Search

CN-121985246-A - Earphone hybrid noise reduction control system with multi-mode sensor fusion

CN121985246ACN 121985246 ACN121985246 ACN 121985246ACN-121985246-A

Abstract

The invention discloses a multimode sensor fused earphone mixed noise reduction control system, which comprises a multimode information acquisition module, a data processing module, a scene recognition and noise processing module and a control feedback module. According to the multi-mode sensor fused earphone mixed noise reduction control system, the multi-mode sensor fusion is adopted, the information of the dual-microphone array, the wearing detection sensor and the motion sensing sensor is combined, the comprehensive perception of environmental noise, wearing state and user motion state is realized, the accuracy and environmental adaptability of noise detection are improved, the problem that a single sensor cannot cope with a complex scene is solved, the dynamic weight adjustment of active noise reduction and passive noise reduction is realized, and the quality of the earphone mixed noise reduction is improved by introducing the directional noise reduction.

Inventors

  • ZHANG XUANJUN

Assignees

  • 安徽芯宇驰半导体科技有限公司

Dates

Publication Date
20260505
Application Date
20260120

Claims (10)

  1. 1. A multimode sensor fused earphone mixed noise reduction control system is characterized by comprising: The multi-mode information acquisition module adopts a multi-type sensor to acquire environmental noise and parameter information used by a user in real time; The data processing module is used for preprocessing the acquired data, establishing a correlation model between the environmental sound intensity and a user hearing safety threshold value and evaluating the hearing state of the user; The scene recognition and noise processing module is used for preferentially determining the wearing state of the earphone by a user, judging the motion state after wearing based on the gesture data, fusing the characteristics of the wearing state parameters and the motion state parameters, and performing synchronous operations of active noise reduction, passive noise reduction and directional noise reduction based on the condition of the fused characteristics; And the control feedback module is used for generating a noise processing instruction, controlling an output audio signal based on the noise processing instruction, dynamically adjusting parameters based on a user noise reduction feedback log and an electromyographic signal, and realizing personalized noise reduction.
  2. 2. The system for controlling the mixed noise reduction of the earphone fused by the multi-modal sensor as set forth in claim 1, wherein the multi-modal sensor of the multi-modal information acquisition module comprises: The microphone array is characterized in that a single-side earphone is integrated with four microphones and adopts the four microphone array, wherein a feedforward microphone is positioned at the outer side of the earphone and used for collecting environmental noise in real time, and a feedback microphone is positioned at the inner side of an ear canal and used for monitoring the actual noise reduction effect and is converted into digital signals through an ADC (analog-to-digital converter) for subsequent processing; the bone conduction vibration sensor is used for capturing voice signals conducted by the jawbone and the fundamental frequency of sound band vibration; The three-axis acceleration sensor is combined with the three-axis gyroscope and is used for collecting the motion parameters of the head of the user; The wearing state of the earphone is automatically identified by adopting a capacitive proximity sensor and a pressure sensor.
  3. 3. The multi-modal sensor fused headphones hybrid noise reduction control system of claim 1, wherein the data processing module evaluates the user's hearing status as: Filtering, denoising and standardizing signals acquired by each sensor, eliminating random interference in the signals of a noise acquisition unit by adopting a Kalman filtering algorithm, eliminating high-frequency jitter in the signals of a motion sensing unit by adopting a moving average filtering method, and standardizing the signals to a [0,1] interval by adopting the standardizing process to ensure that the data are consistent; And then establishing a correlation model between the environmental sound intensity and the user hearing safety threshold by using a deep learning algorithm, and judging the current environmental noise to output a dynamic safety sound pressure threshold curve by means of transfer learning of a large amount of environmental sound data and user hearing data.
  4. 4. A multi-modal sensor fused headphones hybrid noise reduction control system as defined in claim 3, wherein: the forming operation of the safe sound pressure threshold curve comprises the following steps: extracting environmental sound parameters under the historical data, and extracting the situation of the hearing sound pressure of the user under the sequential time based on the same environmental sound parameters; After extracting the hearing sound pressure values according to the sequence time, when the hearing sound pressure values are determined to meet the condition that the hearing sound pressure values are not changed in the set time interval, extracting the hearing sound pressure values, and then averaging a plurality of hearing sound pressure values to obtain a safe sound pressure threshold under the current environment sound parameters; and taking the numerical value change of the environmental sound parameter from small to large as a parameter of a transverse coordinate axis, taking a safety sound pressure threshold value under the corresponding environmental sound parameter as a parameter of a vertical coordinate axis, and performing curve connection on parameter points positioned on the transverse coordinate axis and the vertical coordinate axis to obtain a safety sound pressure threshold value curve.
  5. 5. The multi-modal sensor fused headphones hybrid noise reduction control system of claim 2 wherein the scene recognition and noise processing module determines the user's wear status of headphones as: The wearing state is judged by extracting parameter data acquired by a capacitive proximity sensor and a pressure sensor and a threshold value set by a factory, wherein the capacitive proximity sensor is used for detecting that the attaching distance between an earphone and an ear of a user is a, and the pressure sensor is arranged at the edge of an earmuff and used for detecting that the attaching pressure between the earphone and the ear is b; and the distance threshold value set by a factory is A, the bonding pressure threshold value with the ear is B, and the specific result is as follows: when a < A, B is more than or equal to B, the result I is that the wearing state is stable; secondly, when a is more than or equal to A and B is less than B, the wearing state is loose; and thirdly, when a has no data and b=0, the state is not worn currently.
  6. 6. The multi-modal sensor fused headphones hybrid noise reduction control system of claim 2 wherein the scene recognition and noise processing module determines post-wear motion state based on gesture data to operate as: tracking the three-dimensional motion state of the head of the user in real time through a built-in three-axis gyroscope, and judging the direction and the speed of the head of the user according to data obtained by a three-axis acceleration sensor; taking the position of the sensor when the earphone is worn initially as an initial point, and extracting triaxial acceleration sensor data in the subsequent time to obtain the angular velocity c and the acceleration d of the change position of the earphone sensor at the initial point; obtaining angular velocity intervals [ m, n ] and acceleration intervals [ p, q ] under different motion states according to the record of the historical data; If c < m, d < p, and duration t, then the current user is in a stationary state, if , And when the duration is t, the current user is in a walking state, and when c > n, d > q and the duration is t, the current user is in a running state.
  7. 7. The multi-modal sensor fused headphones hybrid noise reduction control system of claim 1 wherein the scene recognition and noise processing module fuses the characteristics of the wearing state parameters and the motion state parameters to: The wearing state parameter and the movement state parameter data in a real-time state are extracted and compared with a set fitting distance threshold value and an ear fitting pressure threshold value of the user's ear, and whether the change occurs under different movement state parameters is determined; And timely generating an adjustment instruction for feedback based on the changed condition.
  8. 8. The system for controlling the mixed noise reduction of the multimode sensor fused earphone according to claim 2, wherein the synchronous operations of active noise reduction, passive noise reduction and directional noise reduction based on the condition of fusion characteristics in the scene recognition and noise processing module are as follows: The weight of active noise reduction and passive noise reduction is dynamically adjusted by extracting the characteristic parameters of the noise signals after characteristic fusion and combining the wearing state parameters and the motion state parameters; and simultaneously, the sound wave direction of the sound source is positioned based on the motion state parameters, and the reverse sound wave is emitted aiming at the specific direction, so that noise suppression is realized.
  9. 9. The system for controlling mixed noise reduction of a multimode sensor fused earphone of claim 7 wherein said adjusting the weight of active noise reduction and passive noise reduction is: If the wearing state is a stable wearing state, comparing the amplitude of the noise signal with the set signal amplitude by extracting the amplitude of the noise signal; If the amplitude of the extracted noise signal is smaller than the set signal amplitude, setting the active noise reduction weight as u 1 and the passive noise reduction weight as And (2) and If the amplitude of the extracted noise signal is larger than the set signal amplitude, setting the active noise reduction weight as The passive noise reduction weight is And (2) and ; If the wearing state is a loose wearing state, setting the active noise reduction weight as The passive noise reduction weight is And (2) and ; If the wearing state is not the wearing state, the active noise reduction and passive noise reduction functions are automatically closed, so that the energy consumption is reduced.
  10. 10. The multi-modal sensor fused headphones hybrid noise reduction control system of claim 8 wherein the sound-wave direction localization based on motion state parameters operates to: Determining arrival time stamps of the same noise signals received by all microphones in the four microphone arrays, and positioning the sound wave direction of the sound source according to the sequence of the time stamps; And the change of the sound source direction is updated by combining the change of the motion state, so that the noise reduction signal opposite to the sound wave direction of the sound source is generated for cancellation.

Description

Earphone hybrid noise reduction control system with multi-mode sensor fusion Technical Field The invention relates to the technical field of earphone noise reduction, in particular to an earphone mixed noise reduction control system with a multi-mode sensor fusion function. Background With the increasing demands of people on audio experience, earphone noise reduction technology is becoming more and more important. Conventional noise reduction headphones mainly employ a single active noise reduction (ANC) or passive noise reduction (PNC) technique. The method and the system for noise reduction of the earphone based on the multiple noises are disclosed by the reference patent names (patent publication number: CN119893362A, patent publication date: 2025-04-25), and comprise the steps that a multi-array microphone collects effective voice signals of users and signals of different noise sources, a noise reduction mode control system realizes a motion mode, a watching video mode, a conversation mode and an audio loading mode through a user terminal interface, and real state switching among a standard noise reduction mode, a comfortable noise reduction mode and an enhanced noise reduction mode, a wireless earphone noise reduction system, a user terminal earphone noise reduction system and a data computing power cloud service center adopt a three-stage framework of a cloud side, noise signals in the mixed voice signals are removed through a built-in intelligent voice noise reduction method, user experience is improved, a composite information transmission system combining wired and wireless is designed through the cloud transmission framework, long-distance and high-capacity efficient information transmission is realized, and the voice enhancement system performs final processing on the noise signals after noise reduction through the built-in intelligent voice enhancement method, mixed voice and echo interference are eliminated, and high-fidelity voice quality sound signals are output. Based on the description of the above document, the existing earphone realizes noise processing through a single sensor, does not combine information of other sensors to perform comprehensive judgment, and fails to fully consider the dynamic change of environmental noise, the difference of wearing states of the earphone and the influence of the motion state of a user on the noise reduction effect, so that the accuracy of noise detection and the environmental adaptability are poor. Disclosure of Invention Aiming at the defects of the prior art, the invention provides a multimode sensor fused earphone mixed noise reduction control system, which solves the problems that the existing earphone realizes noise treatment through a single sensor, does not combine information of other sensors to carry out comprehensive judgment, and cannot fully consider the dynamic change of environmental noise, the difference of wearing states of the earphone and the influence of the motion state of a user on the noise reduction effect, so that the noise detection precision and the environmental adaptability are poor. In order to achieve the purpose, the invention is realized by the following technical scheme that the earphone mixed noise reduction control system with the multi-mode sensor fusion comprises: The multi-mode information acquisition module adopts a multi-type sensor to acquire environmental noise and parameter information used by a user in real time; The data processing module is used for preprocessing the acquired data, establishing a correlation model between the environmental sound intensity and a user hearing safety threshold value and evaluating the hearing state of the user; The scene recognition and noise processing module is used for preferentially determining the wearing state of the earphone by a user, judging the motion state after wearing based on the gesture data, fusing the characteristics of the wearing state parameters and the motion state parameters, and performing synchronous operations of active noise reduction, passive noise reduction and directional noise reduction based on the condition of the fused characteristics; And the control feedback module is used for generating a noise processing instruction, controlling an output audio signal based on the noise processing instruction, dynamically adjusting parameters based on a user noise reduction feedback log and an electromyographic signal, and realizing personalized noise reduction. Preferably, the multi-type sensor of the multi-mode information acquisition module includes: The microphone array is characterized in that a single-side earphone is integrated with four microphones and adopts the four microphone array, wherein a feedforward microphone is positioned at the outer side of the earphone and used for collecting environmental noise in real time, and a feedback microphone is positioned at the inner side of an ear canal and used for monitoring the actual noise reduction eff