A vehicle-mounted intelligent sound field regulation method based on brain waves and surface acoustic waves

Abstract

The application provides a vehicle-mounted intelligent sound field regulation method based on brain waves and surface acoustic waves, comprising: acquiring an acoustic signal in the vehicle and a brain electrical signal of a driver, wherein the acoustic signal is collected by a surface acoustic wave sensor, and the brain electrical signal is acquired by a brain electrical collection device; processing the acoustic signal, extracting mechanical vibration features related to vocal cord vibration, and distinguishing between a vocal signal and a non-vocal signal based on the mechanical vibration features; processing the brain electrical signal, extracting brain electrical features representing the attention state and / or physiological state of the driver; performing multi-modal fusion analysis on the acoustic feature information corresponding to the vocal signal and the brain electrical features, generating a sound field regulation strategy and / or a safety response strategy corresponding to the state of the driver; adjusting the vehicle-mounted audio output according to the sound field regulation strategy to enhance the vocal signal and / or suppress the non-vocal signal; and / or triggering a corresponding safety response operation according to the safety response strategy.

Description

Technical Field

[0001] This invention relates to the field of human-computer interaction technology for intelligent vehicles, and more specifically, to an in-vehicle intelligent sound field control method based on electroencephalogram (EEG) and surface acoustic waves (SAW). Background Technology

[0002] With the continuous development of intelligent vehicles and human-machine interaction technologies, the importance of in-vehicle audio systems in improving driving safety and user experience is becoming increasingly prominent. In particular, achieving effective separation and intelligent control of human voices and music in complex in-vehicle environments has become a key technological requirement. Currently, in-vehicle sound field control mainly relies on multi-microphone arrays combined with beamforming algorithms for sound source separation, or on simple human-machine control based on driver biosignals. Some solutions also attempt to introduce physical sensors or provide a manual adjustment interface. However, existing technologies have significant shortcomings: on the one hand, traditional microphone solutions are easily affected by environmental noise and electromagnetic interference, making it difficult to achieve stable human voice recognition in mixed sound fields such as music, navigation voice, and wind noise; on the other hand, existing biosignal-based solutions are mostly used only for control command generation, lacking a linkage mechanism with acoustic processing, and cannot dynamically optimize the sound field according to the driver's attention or physiological state.

[0003] The problem is that existing technologies struggle to effectively separate human voices from music in complex in-vehicle environments, and it is also difficult to combine driver status with intelligent sound field control and safety response. Summary of the Invention

[0004] This invention solves the technical problems of existing technologies in achieving effective separation of human voices and music in complex in-vehicle environments, and in achieving intelligent sound field control and safety response in conjunction with the driver's state.

[0005] To address the aforementioned problems, this invention provides an in-vehicle intelligent sound field control method based on electroencephalograms (EEGs) and surface acoustic waves (SAWs), comprising: acquiring in-vehicle acoustic signals and the driver's EEG signals, wherein the acoustic signals are acquired using a surface acoustic wave (SAW) sensor, and the EEG signals are acquired using an EEG acquisition device; processing the acoustic signals to extract mechanical vibration features related to vocal cord vibration, and distinguishing between human voice signals and non-human voice signals based on the mechanical vibration features; processing the EEG signals to extract EEG features characterizing the driver's attention state and / or physiological state; performing multimodal fusion analysis based on the acoustic feature information and EEG features corresponding to the human voice signals to generate a sound field control strategy and / or a safety response strategy corresponding to the driver's state; adjusting the in-vehicle audio output according to the sound field control strategy to enhance the human voice signals and / or suppress the non-human voice signals; and / or triggering corresponding safety response operations according to the safety response strategy.

[0006] Compared with existing technologies, the technical effects achieved by this solution are as follows: By introducing a surface acoustic wave (SAW) sensor to directly collect and identify the mechanical vibration characteristics generated by vocal cord vibration, the extraction of human voice signals no longer relies on traditional airborne acoustic paths, thus effectively improving the accuracy and anti-interference capability of human voice recognition in complex noise environments. Simultaneously, by processing the driver's EEG signals to obtain their attention and physiological states, and performing multimodal fusion analysis with acoustic features, the sound field control can be dynamically adjusted according to the driver's real-time state, thereby improving the perceptibility and responsiveness of key voice information. Furthermore, by linking the sound field control strategy with the safety response strategy, the system can automatically trigger corresponding adjustments or safety measures when abnormal states are detected, thereby improving the intelligence level of the in-vehicle system and driving safety. Overall, this achieves a synergistic improvement in acoustic perception capabilities, interactive response capabilities, and safety assurance capabilities.

[0007] In one possible design, EEG characteristics include at least the energy percentage of the theta wave band and information on energy variations in the delta wave band.

[0008] Compared with existing technologies, the technical effects achieved by this solution are as follows: By introducing the energy proportion of the theta band and the energy change information of the delta band, the system can accurately characterize the driver from two dimensions: attention state and abnormal physiological state. This improves the pertinence and reliability of driver state recognition. The energy proportion of the theta band reflects the driver's level of concentration, which is helpful for timely detection of inattention, while the energy change information of the delta band reflects the trend of abnormal physiological state, which is helpful for early perception of sudden risks. Furthermore, the introduction of the energy proportion of the theta band and the energy change information of the delta band makes the triggering basis of sound field control and safety response clearer and more physiologically grounded, thereby improving the accuracy and stability of overall decision-making and further enhancing the adaptive capability and safety assurance capability of the vehicle system in complex driving scenarios.

[0009] In one possible design, the acoustic feature information corresponding to the human voice signal includes the human voice probability value and / or breathing frequency characteristics.

[0010] Compared with existing technologies, the technical effects achieved by this solution are as follows: By specifically setting the acoustic feature information corresponding to human voice signals as human voice probability values ​​and / or respiratory frequency features, the system can not only quantify the probability of the presence of human voice signals during acoustic analysis, but also further reflect respiratory rhythm information related to human physiological activities, thereby improving the precision and reliability of human voice signal recognition. Among them, the human voice probability value can provide a clear basis for sound field control, which is conducive to improving the accuracy of human voice enhancement and non-human voice suppression. The respiratory frequency feature can provide auxiliary information for driver physiological state assessment, enabling the system to perceive abnormal states from acoustic signals. By introducing human voice probability values ​​and respiratory frequency, acoustic information is no longer limited to the traditional speech recognition level, but is extended to the expression of multi-dimensional information related to human state, thereby enhancing the effectiveness of multimodal fusion analysis and further improving the adaptive control capability and safety assurance capability of the vehicle system in complex environments.

[0011] In one possible design, the multimodal fusion analysis includes: when the energy proportion of the theta wave band is higher than a first preset threshold, determining that the driver is in a distracted state, and further determining the state based on the human voice probability value; if the human voice probability value is higher than a second preset threshold, increasing the output priority of the human voice signal; if the human voice probability value is not higher than the second preset threshold, maintaining the current volume setting.

[0012] Compared with existing technologies, the technical effects achieved by this solution are as follows: By linking the energy proportion of the theta wave band with the probability value of human voice and constructing a hierarchical judgment mechanism based on preset thresholds, the system can, in addition to recognizing the driver's distracted state, further adjust the system based on the presence of human voice signals in the current acoustic environment, thereby avoiding misjudgments caused by relying solely on the driver's state or solely on acoustic information. Specifically, when a decrease in attention is detected and the probability of human voice is high, the system can actively increase the output priority of human voice signals, making key voice information clearer and more discernible; when the conditions are not met, the original volume setting is maintained, thereby avoiding unnecessary interference and over-adjustment. Furthermore, through the fusion judgment and differentiated control methods, the sound field control process can be made more refined, reasonable, and targeted, thereby improving the information transmission efficiency and overall stability of the in-vehicle system in complex driving scenarios, while enhancing driving safety and user experience.

[0013] In one possible design, multimodal fusion analysis includes: when a sudden change in the energy of the delta band is detected within a preset time window, a state determination is further made based on the breathing frequency; if the breathing frequency is higher than a third preset threshold, it is determined that the driver has an abnormal physiological state and a safety response strategy is triggered; if the breathing frequency is not higher than the third preset threshold, a warning reminder is triggered.

[0014] Compared with existing technologies, the technical effects achieved by this solution are as follows: By jointly determining the abrupt changes in delta wave energy within a preset time window and respiratory frequency, the system can more accurately identify the driver's state by combining respiratory characteristics directly related to human physiological activities with the detection of abnormal EEG signals, thereby effectively reducing the risk of misjudgment caused by relying solely on a single EEG feature. Specifically, when the respiratory frequency exceeds a preset threshold, it can be promptly identified as an abnormal physiological state and trigger a safety response strategy, while when no abnormal conditions are met, a warning reminder is executed, achieving a balance between safety and the rationality of intervention. Through the above-mentioned multimodal collaborative judgment mechanism, the identification of sudden risks becomes more reliable and hierarchical, not only improving the accuracy and stability of abnormal state detection but also enhancing the system's ability to distinguish between different levels of risk, thereby further improving the safety assurance capability and response effectiveness of the vehicle system in actual driving scenarios.

[0015] In one possible design, the warning alert operation includes triggering seat vibration to alert the driver.

[0016] Compared with existing technologies, the technical effects achieved by this solution are as follows: by triggering seat vibration, the prompts can be delivered directly to the driver in tactile form, thus avoiding the problems of information interference or neglect caused by relying solely on sound or visual cues; seat vibration can still effectively transmit warning signals when the driver's attention is reduced or in a complex acoustic environment, improving the perceptibility and timeliness of the reminders; at the same time, as a non-acoustic and non-visual prompting method, seat vibration will not place an additional burden on the current audio output or visual interface, which is conducive to maintaining the stability of the sound field control strategy.

[0017] In one possible design, the sound pressure level of the music signal is reduced when the EEG characteristics indicate that the driver's attention is below a fourth preset threshold and the SAW sensor detects the presence of navigation voice.

[0018] Compared with existing technologies, the technical effects achieved by this solution are as follows: By jointly determining the driver's attention state as represented by EEG features and the navigation voice detected by the SAW sensor, the sound field modulation can specifically reduce the sound pressure level of the music signal when the driver's attention is reduced, thereby avoiding the problem of key navigation voice being masked by background music. This method no longer relies on fixed volume strategies or manual intervention, but rather makes adaptive adjustments based on the driver's real-time state and acoustic environment, making the presentation of voice information clearer and more targeted, while improving the effectiveness of information transmission without affecting the normal driving experience.

[0019] In one possible design, the sound field control strategy involves directional amplification of the target human voice signal through bone conduction speakers.

[0020] Compared with existing technologies, the technical effects achieved by this solution are as follows: By introducing bone conduction speakers into the sound field control strategy to directionally enhance the output of the target human voice signal, the human voice information no longer relies entirely on the air propagation path, but acts directly on the driver through bone conduction, thereby effectively reducing the interference of environmental noise and music on key speech, and improving the clarity and perceptibility of human voice information; the bone conduction speakers can maintain the stability of voice transmission in complex acoustic environments, while avoiding excessive interference to the overall in-vehicle sound field.

[0021] In one possible design, the safety response strategy includes controlling the vehicle to perform deceleration and / or issuing alarm signals and / or sending distress messages to the outside.

[0022] Compared with existing technologies, the technical effects achieved by adopting this solution are as follows: The safety response strategy specifically includes controlling the vehicle to perform deceleration operations, issuing alarm signals, and / or sending distress messages to the outside world. This enables the system to directly transition from information perception to actual intervention when abnormal conditions are detected, thus avoiding the limitations of merely providing warnings. This can promptly reduce vehicle operation risks and simultaneously send distress messages to the outside world when the driver's condition is abnormal or there are potential risks, which helps to shorten emergency response time and improve handling efficiency. At the same time, by combining vehicle control with information communication, the safety response can be linked and complete, thereby improving the system's ability to respond in emergency situations.

[0023] In one possible design, multimodal fusion analysis uses an onboard SoC processor to jointly calculate the acoustic features of the human voice signal and the EEG features.

[0024] Compared with existing technologies, the technical effects achieved by this solution are as follows: By using the onboard SoC processor to perform unified joint calculations on the acoustic feature information and EEG features corresponding to human voice signals, multimodal data can be fused and analyzed on the same processing platform, thereby avoiding the latency and data inconsistency problems caused by the decentralized processing between different modules; this method can improve the real-time performance and computational efficiency of multimodal fusion, and ensure the temporal consistency between acoustic information and EEG information, making the fusion results more accurate and reliable; at the same time, relying on the centralized processing capabilities of the onboard SoC, it is beneficial to realize the rapid generation and execution of sound field control and safety response strategies, thereby improving the overall stability and response speed of the system. Attached Figure Description

[0025] Figure 1A flowchart illustrating an in-vehicle intelligent sound field modulation method based on electroencephalograms and surface acoustic waves, provided for an embodiment of this application; Figure 2 The following is an execution flowchart of an in-vehicle intelligent sound field control method based on electroencephalogram (EEG) and surface acoustic waves, provided as an embodiment of this application. Detailed Implementation

[0026] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0027] See Figures 1 to 2 This invention provides a vehicle-mounted intelligent sound field control method based on electroencephalogram (EEG) and surface acoustic waves, comprising: Step S1: Acquire in-vehicle acoustic signals and driver's EEG signals, wherein the acoustic signals are acquired by a surface acoustic wave (SAW) sensor and the EEG signals are acquired by an EEG acquisition device. Step S2: Process the acoustic signal, extract the mechanical vibration features related to vocal cord vibration, and distinguish human voice signals from non-human voice signals based on the mechanical vibration features; Step S3: Process the EEG signals and extract EEG features that characterize the driver's attention state and / or physiological state; Step S4: Perform multimodal fusion analysis based on the acoustic feature information and EEG features corresponding to the human voice signal to generate a sound field control strategy and / or safety response strategy corresponding to the driver's state; Step S5: Adjust the vehicle audio output according to the sound field control strategy to enhance human voice signals and / or suppress non-human voice signals; and / or trigger the corresponding safety response operation according to the safety response strategy.

[0028] Specifically, by introducing a surface acoustic wave (SAW) sensor to directly collect and identify the mechanical vibration characteristics generated by vocal cord vibration, the extraction of human voice signals no longer relies on the traditional airborne acoustic path, thereby effectively improving the accuracy and anti-interference capability of human voice recognition in complex noise environments. Simultaneously, by processing the driver's EEG signals to obtain their attention and physiological states, and performing multimodal fusion analysis with acoustic features, the sound field control can be dynamically adjusted according to the driver's real-time state, thereby improving the perceptibility and responsiveness of key voice information. The SAW sensor uses a center frequency of 38MHz to 42MHz, preferably 40MHz, with a bandwidth of not less than 10MHz and a sensitivity of approximately -60dB. The SAW sensor is used to perceive acoustic signals based on the physical characteristics generated by vocal cord vibration during human voice production. The fundamental frequency range corresponding to human voice is 80Hz to 300Hz. Response to the vibration characteristics in this frequency band facilitates the identification of human voice signals and the effective differentiation of instrument sounds and environmental noise, thereby improving the accuracy and anti-interference capability of acoustic signal acquisition.

[0029] In one embodiment of this application, the EEG characteristics include at least the energy percentage of the theta wave band and the energy variation information of the delta wave band.

[0030] Specifically, in this embodiment, the electroencephalogram (EEG) features are acquired via an EEG sensor. These features include at least the energy percentage of the theta wave band and energy variation information in the delta wave band. The theta wave band typically falls within the range of 4Hz to 8Hz, and its energy percentage reflects the driver's level of concentration. A higher theta wave energy percentage usually indicates that the driver is inattentive or fatigued. The delta wave band typically falls within the range of 0.5Hz to 4Hz, and its energy variation information reflects whether the driver is experiencing abnormal physiological states, such as confusion or sudden discomfort. By selecting theta and delta waves as EEG features, the driver's attention state can be assessed in real time, and abnormal physiological states can be effectively detected, thus providing a reliable physiological basis for sound field control and safety response.

[0031] In one embodiment of this application, the acoustic feature information corresponding to the human voice signal includes the human voice probability value and / or breathing frequency features.

[0032] Specifically, in this embodiment, the acoustic feature information corresponding to the human voice signal includes the human voice probability value. And / or breathing frequency features; wherein, the human voice probability value is used to characterize the likelihood of the current acoustic signal being a human voice, so as to achieve quantitative judgment of the human voice signal; the breathing frequency feature is extracted based on the low-frequency acoustic features generated by vocal cord vibration and airflow changes, and is used to reflect the driver's breathing state. By simultaneously introducing the human voice probability value and breathing frequency features, the acoustic analysis can not only achieve human voice recognition, but also further associate human physiological information, thereby providing richer data support for multimodal fusion analysis.

[0033] In one embodiment of this application, the multimodal fusion analysis includes: when the energy proportion of the theta wave band is higher than a first preset threshold, determining that the driver is in a distracted state, and further determining the state based on the human voice probability value; if the human voice probability value is higher than a second preset threshold, increasing the output priority of the human voice signal; if the human voice probability value is not higher than the second preset threshold, maintaining the current volume setting.

[0034] Specifically, in this embodiment, the first preset threshold is 30%; the second preset threshold is 70%; the multimodal fusion analysis includes: when the energy proportion of the theta wave band is higher than 30%, it is determined that the driver is in a distracted state, and the state is further determined based on the human voice probability value; if the human voice probability value is higher than 70%, it indicates that there is a strong human voice signal in the current environment, and the output priority of the human voice signal is increased to enhance the perceptibility of key voice information; if the human voice probability value is not higher than 70%, the current volume setting is maintained to avoid unnecessary sound field adjustments; through the above-mentioned graded judgment mechanism, the sound field control can simultaneously consider the driver's state and the acoustic environment, improving the targeting and rationality of the control.

[0035] In one embodiment of this application, the multimodal fusion analysis includes: when a sudden change in the energy of the delta band is detected within a preset time window, a state determination is further made based on the breathing frequency; if the breathing frequency is higher than a third preset threshold, it is determined that the driver has an abnormal physiological state and a safety response strategy is triggered; if the breathing frequency is not higher than the third preset threshold, a warning reminder operation is triggered.

[0036] Specifically, in this embodiment, the preset time window is 5 seconds; the third preset threshold is 30 times / minute; the multimodal fusion analysis includes: when a sudden change in the energy of the delta wave band is detected within the 5-second window, it is determined that the driver may have an abnormal physiological state, and the state is further determined based on the breathing rate; if the breathing rate is higher than 30 times / minute, it indicates that the driver has an abnormal breathing condition, thereby determining it as an abnormal physiological state and triggering a safety response strategy; if the breathing rate is not higher than 30 times / minute, a warning reminder operation is triggered; through the joint determination method, the system can distinguish different risk levels, thereby achieving graded response.

[0037] In one embodiment of this application, the warning alert operation includes triggering seat vibration to alert the driver.

[0038] Specifically, in this embodiment, the warning reminder operation includes triggering seat vibration to alert the driver. The seat vibration is achieved through a vibration actuator installed inside the seat. This vibration alert can directly act on the driver in the form of tactile sensation, thereby achieving effective reminder without interfering with the current acoustic environment and improving the perceptibility of the warning information.

[0039] In one embodiment of this application, when the EEG characteristics indicate that the driver's attention is below a fourth preset threshold and the SAW sensor detects the presence of navigation voice, the sound pressure level of the music signal is reduced.

[0040] Specifically, in this embodiment, the fourth preset threshold is 20%; when the EEG characteristics indicate that the driver's attention is below 20% and the SAW sensor detects the presence of navigation voice, the sound pressure level of the music signal is reduced; wherein, the navigation voice can be obtained by recognizing preset voice features in the human voice signal. In this way, key navigation information can be presented first when the driver's attention decreases, thereby improving the effectiveness of information transmission.

[0041] In one embodiment of this application, the sound field modulation strategy includes directional enhancement output of the target human voice signal through a bone conduction loudspeaker.

[0042] Specifically, in this embodiment, the sound field control strategy includes directional enhancement of the target human voice signal through a bone conduction speaker; wherein, the bone conduction speaker transmits sound by transmitting vibration signals to the driver's skull, thereby avoiding interference from environmental noise on human voice information and improving the clarity of the speech signal.

[0043] In one embodiment of this application, the safety response strategy includes controlling the vehicle to perform deceleration and / or issuing an alarm signal and / or sending a distress message to the outside.

[0044] Specifically, in this embodiment, the safety response strategy includes controlling the vehicle to perform deceleration and / or issuing an alarm signal and / or sending a distress message to the outside. The distress message may include vehicle location information and driver status information. In this way, the system can intervene in a timely manner and link with external systems when it detects an abnormal state, thereby improving emergency response capabilities.

[0045] In one embodiment of this application, multimodal fusion analysis uses an onboard SoC processor to jointly calculate the acoustic feature information and EEG features corresponding to human voice signals.

[0046] Specifically, in this embodiment, the multimodal fusion analysis uses an onboard SoC processor to jointly calculate the acoustic feature information and EEG features corresponding to the human voice signal. The onboard SoC processor is used to perform unified processing and fusion analysis of multi-source data, thereby ensuring the real-time performance and consistency of data processing and improving the overall operating efficiency of the system.

[0047] This application achieves the following breakthroughs in the field of vehicle sound field control through multimodal fusion technology: First, it uses a surface acoustic wave (SAW) sensor to directly capture vocal cord vibration characteristics, combined with real-time monitoring of driver attention via electroencephalogram (EEG), improving voice recognition accuracy in complex noise environments. Second, it enhances the coverage of sudden illness recognition through a dual-modal health monitoring mechanism combining acoustic breathing characteristics and delta waves from the EEG. Third, it achieves coordinated control of navigation voice direction enhancement and intelligent music attenuation through a dynamic sound field optimization algorithm, reducing power consumption while remaining fully compatible with existing vehicle hardware. These features comprehensively improve the system's safety, interactivity, and reliability, providing a more complete acoustic solution for intelligent driving.

[0048] In one embodiment of this application, when the vehicle is traveling at high speed, the system detects navigation voice signals within the vehicle using a surface acoustic wave (SAW) sensor, with a high probability value for the corresponding human voice. Simultaneously, analysis of electroencephalogram (EEG) signals reveals a high proportion of energy in the theta wave frequency band, indicating that the driver is in a state of distraction. In this situation, the system generates a sound field modulation strategy based on multimodal fusion analysis results, adjusting the in-vehicle audio output to reduce the sound pressure level of the music signal. Simultaneously, it uses bone conduction speakers to directionally enhance the navigation voice output, improving the clarity and perceptibility of the navigation information. Furthermore, the system can output prompts through the in-vehicle display device to assist in reminding the driver to pay attention to driving safety, thereby improving the transmission of key information without affecting the overall auditory experience.

[0049] In one embodiment of this application, when the system detects abnormal breathing characteristics through a surface acoustic wave (SAW) sensor, and the breathing rate is significantly higher than the normal range, and simultaneously detects a significant increase in delta wave frequency energy within a preset time window through electroencephalogram (EEG) signal analysis, the system determines that the driver may have an abnormal physiological state based on multimodal fusion analysis, and generates a safety response strategy accordingly. This strategy controls the vehicle to decelerate and triggers a vehicle alarm device to provide a warning. If the driver does not respond effectively, the system can also send a distress signal to the outside world. This distress signal includes vehicle location information and driver-related status information, thereby enabling timely intervention and external coordination in emergency situations and improving driving safety.

[0050] While the present invention has been disclosed above, it is not limited thereto. Any person skilled in the art can make various modifications and alterations without departing from the spirit and scope of the invention; therefore, the scope of protection of the present invention should be determined by the scope defined in the claims.

Claims

1. A vehicle-mounted intelligent sound field control method based on electroencephalogram (EEG) and surface acoustic waves, characterized in that, include: Acoustic signals from inside the vehicle and electroencephalogram (EEG) signals from the driver are acquired, wherein the acoustic signals are acquired by a surface acoustic wave (SAW) sensor and the EEG signals are acquired by an EEG acquisition device. The acoustic signal is processed to extract mechanical vibration features related to vocal cord vibration, and human voice signals are distinguished from non-human voice signals based on the mechanical vibration features. The EEG signals are processed to extract EEG features that characterize the driver’s attention state and / or physiological state; Based on the acoustic feature information corresponding to the human voice signal and the EEG features, a multimodal fusion analysis is performed to generate a sound field control strategy and / or safety response strategy corresponding to the driver's state. The vehicle audio output is adjusted according to the sound field control strategy to enhance human voice signals and / or suppress non-human voice signals; and / or, a corresponding safety response operation is triggered according to the safety response strategy.

2. The vehicle-mounted intelligent sound field control method based on electroencephalogram (EEG) and surface acoustic waves according to claim 1, characterized in that, The EEG characteristics include at least the energy percentage of the theta wave band and the energy variation information of the delta wave band.

3. The vehicle-mounted intelligent sound field control method based on electroencephalogram (EEG) and surface acoustic waves according to claim 2, characterized in that, The acoustic feature information corresponding to the human voice signal includes the human voice probability value and / or respiratory frequency features.

4. The vehicle-mounted intelligent sound field control method based on electroencephalogram (EEG) and surface acoustic waves according to claim 3, characterized in that, The multimodal fusion analysis includes: When the energy proportion of the theta wave band is higher than the first preset threshold, it is determined that the driver is in a state of distraction, and the state is further determined based on the human voice probability value; if the human voice probability value is higher than the second preset threshold, the output priority of the human voice signal is increased; if the human voice probability value is not higher than the second preset threshold, the current volume setting is maintained.

5. The vehicle-mounted intelligent sound field control method based on electroencephalogram (EEG) and surface acoustic waves according to claim 3, characterized in that, The multimodal fusion analysis includes: When a sudden change in the energy of the delta wave band is detected within a preset time window, the state is further determined based on the breathing rate; if the breathing rate is higher than the third preset threshold, it is determined that the driver has an abnormal physiological state and a safety response strategy is triggered; if the breathing rate is not higher than the third preset threshold, a warning reminder is triggered.

6. The vehicle-mounted intelligent sound field control method based on electroencephalogram (EEG) and surface acoustic waves according to claim 5, characterized in that, The warning alert operation includes triggering seat vibration to alert the driver.

7. The vehicle-mounted intelligent sound field control method based on electroencephalogram (EEG) and surface acoustic waves according to claim 3, characterized in that, When the EEG characteristics indicate that the driver's attention is below a fourth preset threshold and the SAW sensor detects the presence of navigation voice, the sound pressure level of the music signal is reduced.

8. The vehicle-mounted intelligent sound field control method based on electroencephalogram (EEG) and surface acoustic waves according to claim 1, characterized in that, The sound field control strategy includes directional enhancement of the target human voice signal through bone conduction loudspeakers.

9. The vehicle-mounted intelligent sound field control method based on electroencephalogram (EEG) and surface acoustic waves according to claim 1, characterized in that, The safety response strategy includes controlling the vehicle to perform deceleration and / or issuing alarm signals and / or sending distress messages to the outside world.

10. The vehicle-mounted intelligent sound field control method based on electroencephalograms and surface acoustic waves according to claim 1, characterized in that, The multimodal fusion analysis uses an onboard SoC processor to jointly calculate the acoustic feature information and EEG features corresponding to the human voice signal.

Images