[0080] Physiological data: Wearable biosensors, such as heart rate monitors, pulse oximeters, and skin conductance sensors, are used to collect real-time physiological indicators such as heart rate, blood oxygen saturation, and skin conductance. These sensors can accurately measure changes in physiological parameters, reflecting the user's physiological state in different scenarios. For example, when a user is under stress or fatigue, their heart rate may increase, and their skin conductance may be enhanced.

[0081] Motion data: Using inertial sensors such as accelerometers and gyroscopes, information such as the user's motion acceleration and angular velocity is collected. Through integral algorithms, parameters such as the user's motion speed and displacement are calculated to determine the user's motion state, such as being stationary, walking, or running.

[0082] The scenario-physiology-loss three-dimensional adaptive model can dynamically reflect the optimal fitting parameters under different hearing scenarios, user physiological states, and hearing loss conditions.

[0083] It should be noted that different regions may have different acoustic environment characteristics and fitting needs. For example, in an office setting, more emphasis is placed on compensating for speech clarity, while in a noisy outdoor environment, the focus is on improving the overall signal-to-noise ratio. Therefore, different fitting strategies need to be defined according to different situations.

[0084] The compensation parameter generation module dynamically allocates the priority of the three types of data based on the real-time fused multi-source fitting data and the scenario-physiology-loss three-dimensional adaptive model data, and generates initial compensation parameters.

[0085] The priority allocation is dynamically adjusted based on the degree of influence of data on the hearing compensation effect;

[0086] The initial compensation parameters include, but are not limited to, gain, compression ratio, and noise reduction threshold.

[0087] For example, the impact of various types of data on compensation parameters varies depending on the scenario and user state. For instance, in a noisy environment, environmental acoustic feature data has a higher priority; when the user is under stress, physiological indicator data needs to be given higher priority.

[0088] The initial compensation parameters are input into the third-order LSTM acoustic compensation engine configured in the acoustic compensation module. Combined with the audio signal collected according to the predefined real-time hearing test strategy, the acoustic compensation signal is output and transmitted to the execution and feedback module.

[0089] The execution and feedback module converts the acoustic compensation signal into executable instructions for the main control chip of the conduction device, drives the conduction device to perform acoustic compensation, and outputs the gain curve in real time. At the same time, it collects user feedback information on the current acoustic compensation effect.

[0090] It should be noted that the execution and feedback module converts the acoustic compensation signal into actual executable instructions to drive the hearing aid to work, and transmits the feedback information back to the data acquisition module. This allows the system to continuously adjust and optimize the subsequent fitting process based on the user's actual experience, forming an adaptive closed-loop system that continuously improves the acoustic compensation effect of the hearing aid.

[0091] The data acquisition module includes: an acoustic sensor unit, a physiological index sensor unit, and a motion sensor unit;

[0092] Acoustic sensor unit, used to collect environmental acoustic feature data;

[0093] Physiological indicator sensor unit, used to collect physiological indicator data;

[0094] Motion sensor unit, used to collect motion data.

[0095] The 3D modeling module includes: a data fusion unit and a model building unit;

[0096] The data fusion unit is used to perform time synchronization and normalization processing on multi-source fitting data;

[0097] The model building unit is used to combine regional fitting strategies to build a three-dimensional adaptive model of scenario, physiology, and wear.

[0098] The compensation parameter generation module includes: a priority allocation unit and a parameter generation unit;

[0099] The priority allocation unit is used to dynamically allocate the priority of environmental acoustic feature data, physiological index data, and motion data.

[0100] The parameter generation unit is used to generate initial compensation parameters according to priority.

[0101] The acoustic compensation module includes: an engine configuration unit and a signal processing unit;

[0102] The engine configuration unit is used to configure the structure and parameters of the third-order LSTM acoustic compensation engine.

[0103] The signal processing unit is used to combine the audio signals acquired by the real-time hearing test strategy and output an acoustic compensation signal.

[0104] Example 2

[0105] Please see Figure 2 and Figure 3 In this embodiment, the 3D modeling module performs real-time fusion of multi-source fitting data collected by the data acquisition module, and constructs a scene-physiology-weariness 3D adaptive model by combining it with a predefined regional fitting strategy, including:

[0106] A1: Collect multi-source fitting data; the multi-source fitting data includes environmental acoustic characteristic data, physiological index data, and motion data;

[0107] Furthermore, the method for acquiring the environmental acoustic feature data is as follows: multiple microphone arrays of different types are arranged at different positions of the conduction device to acquire environmental sound signals of different directions and frequency bands. The time-domain sound signals are converted into frequency-domain signals through Fourier transform, and environmental acoustic feature data including environmental noise intensity, main noise frequency components, and reverberation time are extracted. Here, Fourier transform is a prior art in this field and is not an inventive solution of this application, so it will not be described in detail here.

[0108] Furthermore, the physiological index data includes ear canal shape data, eardrum thickness data, and auditory nerve response data; the ear canal shape data is obtained through three-dimensional laser scanning technology, the eardrum thickness data is collected using ultrasound measurement technology, and the auditory nerve response data is recorded under specific audio stimulation using an electroencephalogram (EEG) signal acquisition device.

[0109] Furthermore, the motion data is collected in real time using a nine-axis inertial sensor to acquire the acceleration, angular velocity, and magnetic field strength data of the wearer of the conduction device. The sensor data is fused and processed by a complementary filtering algorithm to extract the wearer's motion state, motion amplitude, motion direction, and other motion data.

[0110] A2: Define a regional fitting strategy; the defined regional fitting strategy includes defining a set of environmental scene categories, a set of physiological state categories, and a set of loss characteristics;

[0111] Define a set of environmental scene categories, such as quiet, noisy, and music scenes;

[0112] Define a set of physiological state categories, such as relaxation, tension, and exercise.

[0113] Define a set of loss characteristics, such as frequency response curves and sensitivity thresholds.

[0114] A3: A three-dimensional strategy matrix is ​​established using environmental scene category, physiological state category, and loss characteristics as three dimensions; each element in the three-dimensional strategy matrix corresponds to a scene-physiology-loss combination, which is used to store fitting strategy information under the combination; the fitting strategy information includes gain adjustment parameters and noise reduction parameters.

[0115] Furthermore, the three-dimensional strategy matrix is ​​stored using a hash table data structure, with the scenario-physiology-loss combination as the hash key and the fitting strategy information as the hash value. When performing data queries and calls, the corresponding fitting strategy information can be quickly located through the hash function, thereby improving the system's operating efficiency.

[0116] A4: Assign initial weights to fitting strategies under different scenarios, physiological conditions, and wear-out combinations;

[0117] A5: For environmental acoustic feature data, physiological index data and motion data, Kalman filter models are established respectively, and the Kalman filter prediction results from different data sources are weighted and fused to obtain fused time series data;

[0118] For example, taking environmental acoustic feature data as an example, parameters such as sound pressure level and frequency are used as state variables. Based on historical data and current measurement values, the Kalman filter algorithm is used for state estimation and prediction. By continuously updating the state estimate and covariance matrix, the prediction accuracy of data change trends is improved. The Kalman filter algorithm is existing technology in this field and is not an inventive solution of this application, so it will not be described in detail here.

[0119] A6: For the questionable portion in the fused time series data, a basic probability assignment function is defined for each fitting data source to map the fitting data to different propositions and assign corresponding probability values. At the same time, the basic probability assignment functions of different fitting data sources are synthesized using the synthesis rules of DS evidence theory to obtain high-confidence fused data. The questionable portion refers to measurement errors and individual differences in physiological indicator data.

[0120] For example, heart rate data can be mapped to propositions such as calm, tense, and fatigue, and corresponding probabilities can be assigned.

[0121] A7: Feature extraction is performed on the high-confidence fusion data to obtain feature vectors of different feature dimensions. The importance weight of each feature dimension is calculated through the attention mechanism, and the final attention fusion fitting data is obtained after weighted fusion.

[0122] A8: Using machine learning algorithms, an environmental scene recognition sub-model is constructed by taking environmental acoustic feature data and motion data from the attention fusion fitting data as input, and a physiological state monitoring sub-model is constructed by taking physiological indicator data and environmental acoustic feature data from the attention fusion fitting data as input.

[0123] A9: By analyzing equipment usage data and periodic calibration results, and combining them with current input data, a loss characteristic analysis sub-model is established;

[0124] Furthermore, the environmental scene recognition sub-model automatically identifies the current environmental scene category based on the characteristics of the input fitting data and outputs it; the physiological state monitoring sub-model monitors the user's physiological state in real time and classifies it into the corresponding physiological state category; the loss feature analysis sub-model estimates the device's loss characteristics, such as the degree of high-frequency loss and the degree of low-frequency loss, and outputs them.

[0125] A10: The output results of the environmental scene recognition sub-model, the physiological state monitoring sub-model, and the loss feature analysis sub-model are mapped to a three-dimensional feature space to form a scene-physiology-loss three-dimensional adaptive model; in the three-dimensional feature space, each point represents a specific scene-physiology-loss combination.

[0126] The compensation parameter generation module dynamically allocates the priority of the three types of data based on the real-time fused multi-source fitting data and the scene-physiology-loss three-dimensional adaptive model data, and generates initial compensation parameters, including:

[0127] B1: Preprocess the multi-source fitting data after real-time fusion; the preprocessing includes data normalization, outlier removal, and missing value imputation.

[0128] B2: Based on the three-dimensional adaptive model of scenario-physiology-loss, determine the reference range of gain adjustment parameters and noise reduction parameters in the fitting strategy information corresponding to the current scenario-physiology-loss combination;

[0129] B3: Assign initial priority weights to the environmental acoustic feature data based on the noise type, noise intensity, and output results of the environmental scene recognition sub-model in the environmental acoustic feature data; for example, when the noise intensity exceeds the preset noise threshold and the environmental scene is identified as a noisy public place, increase the priority weight of the environmental acoustic feature data.

[0130] Furthermore, the setting of the noise threshold needs to be considered in conjunction with the specific application scenario, equipment performance, and human perception characteristics. It is typically determined using a combination of a baseline reference value and dynamic adjustment based on the scenario, specifically including:

[0131] (1) Setting the reference threshold based on acoustic standards

[0132] Based on internationally accepted acoustic standards and combined with human subjective perception of different noise intensities, its unit is decibel dB(A), with A-weighting being more consistent with human auditory perception.

[0133] Quiet environment benchmark: In scenarios such as residences and offices, the preset basic threshold is usually 40-50dB(A). When the noise intensity is less than or equal to this preset basic threshold, it is judged as low noise, and the environmental acoustic data has a lower priority.

[0134] Noisy environment benchmark: In scenarios such as streets and shopping malls, the preset benchmark threshold is usually 60-70dB(A). When the noise intensity is greater than this preset benchmark threshold, it is judged as medium to high noise, and the priority of environmental acoustic data needs to be increased.

[0135] (2) Dynamic adjustment based on scene type

[0136] Based on the baseline threshold, the threshold is further refined according to the scene category output by the environmental scene recognition sub-model:

[0137] In quiet environments such as libraries and bedrooms: the threshold is reduced to 30-40 dB(A). Even low noise, such as a whisper at 35 dB(A), can affect the experience, so acoustic data should be given higher priority.

[0138] In noisy public places, such as subways and concerts: the threshold is increased to 70-85 dB(A), and the background noise in such scenarios is relatively high;

[0139] For dynamic scenarios, such as driving or walking: Set a threshold range, such as 50-75dB(A), and adjust it in real time according to noise fluctuations in the scenario. For example, when the vehicle speed increases, the threshold will increase by 10-15dB(A).

[0140] B4: Based on the degree of hearing loss, the trend of hearing change, and the output of the physiological state monitoring sub-model in the physiological indicator data, assign initial priority weights to the physiological indicator data; for example, when the degree of hearing loss is severe and the physiological state monitoring indicates a state of fatigue, appropriately increase the priority weight of the physiological indicator data.

[0141] Furthermore, initial priority weights are assigned to the physiological indicator data, including:

[0142] B4.1: Establish a functional relationship between the degree of hearing loss and the priority weight. The more severe the hearing loss, the higher the corresponding initial priority weight.

[0143] B4.2: Based on the trend of hearing changes, if hearing declines in a short period of time, the priority weight of physiological indicator data will be increased.

[0144] B4.3: Based on the output of the physiological state monitoring sub-model, if the user is detected to be in a state of tension, anxiety or other adverse physiological state that affects hearing perception, the priority weight of the physiological indicator data will be increased.

[0145] B5: Assign initial priority weights to the motion data by combining the output of the motion intensity, motion type, and wear characteristic analysis sub-model in the motion data; for example, when the motion intensity is high and the wear characteristic analysis shows that the equipment wear is aggravated, increase the priority weight of the motion data;

[0146] Furthermore, initial priority weights are assigned to the motion data, including:

[0147] B5.1: Establish the correspondence between exercise intensity and priority weight. The greater the exercise intensity, the higher the corresponding initial priority weight.

[0148] B5.2: Based on the type of exercise, if the type of exercise is a high-intensity vibration exercise, such as running or jumping, then increase the priority weight of the exercise data;

[0149] B5.3: Combine the output results of the loss characteristic analysis sub-model. If the equipment loss is high and related to motion, further increase the priority weight of motion data.

[0150] B6: Construct a comprehensive priority evaluation model based on fuzzy logic. Input the initial priority weights of environmental acoustic feature data, physiological index data, and motion data into the comprehensive priority evaluation model. Take into account the mutual influence between the three types of data to obtain the final dynamic priority weights.

[0151] Furthermore, when constructing a comprehensive priority evaluation model based on fuzzy logic, the following steps are included:

[0152] B6.1: Determine the initial priority weights of the fuzzy input variables, namely, environmental acoustic feature data, physiological index data, and motion data;

[0153] B6.2: Define fuzzy sets, and define different fuzzy sets for each fuzzy input variable, such as low priority, medium priority, and high priority;

[0154] B6.3: Establish a fuzzy rule base and formulate fuzzy rules based on actual application scenarios and expert experience to describe the relationship between different combinations of fuzzy input variables and the final priority weights;

[0155] B6.4: A fuzzy inference method is adopted to perform inference based on fuzzy input variables and a fuzzy rule base to obtain fuzzy output results. The fuzzy inference method is prior art in this field and is not an inventive solution of this application, so it will not be described in detail here.

[0156] B6.5: Defuzzify the fuzzy output results to obtain the final dynamic priority weights.

[0157] B7: Based on the final dynamic priority weight, the multi-source fitting data after real-time fusion is weighted to obtain weighted fusion data;

[0158] B8: Using weighted fused data as input, and combining the reference ranges of the corresponding gain adjustment parameters and noise reduction parameters in the scene-physiology-loss three-dimensional adaptive model, a genetic algorithm is used for optimization search to generate initial compensation parameters; the initial compensation parameters include initial gain parameters and initial noise reduction parameters.

[0159] Furthermore, a genetic algorithm is used for optimization search to generate initial compensation parameters, including:

[0160] B8.1: Determine the encoding method of the genetic algorithm, and encode the initial gain parameters and initial noise reduction parameters into chromosomes;

[0161] B8.2: Define the fitness function, which comprehensively considers the degree of matching of the compensation parameters with environmental acoustic feature data, physiological index data and motion data, as well as the degree of closeness to the reference range in the scene-physiology-loss three-dimensional adaptive model;

[0162] B8.3: Set the parameters of the genetic algorithm, including population size, crossover probability, and mutation probability. The genetic algorithm is prior art in this field and is not an inventive solution of this application, so it will not be described in detail here.

[0163] B8.4: Perform population initialization operations, randomly generating a predetermined number of chromosomes as the initial population;

[0164] B8.5: Perform selection, crossover, and mutation operations, select individuals in the population according to the fitness function, and generate new individuals through crossover and mutation operations. The selection, crossover, and mutation operations are prior art in this field and are not the inventive solution of this application, and will not be described in detail here.

[0165] B8.6: Repeat the selection, crossover, and mutation operations until the preset termination conditions are met, such as reaching the preset maximum number of iterations or the fitness function value reaching the preset threshold, and output the optimal chromosome, i.e. the initial compensation parameters.

[0166] When assigning initial priority weights to environmental acoustic feature data in B3, the following is included:

[0167] B3.1: Establish a mapping table between noise type and priority weight. Different types of noise correspond to different initial priority weight adjustment values. For example, sudden high-frequency noise corresponds to a larger initial priority weight adjustment value.

[0168] B3.2: The initial priority weights are dynamically adjusted based on the comparison between the noise intensity and the preset noise threshold.

[0169] When the noise intensity exceeds the preset noise threshold, the priority weight of the environmental acoustic feature data is increased according to the preset proportional coefficient.

[0170] Furthermore, the preset scaling factor is used to quantify the increase in priority weight of environmental acoustic feature data when the noise intensity exceeds the threshold. Its setting needs to balance response sensitivity and system stability, and is usually determined by combining the degree of noise exceeding the standard or scene characteristics. The specific method is as follows:

[0171] (1) Stepped proportional coefficient based on the noise exceedance amplitude:

[0172] The noise intensity is divided into intervals based on the difference between the noise intensity and the preset threshold. A fixed proportional coefficient is assigned to each interval to avoid over-adjustment or under-adjustment caused by a single coefficient. Specifically, the proportional coefficient is set to 1.2-1.5 when the noise intensity exceeds the limit slightly (ΔdB≤10dB), 1.5-2.0 when the noise intensity exceeds the limit moderately (10dB<ΔdB≤20dB), and 2.0-3.0 when the noise intensity exceeds the limit severely (ΔdB>20dB).

[0173] (2) Combine the dynamic scaling factor of scene category

[0174] Different scenarios have different noise tolerances, requiring specific proportional coefficients to be matched for each scenario. This avoids unreasonable adjustments caused by uniform changes across scenarios. For example: in quiet scenarios, even if the noise level is slightly exceeded, the proportional coefficient is set to 1.5-2.0 because quiet environments are more sensitive to noise and acoustic data needs to be prioritized; in noisy public places, when the noise level is moderately exceeded, the proportional coefficient is set to 1.2-1.5 because the background noise of the scene is high; in dynamic scenarios, the coefficient is adjusted according to the intensity of activity within the scene. For example, when the vehicle speed is >60km / h, the proportional coefficient is reduced by 10%-20% to avoid frequent noise fluctuations caused by vehicle bumps triggering weight adjustments.

[0175] B3.3: Based on the output of the environmental scene recognition sub-model, if the identified environmental scene is an environmental sound concern scene, the priority weight of the environmental acoustic feature data is increased by a dynamic scaling factor; the environmental sound concern scene refers to a scene that requires high attention to environmental sound, including but not limited to the examination site.

[0176] The initial compensation parameters are input into the third-order LSTM acoustic compensation engine configured within the acoustic compensation module. Combined with the hearing audio signals acquired according to a predefined real-time hearing test strategy, the acoustic compensation signal is output, including:

[0177] C1: Audio signals are acquired under different scenarios, physiological conditions, and hearing loss combinations according to a predefined real-time hearing test strategy; the audio signals include the original audio signals and the audio signals after initial compensation.

[0178] The predefined real-time hearing test strategy is as follows: every 500 milliseconds, a set of test audio signals containing different frequencies and intensities is input into the conduction device, and the output audio signal processed by the conduction device is collected at the same time. The difference between the input and output audio signals is compared to obtain the current acoustic performance data of the conduction device.

[0179] C2: Preprocess the acquired audio signal; preprocessing includes operations such as pre-emphasis, framing, and windowing to improve the quality of the audio signal and the accuracy of feature extraction;

[0180] C3: Input the preprocessed audio signal into the third-order LSTM acoustic compensation engine; the third-order LSTM acoustic compensation engine includes three LSTM layers, each of which is used to extract features of the audio signal at different time scales;

[0181] Furthermore, the internal structure of the third-order LSTM acoustic compensation engine includes three layers of long short-term memory network, with each LSTM layer containing 256 memory units.

[0182] C4: In the third-order LSTM acoustic compensation engine, the initial compensation parameters are used as input information and fused with the audio signal features to obtain the fused audio signal features.

[0183] Furthermore, when fusing the initial compensation parameters as input information with audio signal features, the process includes:

[0184] C4.1: Normalize the initial compensation parameters to match the range of audio signal characteristics;

[0185] C4.2: The normalized initial compensation parameters are concatenated with the output features of each LSTM layer to form a new fused feature vector;

[0186] C4.3: During the splicing process, different weights are assigned to the initial compensation parameters and audio signal features to adjust their importance in the fused feature vector, thus obtaining the fused audio signal features.

[0187] C5: The third-order LSTM acoustic compensation engine calculates the preliminary results of the acoustic compensation signal through forward propagation based on the characteristics of the input fused audio signal.

[0188] C6: Post-processing the preliminary results of the acoustic compensation signal; the post-processing includes amplitude limiting and smoothing operations to avoid distortion or abrupt changes in the acoustic compensation signal;

[0189] C7: Outputs the post-processed acoustic compensation signal.

[0190] The parameter settings for each LSTM layer of the third-order LSTM acoustic compensation engine in C3 are as follows:

[0191] C3.1: The first layer of the LSTM is used to extract short-term features of the audio signal. The number of hidden layer nodes is set to 'a' to capture local changes in the audio signal.

[0192] C3.2: The second layer of the LSTM is used to extract the mid-term features of the audio signal. The number of hidden layer nodes is set to b, taking into account both short-term and long-term features.

[0193] C3.3: The third layer of the LSTM is used to extract long-term features of the audio signal. The number of hidden layer nodes is set to c, and the following conditions are met: To capture the overall trend of the audio signal;

[0194] C3.4: Each LSTM layer uses dropout technology to randomly discard a predetermined number of neurons to prevent overfitting.

[0195] The execution and feedback module converts the acoustic compensation signal into executable instructions for the main control chip of the conduction device, drives the conduction device to perform acoustic compensation, and outputs the gain curve in real time. Simultaneously, it collects user feedback on the current acoustic compensation effect, including:

[0196] D1: Based on the instruction format requirements of the main control chip of the conduction device, convert the acoustic compensation signal into executable instructions for the main control chip of the conduction device;

[0197] The process of converting the acoustic compensation signal into executable instructions for the main control chip of the conduction device is as follows: the digital acoustic compensation signal is converted into an analog signal through a digital-to-analog converter chip, and then the analog signal is encoded into digital code that conforms to the instruction format of the main control chip using a communication protocol and transmitted to the main control chip.

[0198] D2: Send the executable instructions from the main control chip to the main control chip of the conduction device to drive the acoustic compensation module of the conduction device to perform acoustic compensation operation;

[0199] D3: During the acoustic compensation process, the output signal of the conduction device is acquired in real time, and the gain curve is calculated based on the output signal. The gain curve reflects the gain change at different frequencies.

[0200] D4: Display the current acoustic compensation effect to the user through a user interface or wireless communication, and collect user feedback on the acoustic compensation effect; the feedback includes user evaluations of sound clarity, comfort, and volume.

[0201] Furthermore, collecting user feedback on the acoustic compensation effect includes setting up star rating and text evaluation input boxes in the system, as well as equipping it with voice feedback function. Users can provide feedback on their satisfaction with the current acoustic compensation effect and suggestions for improvement by clicking on the star rating, entering text or voice description.

[0202] D5: Quantify user feedback information and convert it into numerical indicators that can be used for subsequent analysis and optimization.

[0203] Furthermore, user feedback information is quantified, including:

[0204] D5.1: Assign a corresponding numerical score to each feedback rating option. For example, a five-star rating corresponds to 5 points, and a one-star rating corresponds to 1 point.

[0205] D5.2: Calculate the average score or weighted average score of each feedback indicator based on the evaluation options selected by the user;

[0206] D5.3: Store the quantified feedback information in the database for subsequent data analysis and system optimization.

[0207] The calculation of the gain curve in D3 includes:

[0208] D3.1: Determine the frequency range and number of sampling points for the gain curve; the frequency range covers the range of audio frequencies that the conduction device can process;

[0209] D3.2: At each sampling point, calculate the amplitude ratio of the output signal to the input signal of the conduction device to obtain the gain value at the sound wave frequency corresponding to the current sampling point;

[0210] Furthermore, in acoustic or signal processing scenarios, the performance analysis of conduction devices usually needs to cover a preset frequency range, such as 20Hz-20kHz for human hearing. During the analysis, multiple sampling points are selected at preset intervals, and each sampling point corresponds to a specific frequency value, such as 100Hz, 500Hz, 1kHz, etc.

[0211] D3.3: Connect the gain values ​​of all sampling points to form a gain curve, and smooth the gain curve to reduce noise interference.

[0212] Example 3

[0213] A conductive device includes a conductive device body, a data acquisition sensor group, a main control chip, an acoustic compensation module, and a user interface;

[0214] The main body of the conductive device is the main structure of the device, used for installing and fixing components;

[0215] The data acquisition sensor group is installed on the main body of the transmission device and is used to collect environmental acoustic feature data of the user's environment, physiological index data of the user, and motion data of the user, and transmit the collected data to the main control chip.

[0216] The main control chip receives data transmitted by the data acquisition sensor group and runs the conduction device fitting management system. Based on the real-time fused multi-source fitting data and the scene-physiology-loss three-dimensional adaptive model data, it dynamically generates compensation parameters and controls the acoustic compensation module to perform acoustic compensation.

[0217] The acoustic compensation module receives compensation parameters transmitted by the main control chip, performs compensation processing on the received audio signal, and outputs the compensated audio signal.

[0218] The user interface is located on the main body of the transmission device and is used to display the current acoustic compensation effect, receive user feedback on the acoustic compensation effect, and transmit the feedback information to the main control chip.

[0219] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments under the guidance of the present invention without departing from the spirit and scope of the present invention. All of these variations are within the protection scope of the present invention.

[0220] If the technical solution disclosed herein involves personal information, the product using this technical solution has clearly informed the user of the personal information processing rules and obtained the user's voluntary consent before processing the personal information. If the technical solution disclosed herein involves sensitive personal information, the product using this technical solution has obtained the user's separate consent before processing the sensitive personal information, and also meets the requirement of "express consent". For example, at personal information collection devices such as cameras, clear and prominent signs are set up to inform users that they have entered the scope of personal information collection and that personal information will be collected. If an individual voluntarily enters the collection scope, it is deemed that they have agreed to the collection of their personal information; or on the personal information processing device, with clear signs / information informing users of the personal information processing rules, authorization is obtained from the individual through pop-up information or by asking the individual to upload their personal information; wherein, the personal information processing rules may include information such as the personal information processor, the purpose of personal information processing, the processing method, and the types of personal information processed.

Claims

1. A conductive device fitting management system characterized by, Comprise: Data acquisition module, three-dimensional modeling module, compensation parameter generation module, acoustic compensation module, execution and feedback module; The three-dimensional modeling module fuses the multi-source fitting data collected by the data acquisition module in real time, combines the pre-defined regional fitting strategy, and constructs a scene-physiology-loss three-dimensional adaptive model; The compensation parameter generation module dynamically allocs the priority of the three types of data according to the real-time fused multi-source fitting data and the scene-physiology-loss three-dimensional adaptive model data, and generates initial compensation parameters; The initial compensation parameters are input into the third-order LSTM acoustic compensation engine configured in the acoustic compensation module, combined with the audio signals collected according to the pre-defined real-time hearing test strategy, to output acoustic compensation signals and transmit them to the execution and feedback module; The execution and feedback module converts the acoustic compensation signal into a conductive device main control chip executable instruction, drives the conductive device to perform acoustic compensation, and outputs the gain curve in real time, while collecting user feedback information on the current acoustic compensation effect; The construction of the scene-physiology-loss three-dimensional adaptive model comprises: A1: Collect multi-source fitting data; the multi-source fitting data includes environmental acoustic feature data, physiological index data, and motion data; A2: Define regional fitting strategy; the defined regional fitting strategy includes defining environmental scene category set, physiological state category set, and loss feature set; A3: Establish a three-dimensional strategy matrix with environmental scene category, physiological state category and loss feature as three dimensions; each element in the three-dimensional strategy matrix corresponds to a scene-physiology-loss combination, which is used to store the fitting strategy information under the combination; the fitting strategy information includes gain adjustment parameters and noise reduction parameters; A4: Assign initial weights to the fitting strategies under different scene-physiology-loss combinations; A5: For environmental acoustic feature data, physiological index data and motion data, respectively establish Kalman filter model, and weight the Kalman filter prediction results of different data sources to obtain fused time series data; A6: For the suspicious part in the fused time series data, define a basic probability assignment function for each fitting data source, map the fitting data to different propositions, and assign corresponding probability values, and at the same time, use the combination rule of D-S evidence theory to combine the basic probability assignment functions of different fitting data sources to obtain high confidence fusion data; the suspicious part is the measurement error and individual difference in physiological index data; A7: Feature extraction is performed on the high confidence fusion data to obtain feature vectors of different feature dimensions, the importance weight of each feature dimension is calculated through attention mechanism, and the final attention fusion fitting data is obtained after weighted fusion; The construction of the scene-physiology-loss three-dimensional adaptive model further comprises: A8: Use machine learning algorithm to take environmental acoustic feature data and motion data in attention fusion fitting data as input to construct an environmental scene recognition sub-model, and take physiological index data and environmental acoustic feature data in attention fusion fitting data as input to construct a physiological state monitoring sub-model; A9: Establishing a loss feature analysis sub-model by analyzing the device usage data and the periodic calibration results, combined with the current input data; A10: Mapping the output results of the environmental scene recognition sub-model, the physiological state monitoring sub-model and the loss feature analysis sub-model into a three-dimensional feature space to form a scene-physiological loss three-dimensional adaptive model; in the three-dimensional feature space, each point represents a specific scene-physiological loss combination.

2. A fitting management system for a conductive device as defined in claim 1, wherein The compensation parameter generation module dynamically allocates the priority of the three types of data according to the real-time fused multi-source fitting data and the scene-physiological loss three-dimensional adaptive model data, and generates initial compensation parameters, including: B1: Preprocessing the real-time fused multi-source fitting data; the preprocessing includes data normalization, removing outliers and missing value filling; B2: Based on the scene-physiological loss three-dimensional adaptive model, determining the reference range of the gain adjustment parameter and the noise reduction parameter in the fitting strategy information corresponding to the current scene-physiological loss combination; B3: Assigning an initial priority weight to the environmental acoustic feature data according to the noise type, noise intensity in the environmental acoustic feature data and the output result of the environmental scene recognition sub-model; B4: Assigning an initial priority weight to the physiological index data according to the degree of hearing loss, hearing change trend in the physiological index data and the output result of the physiological state monitoring sub-model; B5: Assigning an initial priority weight to the motion data according to the motion intensity, motion type in the motion data and the output result of the loss feature analysis sub-model; B6: Building a comprehensive priority evaluation model based on fuzzy logic, inputting the initial priority weights of the environmental acoustic feature data, physiological index data and motion data into the comprehensive priority evaluation model, and comprehensively considering the mutual influence between the three types of data to obtain the final dynamic priority weight; B7: According to the final dynamic priority weight, performing weighted processing on the real-time fused multi-source fitting data to obtain weighted fusion data; B8: Taking the weighted fusion data as input, combining the reference range of the corresponding gain adjustment parameter and noise reduction parameter in the scene-physiological loss three-dimensional adaptive model, and using genetic algorithm for optimization search to generate initial compensation parameters; the initial compensation parameters include initial gain parameters and initial noise reduction parameters.

3. A fitting management system for a conductive device as defined in claim 2, wherein, The initial priority weight assignment for the environmental acoustic feature data includes: B3.1: Establishing a mapping relationship table between noise type and priority weight, different types of noise corresponding to different initial priority weight adjustment values; B3.2: Dynamically adjusting the initial priority weight according to the comparison result of the noise intensity and the preset noise threshold; When the noise intensity exceeds the preset noise threshold, the priority weight of the environmental acoustic feature data is increased by a preset proportion coefficient; B3.3: Combined with the output result of the environmental scene recognition sub-model, if the recognized environmental scene is an environmental sound attention scene, the priority weight of the environmental acoustic feature data is increased by a dynamic proportion coefficient; the environmental sound attention scene refers to a scene that needs to pay attention to environmental sound.

4. A fitting management system for a conductive device as defined in claim 3, wherein, The initial compensation parameter is input into a third-order LSTM acoustic compensation engine configured in the acoustic compensation module, combined with audio signals collected according to a pre-defined real-time hearing test strategy, to output an acoustic compensation signal, including: C1: Collecting audio signals under different scene-physiology-loss combinations according to a pre-defined real-time hearing test strategy; the audio signals include original audio signals and audio signals after initial compensation; The pre-defined real-time hearing test strategy is: every 500 milliseconds, input a set of test audio signals containing different frequencies and different intensities to the transmission device, collect the output audio signals processed by the transmission device, compare the differences between the input and output audio signals, and obtain the current acoustic performance data of the transmission device; C2: Preprocessing the collected audio signals; C3: Inputting the preprocessed audio signals into a third-order LSTM acoustic compensation engine; the third-order LSTM acoustic compensation engine includes three LSTM layers, and each LSTM layer is used to extract features of the audio signals in different time scales; C4: In the third-order LSTM acoustic compensation engine, the initial compensation parameter is taken as input information and fused with the audio signal features to obtain fused audio signal features; C5: The third-order LSTM acoustic compensation engine calculates the preliminary result of the acoustic compensation signal through forward propagation according to the input fused audio signal features; C6: Post-processing the preliminary result of the acoustic compensation signal; the post-processing includes amplitude limiting and smoothing operations; C7: Outputting the post-processed acoustic compensation signal.

5. A fitting management system for a conductive device as defined in claim 4, wherein, The parameter settings of each LSTM layer of the third-order LSTM acoustic compensation engine are as follows: C3.1: The first layer of LSTM is used to extract short-term features of the audio signal, and the number of hidden layer nodes is set to a ; C3.2: The second layer of LSTM is used to extract the mid-term features of the audio signal, and the number of hidden layer nodes is set to b ; C3.3: The third layer of LSTM is used to extract long-term features of the audio signal, and the number of hidden layer nodes is set to c , and satisfies ; C3.4: Each LSTM layer adopts a dropout technology to randomly discard a preset portion of neurons.

6. A fitting management system for a conductive device as defined in claim 5, wherein, The execution and feedback module converts the acoustic compensation signal into executable instructions of the transmission device master chip, drives the transmission device to perform acoustic compensation, and outputs the gain curve in real time, while collecting feedback information of the user on the current acoustic compensation effect, including: D1: Converting the acoustic compensation signal into executable instructions of the transmission device master chip according to the instruction format requirements of the transmission device master chip; The process of converting the acoustic compensation signal into executable instructions of the transmission device master chip is: converting the digital acoustic compensation signal into an analog signal through a digital-to-analog conversion chip, and then encoding the analog signal into a digital code conforming to the instruction format of the master chip using a communication protocol and transmitting it to the master chip; D2: Sending the executable instructions of the master chip to the transmission device master chip to drive the acoustic compensation module of the transmission device to perform acoustic compensation operations; D3: In the acoustic compensation process, the output signal of the transmission device is collected in real time, and the gain curve is calculated according to the output signal; D4: Through a user interface or wireless communication, the current acoustic compensation effect is displayed to the user, and feedback information of the user on the acoustic compensation effect is collected; the feedback information includes the user's evaluation on the sound clarity, comfort, and volume; D5: Quantitatively processing the user feedback information to convert it into numerical indicators.

7. A fitting management system for a conductive device as defined in claim 6, wherein, The calculation gain curve comprises: D3.1: determining the frequency range and the number of sampling points of the gain curve; the frequency range covers the audio frequency range that the conduction device can process; D3.2: at each sampling point, calculate the amplitude ratio of the conduction device output signal and the input signal to obtain the gain value at the sound frequency corresponding to the current sampling point; D3.3: connect the gain values of all sampling points to form a gain curve, and perform smoothing processing on the gain curve.

8. A conducting device, characterized by The conduction device comprises a conduction device body, a data acquisition sensor group, a master control chip, an acoustic compensation module, and a user interaction interface; The conduction device body is the main structure of the device, used for installing and fixing components; The data acquisition sensor group is arranged on the conduction device body and used for acquiring environmental acoustic characteristic data of the environment where the user is located, physiological index data of the user, and motion data of the user, and transmitting the acquired data to the master control chip; The master control chip receives the data transmitted by the data acquisition sensor group and runs the conduction device fitting management system, dynamically generates compensation parameters according to the real-time fused multi-source fitting data and the scene-physiology-loss three-dimensional adaptive model data, and controls the acoustic compensation module to perform acoustic compensation; The acoustic compensation module receives the compensation parameters transmitted by the master control chip, performs compensation processing on the received audio signal, and outputs the compensated audio signal; The user interaction interface is arranged on the conduction device body and used for displaying the current acoustic compensation effect and receiving feedback information of the user on the acoustic compensation effect, and transmitting the feedback information to the master control chip; A scene-physiology-loss three-dimensional adaptive model is constructed, comprising: A1: acquiring multi-source fitting data; the multi-source fitting data comprises environmental acoustic characteristic data, physiological index data, and motion data; A2: defining a regional fitting strategy; the defined regional fitting strategy comprises defining an environmental scene category set, a physiological state category set, and a loss characteristic set; A3: establishing a three-dimensional strategy matrix with the environmental scene category, the physiological state category, and the loss characteristic as three dimensions; each element in the three-dimensional strategy matrix corresponds to a scene-physiology-loss combination, and is used for storing fitting strategy information under the combination; the fitting strategy information comprises gain adjustment parameters and noise reduction parameters; A4: assigning initial weights to fitting strategies under different scene-physiology-loss combinations; A5: for the environmental acoustic characteristic data, the physiological index data, and the motion data, respectively establishing Kalman filter models, and performing weighted fusion on the Kalman filter prediction results of different data sources to obtain fused time series data; A6: defining a basic probability assignment function for each fitting data source for the suspicious part in the fused time series data, mapping the fitting data to different propositions, and assigning corresponding probability values, and simultaneously synthesizing the basic probability assignment functions of different fitting data sources by using the synthesis rule of D-S evidence theory to obtain high-confidence fused data; the suspicious part is measurement error and individual difference in the physiological index data; A7: Feature extraction is performed on the high-confidence fusion data to obtain feature vectors of different feature dimensions, importance weights of each feature dimension are calculated through an attention mechanism, and finally, attention fusion verification data are obtained after weighted fusion; The constructed scene-physiology-wear three-dimensional adaptive model further comprises: A8: An environmental scene recognition sub-model is constructed by taking environmental acoustic feature data and motion data in the attention fusion verification data as inputs, and a physiological state monitoring sub-model is constructed by taking physiological index data and environmental acoustic feature data in the attention fusion verification data as inputs; A9: A wear characteristic analysis sub-model is established by analyzing device usage data and periodic calibration results, combined with current input data; A10: The output results of the environmental scene recognition sub-model, the physiological state monitoring sub-model and the wear characteristic analysis sub-model are mapped into a three-dimensional feature space to form a scene-physiology-wear three-dimensional adaptive model; in the three-dimensional feature space, each point represents a specific scene-physiology-wear combination.

Images