Active noise control based in-vehicle noise reduction system and method for electric vehicle

Abstract

This invention discloses an in-vehicle noise reduction system and method for electric vehicles based on active noise control (AVAS). The system includes an AVAS controller, a multi-source signal acquisition module, a dynamic gain adjustment module, an adaptive filtering and processing module, an in-vehicle noise reduction execution module, an external warning compensation module, an error sensing module, and a central control unit. The adaptive filtering and processing module inputs the AVAS reference signal and error signal into a wide-narrow band hybrid ANC algorithm, ultimately generating an anti-phase sound wave control signal with the same amplitude but opposite phase to the original AVAS sound entering the vehicle. The in-vehicle noise reduction execution module causes acoustic interference and mutual cancellation between the played anti-phase sound wave and the original AVAS noise entering the vehicle through the vehicle body path near the passenger's ear canal. This invention not only intelligently and dynamically cancels the AVAS warning noise entering the vehicle, but also maintains or even optimizes the warning sound effect outside the vehicle.

Description

Technical Field

[0001] This invention relates to the field of vehicle noise control technology, specifically to an active noise control-based AVAS (Active Noise Reduction System) in-vehicle noise reduction system and method for electric vehicles. Background Technology

[0002] To address the potential safety hazards to pedestrians (especially visually impaired individuals) caused by the low noise levels of electric vehicle powertrains at low speeds, major automotive markets worldwide have enacted regulations mandating the installation of acoustic vehicle alert systems (AVAS) on the exterior of vehicles. These systems, applied to electric vehicles (EVs) and hybrid electric vehicles (HEVs), emit specific warning sounds (typically including simulated motor noise, a buzzer at a specific frequency, or synthesized sound effects) when the vehicle is traveling at low speeds (usually below 20 km / h) or reversing, to alert pedestrians to the presence of the vehicle.

[0003] However, while existing AVAS systems warn pedestrians, the sounds they emit inevitably enter the vehicle through the vehicle structure and airflow. Since electric vehicles are very quiet at low speeds, this externally introduced artificial warning sound can become a new and unpleasant noise source in the cabin. Especially under conditions of frequent start-stop operations, continuous or intermittent warning sounds can severely disrupt the cabin's quietness and affect driving comfort.

[0004] Currently, the control of in-vehicle noise mainly employs passive sound insulation (such as adding sound insulation materials) and active noise control (ANC) technology. However, traditional sound insulation methods increase vehicle weight and cost, and have limited effectiveness against low-frequency noise; while existing in-vehicle ANC technology is mainly used to suppress broadband or narrowband noise such as engine order noise, road noise, and wind noise, and its reference signal usually comes from engine speed sensor or vibration accelerometer.

[0005] Given the unique characteristics of AVAS noise (① its sound source is clear and the signal is known, originating from the external AVAS controller; ② its generation and propagation path differs from traditional powertrain noise; ③ it is necessary to suppress its interference inside the vehicle while strictly ensuring its acoustic performance outside the vehicle complies with regulations), existing technologies have not designed a solution to effectively isolate or eliminate AVAS noise interference inside the vehicle without affecting the pedestrian warning function outside the vehicle. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention proposes an in-vehicle noise reduction system and method for electric vehicles based on active noise control. This system can not only intelligently and dynamically cancel the AVAS warning noise entering the vehicle, but also maintain or even optimize the warning sound effect outside the vehicle, thereby achieving the dual goals of "quiet interior and warning exterior".

[0007] To achieve the above objectives, the present invention provides an AVAS in-vehicle noise reduction system for electric vehicles based on active noise control, which is characterized by including an AVAS controller, a multi-source signal acquisition module, a dynamic gain adjustment module, an adaptive filtering and processing module, an in-vehicle noise reduction execution module, an external warning compensation module, an error sensing module, and a central control unit. The AVAS controller is used to generate an original AVAS warning sound signal that complies with regulatory requirements when the triggering conditions are met. The multi-source signal acquisition module is used to acquire the original AVAS warning sound signal, vehicle motion state parameters, and error signal at the noise reduction target point in real time. The dynamic gain adjustment module is used to dynamically adjust the gain of the original AVAS warning sound signal according to the real-time vehicle motion state parameters, and output the adjusted AVAS reference signal. The adaptive filtering and processing module is used to input the AVAS reference signal and error signal into the wide and narrow band hybrid ANC algorithm. The hybrid wideband and narrowband ANC algorithm first inputs the AVAS reference signal and error signal into a linear prediction filter. This linear prediction filter separates the deterministic narrowband component and the residual wideband random component contained in the signal. The narrowband component generates a sinusoidal reference signal corresponding to the AVAS fundamental frequency and harmonics through a harmonic synthesizer, and then sends the sinusoidal reference signal to a narrowband adaptive filter based on the FxLMS algorithm. The wideband component directly uses the AVAS reference signal or error signal as a random reference signal, and then sends the random reference signal to a wideband adaptive filter based on the FxLMS algorithm. Finally, the output signal of the narrowband adaptive filter is compared with that of the wideband adaptive filter. ) The output signal is summed by a summer; finally, an anti-phase sound wave control signal with the same amplitude but opposite phase as the original AVAS sound entering the vehicle is generated. The in-vehicle noise reduction execution module is used to receive and play out the anti-phase sound wave control signal, so that the played anti-phase sound wave and the original AVAS noise that enters the vehicle through the vehicle body path produce acoustic interference and cancel each other out near the passenger's ear canal. The central control unit is communicatively connected to all modules, and is used to coordinate the operation of each module, handle algorithm scheduling, and manage the start and stop of the system.

[0008] Furthermore, the dynamic gain adjustment module includes an internally configured two-dimensional lookup table and an algorithm model. The two-dimensional lookup table is used to calibrate the recommended gain coefficient of the original AVAS warning sound signal under different combinations of vehicle motion state parameters. The gain coefficient of the original AVAS warning sound signal is determined by the two-dimensional lookup table, and then the adjusted AVAS reference signal is output by the algorithm model.

[0009] Furthermore, in the dynamic gain adjustment module, the algorithm model is expressed by the following formula: x’(n)=G*x(n) In the formula, x’(n) This indicates the adjusted AVAS reference signal. G Indicates the gain coefficient. x(n) This indicates the original AVAS warning tone signal.

[0010] Furthermore, in the adaptive filtering and processing module, the wide-narrow band hybrid ANC algorithm also includes a time-frequency masking module. The time-frequency masking module obtains the masking coefficient by performing short-time Fourier transform analysis on the AVAS reference signal and the error signal, and then uses the masking coefficient to identify which time-frequency units' energy mainly comes from AVAS noise.

[0011] Furthermore, in the adaptive filtering and processing module, the wide-narrow band hybrid ANC algorithm also includes a secondary path online identification module. When the secondary path online identification module is triggered, the central control unit controls the adaptive filtering and processing module to pause adaptation and injects self-noise into the in-vehicle noise reduction execution module as an identification excitation signal. Then, the response signal collected by the error sensing module is input together with the excitation signal, and the current secondary path impulse response model is estimated online through the adaptive algorithm. After the identification is completed, the updated secondary path impulse response model is injected into the filter update path of the adaptive filtering and processing module, replacing the original fixed model.

[0012] Furthermore, the external warning compensation module is used to digitally filter the original AVAS warning sound signal according to the external acoustic transfer function model and the current vehicle speed, calculate gain compensation, and output the external warning signal so that the warning sound emitted by the dedicated external speaker meets the sound pressure level requirements at the measurement points and key external frequency bands specified by regulations.

[0013] Furthermore, the error sensing module is an error microphone array used to collect error signals, response signals, and residual noise after noise reduction. 。

[0014] This invention also designs an in-vehicle noise reduction method for electric vehicles based on active noise control (AVAS), which is characterized by including the following steps: S101) System self-test, load preset parameters; S102) Real-time signal acquisition; S103) Determine if AVAS is activated and the vehicle speed is below the set threshold; if yes, enter the noise reduction main loop (S104)~S110); if no, the system goes into standby mode and returns to S102). S104) Based on the current vehicle motion state parameters, consult a two-dimensional lookup table, calculate and apply the gain coefficient G to obtain an AVAS reference signal suitable for ANC processing. x’(n) ; S105) AVAS reference signal x’(n) Sum of error signals e(n) Input a wide-band and narrow-band hybrid ANC algorithm to calculate the optimal anti-phase acoustic control signal at the current moment. y(n) ; S106) The phase-reversing acoustic wave control signal y(n) After D / A conversion and power amplification, the signal is used to drive the car's speakers for playback. S107) Original AVAS warning tone signal x(n) Filtering and gain compensation are performed to generate external warning signals. x ext (n) And play; S108) Collect residual noise after noise reduction e new (n) ; S109) Determine whether the error signal energy remains above the threshold for a preset time period, or whether the periodic identification time has been reached; if yes, proceed to S110) to perform online identification of the secondary path; if no, directly use residual noise. e new (n) Update the filter weights for the next time step and return to S102 to proceed to the next loop; S110) Pause the main noise reduction loop, inject the identification signal, acquire the response signal, and update the secondary path impulse response model. ; S111) Update the filter weight coefficients. After completion, return to the main noise reduction loop to continue noise reduction.

[0015] The advantages of this invention are: 1. Precise noise reduction and improved quietness: This invention directly targets the specific noise source AVAS, and utilizes its known signal characteristics to achieve precise and active cancellation in combination with ANC technology. This can significantly reduce the perceptibility of AVAS noise in the vehicle and effectively improve cabin quietness and ride comfort at low speeds. 2. Safety Guarantee and Compliance with Regulations: Through an independent external warning compensation channel, this invention ensures that the sound pressure level and frequency characteristics of the external warning sound meet the regulatory requirements of various countries under any operating conditions, without weakening its safety warning function for pedestrians, thus achieving a balance between safety and comfort. 3. Intelligent adaptation and strong robustness: This invention can adapt to changes in AVAS volume and frequency under different vehicle operating states (such as acceleration, deceleration, and reversing) through multi-source signal fusion (vehicle parameters, reference signals, and error signals) and dynamic gain adjustment. The secondary path online identification function further enhances the system's robustness to changes in the in-vehicle acoustic environment. 4. Cost optimization and easy integration: This invention makes full use of the hardware already equipped in existing vehicles, such as AVAS controllers, car audio speakers (partially used for noise reduction), CAN bus, and ANC system error microphones and processors already available in some mid-to-high-end models. The main functions can be realized through software upgrades and algorithm integration. The cost of adding new hardware is low, making it easy to promote on existing vehicle platforms.

[0016] This invention relates to an AVAS (Active Noise Reduction System) in-vehicle noise reduction system and method for electric vehicles based on active noise control. It can not only intelligently and dynamically cancel the AVAS warning noise entering the vehicle, but also maintain or even optimize the warning sound effect outside the vehicle, thereby achieving the dual goal of "quiet inside the vehicle and warning outside the vehicle". Attached Figure Description

[0017] Figure 1 This is a block diagram illustrating the overall architecture principle of the AVAS (Active Noise Reduction System) in-vehicle noise reduction system for electric vehicles based on active noise control in this invention. Figure 2 This is a block diagram of the wideband and narrowband hybrid ANC algorithm in this invention; Figure 3 This is a flowchart illustrating the steps of the AVAS (Active Noise Reduction System) in-vehicle noise reduction method for electric vehicles based on active noise control in this invention. Detailed Implementation

[0018] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0019] In the description of this invention, it should be understood that the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention.

[0020] like Figure 1As shown, the present invention discloses an AVAS in-vehicle noise reduction system for electric vehicles based on active noise control, comprising an AVAS controller 100, a multi-source signal acquisition module 200, a dynamic gain adjustment module 300, an adaptive filtering and processing module 400, an in-vehicle noise reduction execution module 500, an external warning compensation module 600, an error sensing module 700, and a central control unit 800.

[0021] The AVAS controller 100 is an inherent component of the vehicle. When the triggering conditions are met (such as vehicle speed <20km / h, vehicle not in P gear), it generates an original AVAS warning sound signal that meets regulatory requirements. One signal is sent to the multi-source signal acquisition module 200, and the other is sent to the external warning compensation module 600, which can be directly connected to the external speaker.

[0022] The multi-source signal acquisition module 200 includes an AVAS signal acquisition unit 210, a vehicle status acquisition unit 220, and an error signal acquisition unit 230.

[0023] The AVAS signal acquisition unit 210 is used to directly acquire the original AVAS warning sound signal from the output terminal of the AVAS controller 100. x(n) .

[0024] The vehicle status acquisition unit 220 is used to acquire real-time vehicle speed via the vehicle CAN bus. v Motor speed RPM Accelerator pedal opening Acc pedal Vehicle motion state parameters, etc.

[0025] The error signal acquisition unit 230 is used to acquire the error signal at the noise reduction target point using the in-vehicle error microphone. e(n) In this embodiment, the in-vehicle error microphone consists of one or more high-precision microphones arranged in the area around the passenger's head (such as near the overhead reading light).

[0026] The multi-source signal acquisition module 200 provides the feedforward and feedback information required for the system to perform accurate noise cancellation.

[0027] The dynamic gain adjustment module 300 is connected to the multi-source signal acquisition module 200, receives vehicle motion state parameters, and sets up a two-dimensional lookup table and algorithm model internally. Based on the real-time vehicle motion state parameters, it determines the gain coefficient of the original AVAS warning sound signal through the two-dimensional lookup table, and then outputs the adjusted AVAS reference signal through the algorithm model.

[0028] The dynamic gain adjustment module 300 is used to dynamically adjust the gain of the AVAS reference signal so that it more accurately reflects the noise characteristics actually transmitted into the vehicle, and avoids the ANC system from becoming unstable or diverging due to sudden changes in the original AVAS warning sound signal.

[0029] The two-dimensional lookup table is used to calibrate different combinations of vehicle motion state parameters, for example ( , Under the combination, the original AVAS warning tone signal x(n) Recommended gain factor. For example, AVAS may not work or the volume may be very low when the vehicle is stationary, the volume may be moderate when driving slowly at low speeds, and the volume or frequency may change during rapid acceleration.

[0030] In this embodiment, when the vehicle is stationary ( When ), the gain coefficient Set to 0 (AVAS not working); when the vehicle is moving slowly at 10km / h, the gain coefficient is... Set to 1.0 (standard gain); when the system detects rapid acceleration (accelerator pedal opening) Acc pedal When the opening is large, the gain coefficient will be... Fine-tuned to 1.1 to anticipate a potential increase in incoming noise.

[0031] The algorithm model is expressed by the following formula. x’(n)=G*x(n) In the formula, x’(n) This indicates the adjusted AVAS reference signal. G Indicates the gain coefficient. x(n) This indicates the original AVAS warning tone signal.

[0032] The adaptive filtering and processing module 400 is the core processing unit of the system. It receives the AVAS reference signal after dynamic gain adjustment. x’(n) Sum of error signals e(n) and AVAS reference signal x’(n) Sum of error signals e(n) Input a wide-band and narrow-band hybrid ANC algorithm.

[0033] The wideband and narrowband hybrid ANC algorithm includes first processing the AVAS reference signal. x’(n) Sum of error signals e(n) The input is a linear prediction filter 410, which separates the deterministic narrowband (harmonic) components and residual broadband random components contained in the signal. The narrowband portion is then used by a harmonic synthesizer to generate a sinusoidal reference signal corresponding to the AVAS fundamental frequency and harmonics. xs (n) Then the sine wave reference signal x s (n) Feed into a narrowband adaptive filter based on the FxLMS algorithm W s (z) The broadband portion directly utilizes the AVAS reference signal. x’(n) or error signal e(n) As a random reference signal x b (n) Then the random reference signal x b (n) Feed into a broadband adaptive filter based on the FxLMS algorithm W b (z) Then the narrowband adaptive filter W s (z) Output signal and broadband adaptive filter W b (z) The output signal is summed by a summer; ultimately, an anti-phase sound wave control signal with the same amplitude but opposite phase as the original AVAS sound entering the vehicle is generated. y(n), The detailed process is attached. Figure 2 As shown.

[0034] Preferably, the wideband / narrowband hybrid ANC algorithm further includes a time-frequency masking module 420, which masks the AVAS reference signal. x’(n) Sum of error signals e(n) The masking coefficient is obtained by performing short-time Fourier transform (STFT) analysis. M(t,f) Then through the masking coefficient M(t,f) Identifying which time-frequency units' energy primarily comes from AVAS noise allows the wide-narrowband hybrid ANC algorithm to concentrate resources on suppressing AVAS noise, while avoiding accidental suppression of other desired sounds (such as navigation voice, music, etc.) that overlap with the AVAS frequency band, thus optimizing the generation of antiphase sound waves.

[0035] The in-vehicle noise reduction execution module 500 is used to receive anti-phase acoustic wave control signals. y(n) The sound waves are played back, causing acoustic interference and mutual cancellation between the back-phase sound waves and the original AVAS noise that enters the vehicle through the vehicle body path near the passenger's ear canal, thereby achieving noise reduction.

[0036] Specifically, the in-vehicle noise cancellation module 500 includes some mid-bass speakers from the vehicle audio system (located in the doors and below the dashboard) and / or headrest speakers dedicated to active noise cancellation. A power amplifier (not shown) converts the inverted sound wave control signal... y(n) It is converted into an analog signal and used to drive a speaker to produce sound, generating canceling sound waves.

[0037] The external warning compensation module 600 is used to ensure the effectiveness of external warnings and receives the original AVAS warning sound signal. x (n)， The external acoustic transfer function model is pre-calibrated, and based on the acoustic transfer function model and the current vehicle speed... v For the original AVAS warning tone signal x(n) Digital filtering is performed, gain compensation is calculated, and an external warning signal is output. x ext (n) This ensures that the warning sound emitted by the dedicated external speaker meets the sound pressure level requirements at the measurement points and key external frequency bands specified by regulations. This guarantees that while noise cancellation is performed inside the vehicle, the warning volume outside the vehicle is sufficient and compliant.

[0038] The measurement point specified by regulations is a dedicated external warning speaker located in the front bumper or grille of the vehicle, with the key external frequency band being 1.5-6kHz.

[0039] The error sensing module 700 is the aforementioned error microphone array, used to continuously monitor the noise reduction effect and collect error signals. e(n) and response signal d(n) .

[0040] The central control unit 800, as the brain of the system, communicates with all modules to coordinate the operation of each module, handle algorithm scheduling, and manage the start and stop of the system. For example, it can activate the noise reduction function of the present invention only when AVAS is activated and the vehicle speed is below a threshold.

[0041] The central control unit 800 is integrated into the vehicle's existing domain controller. In this embodiment, it is integrated into the in-vehicle infotainment system or a separate ANC controller.

[0042] Example 2 The difference between this embodiment 2 and embodiment 1 is that the wide and narrow band hybrid ANC algorithm also includes a secondary path online identification module 430.

[0043] The secondary path online identification module 430 identifies the system at the initial power-on stage or upon detecting an error signal. e(n)Automatically triggered when the noise level remains consistently high. Upon triggering, the central control unit 800 controls the adaptive filtering and processing module 400 to pause adaptive operation and injects a piece of self-noise (such as white noise) into the in-vehicle noise reduction execution module 500 as an identification excitation signal. u(n) Then the response signal collected by the error sensing module 700 d(n) With excitation signal u(n) The input is combined with the input data, and the current secondary path impulse response model is estimated online using algorithms such as LMS. After identification, the updated secondary path impulse response model will be used. The filter update path of the adaptive filtering and processing module 400 is injected to replace the original fixed model, so that the system can adapt to changes in the acoustic environment caused by changes in occupants, opening and closing of windows, etc., and always maintain the best noise reduction performance.

[0044] In this embodiment, the error signal e(n) Impulse response model via secondary path The filtered reference signal and the masking coefficients generated by the time-frequency masking module 420 M(t,f) They work together in the update process of the filter weight coefficients.

[0045] The update process of the filter weight coefficients is expressed by the following formula. W(n + 1)=W(n)+μ*e(n)x’(n) In the formula, W(n + 1) This represents the updated filter weight coefficients. W(n) This represents the filter weight coefficients before the update. μ Indicates the step size factor. e(n) Indicates the error signal. x’(n) This represents the impulse response model through the secondary path. The filtered reference signal.

[0046] Example 3 This embodiment 3 designs an in-vehicle noise reduction method for electric vehicles based on active noise control (AVAS), such as... Figure 3 As shown, it includes the following steps: S101) System initialization. After power-on, the system performs a self-test and loads preset parameters, such as a two-dimensional lookup table, an initial secondary path impulse response model, and an external acoustic transfer function model. S102) Real-time signal acquisition. Specifically, it synchronously acquires the original AVAS warning sound signal and the vehicle CAN signal (vehicle speed). (etc.), and error signals ; S103) Judgment and Activation. Determine if AVAS is activated and the vehicle speed is below the set threshold; if yes, enter the noise reduction main loop (S104)~S110); if no, the system goes into standby mode and returns to S102). Preferably, the set threshold for vehicle speed is 20 km / h.

[0047] S104) Dynamic gain adjustment. Based on the current vehicle speed. v and motor speed RPM By consulting a two-dimensional lookup table, calculating and applying the gain coefficient G, the adjusted AVAS reference signal is obtained. x’(n) ; S105) Adaptive filtering processing. The AVAS reference signal... x’(n) Sum of error signals e(n) Input a wide-band and narrow-band hybrid ANC algorithm to calculate the optimal anti-phase acoustic control signal at the current moment. y(n) This step continues to use the latest secondary path impulse response model. .

[0048] S106) In-vehicle cancellation. This cancels out the phase-reversing acoustic wave control signal. y(n) After D / A conversion and power amplification, the sound is driven by the in-vehicle speakers, which physically superimpose with the AVAS noise entering the vehicle to achieve active noise cancellation. S107) External compensation. This applies to the original AVAS warning sound signal. x(n) Filtering and gain compensation are performed to generate external warning signals. x ext (n) Simultaneous broadcasting to ensure compliance with regulations; S108) Effect monitoring and error feedback. An error microphone collects the residual noise after noise reduction. e new (n)。

[0049] S109) Determine if the model needs to be updated. Determine if the error signal energy remains above a threshold for a preset time, or if the periodic identification time has been reached. If yes, proceed to S110) to perform online identification of secondary paths; otherwise, directly use residual noise. e new (n) Update the filter weights for the next time step and return to S102 to proceed to the next loop; S110) Secondary path online identification. The main noise reduction loop is paused, the identification signal is injected, the response signal is acquired, and the secondary path impulse response model is updated. ; S111) Update the filter weight coefficients. After completion, return to the main noise reduction loop to continue noise reduction.

[0050] Through the above closed-loop process, the system achieves continuous, adaptive, and high-precision cancellation of AVAS noise inside the vehicle, while maintaining compliant and effective pedestrian warnings outside the vehicle.

[0051] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.

Claims

1. An AVAS (Active Noise Reduction System) for electric vehicles based on active noise control, characterized in that: It includes an AVAS controller (100), a multi-source signal acquisition module (200), a dynamic gain adjustment module (300), an adaptive filtering and processing module (400), an in-vehicle noise reduction execution module (500), an external warning compensation module (600), an error sensing module (700), and a central control unit (800). The AVAS controller (100) is used to generate an original AVAS warning tone signal that complies with regulatory requirements when the triggering conditions are met; The multi-source signal acquisition module (200) is used to acquire the original AVAS warning sound signal, vehicle motion state parameters, and error signal at the noise reduction target point in real time; The dynamic gain adjustment module (300) is used to dynamically adjust the gain of the original AVAS warning sound signal according to the real-time vehicle motion state parameters, and output an AVAS reference signal suitable for ANC processing. The adaptive filtering and processing module (400) is used to input the AVAS reference signal and error signal into the wide and narrow band hybrid ANC algorithm; The wideband-narrowband hybrid ANC algorithm first inputs the AVAS reference signal and error signal into a linear prediction filter (410). This linear prediction filter (410) separates the deterministic narrowband component and the residual wideband random component contained in the signal. The narrowband component generates a sinusoidal reference signal corresponding to the AVAS fundamental frequency and harmonics through a harmonic synthesizer, and then sends the sinusoidal reference signal into a narrowband adaptive filter based on the FxLMS algorithm. The wideband component directly uses the AVAS reference signal or error signal as a random reference signal, and then sends the random reference signal into a wideband adaptive filter based on the FxLMS algorithm. Finally, the output signal of the narrowband adaptive filter is compared with that of the wideband adaptive filter. ) The output signal is summed by a summer; finally, an anti-phase sound wave control signal with the same amplitude but opposite phase as the original AVAS sound entering the vehicle is generated. The in-vehicle noise reduction execution module (500) is used to receive and play out the anti-phase sound wave control signal, so that the played anti-phase sound wave and the original AVAS noise that enters the vehicle through the vehicle body path produce acoustic interference and cancel each other out near the passenger's ear canal. The central control unit (800) is connected to all modules and is used to coordinate the operation of each module, handle algorithm scheduling, and manage the start and stop of the system.

2. The AVAS in-vehicle noise reduction system for electric vehicles based on active noise control according to claim 1, characterized in that: The dynamic gain adjustment module (300) includes an internally configured two-dimensional lookup table and an algorithm model. The two-dimensional lookup table is used to calibrate the recommended gain coefficient of the original AVAS warning sound signal under different combinations of vehicle motion state parameters. The gain coefficient of the original AVAS warning sound signal is determined by the two-dimensional lookup table, and then the adjusted AVAS reference signal is output by the algorithm model.

3. The AVAS in-vehicle noise reduction system for electric vehicles based on active noise control according to claim 2, characterized in that: In the dynamic gain adjustment module (300), the algorithm model is expressed by the following formula. x'(n) = G*x(n) In the formula, x'(n) This indicates the adjusted AVAS reference signal. G Indicates the gain coefficient. x(n) This indicates the original AVAS warning tone signal.

4. The AVAS in-vehicle noise reduction system for electric vehicles based on active noise control according to claim 1, characterized in that: In the adaptive filtering and processing module (400), the wide and narrow band hybrid ANC algorithm also includes a time-frequency masking module (420). The time-frequency masking module (420) obtains the masking coefficient by performing short-time Fourier transform analysis on the AVAS reference signal and the error signal, and then identifies which time-frequency units' energy mainly comes from AVAS noise through the masking coefficient.

5. The AVAS in-vehicle noise reduction system for electric vehicles based on active noise control according to claim 4, characterized in that: In the adaptive filtering and processing module (400), the wide and narrow band hybrid ANC algorithm also includes a secondary path online identification module (430). When the secondary path online identification module (430) is triggered, the central control unit (800) controls the adaptive filtering and processing module (400) to pause the adaptation and inject self-noise into the in-vehicle noise reduction execution module (500) as an identification excitation signal; then the response signal collected by the error sensing module (700) is input together with the excitation signal, and the current secondary path impulse response model is estimated online through the adaptive algorithm. After identification, the updated secondary path impulse response model is injected into the filter update path of the adaptive filtering and processing module (400) to replace the original fixed model.

6. The AVAS in-vehicle noise reduction system for electric vehicles based on active noise control according to claim 1, characterized in that: The external warning compensation module (600) is used to digitally filter the original AVAS warning sound signal according to the external acoustic transfer function model and the current vehicle speed, calculate gain compensation, and output the external warning signal so that the warning sound emitted by the dedicated external loudspeaker meets the sound pressure level requirements at the measurement points and key external frequency bands specified by regulations.

7. The AVAS in-vehicle noise reduction system for electric vehicles based on active noise control according to claim 6, characterized in that: The error sensing module (700) is an error microphone array used to collect error signals, response signals, and residual noise after noise reduction. 。 8. A method for noise reduction inside an electric vehicle based on active noise control (AVAS), characterized in that, Includes the following steps: S101) System self-test, load preset parameters; S102) Real-time signal acquisition; S103) Determine if AVAS is activated and the vehicle speed is below the set threshold; if yes, enter the noise reduction main loop (S104)~S110); if no, the system goes into standby mode and returns to S102). S104) Based on the current vehicle motion state parameters, consult a two-dimensional lookup table, calculate and apply the gain coefficient G to obtain an AVAS reference signal suitable for ANC processing. x'(n) ; S105) AVAS reference signal x'(n) Sum of error signals e(n) Input a wide-band and narrow-band hybrid ANC algorithm to calculate the optimal anti-phase acoustic control signal at the current moment. y(n) ; S106) The phase-reversing acoustic wave control signal y(n) After D / A conversion and power amplification, the signal is used to drive the car's speakers for playback. S107) Original AVAS warning tone signal x(n) Filtering and gain compensation are performed to generate external warning signals. x ext (n) And play; S108) Collect residual noise after noise reduction e new (n) ; S109) Determine whether the error signal energy is continuously higher than the threshold within a preset time, or whether the periodic identification time has been reached; If so, proceed to S110) for online identification of secondary pathways; If not, then use the residual noise directly. e new (n) Update the filter weights for the next time step and return to S102 to proceed to the next loop; S110) Pause the main noise reduction loop, inject the identification signal, acquire the response signal, and update the secondary path impulse response model. ; S111) Update the filter weight coefficients. After completion, return to the main noise reduction loop to continue noise reduction.

Images