Electrostatic vibrating diaphragm car audio

Abstract

The invention relates to an electrostatic vibrating diaphragm automobile audio which is mainly used for an electric vehicle and can compensate vehicle-mounted electromagnetic interference audio and solve the problem of electromagnetic interference of a motor PWM switch and rotating speed harmonics on an audio system. The electrostatic diaphragm automobile sound system comprises an electrostatic diaphragm loudspeaker and an interference feature extraction module, the rotating speed of a motor and the PWM frequency are obtained through a CAN bus, and a dynamic interference feature set is analyzed and extracted through adaptive window length FFT; the EMI quantification module is used for calculating an electromagnetic pollution index at a frequency band of 8kHz-20kHz, and triggering compensation when the electromagnetic pollution index exceeds a threshold value of-40dB; the dynamic notch compensation module is used for generating an inversion signal; according to the double-loop feedback control module, the amplitude is adjusted through PID of an inner loop, and the phase is optimized through gradient descent of an outer loop so as to minimize THD. The method comprises five steps of signal acquisition, feature extraction, interference quantification, dynamic compensation and signal injection. Interference frequency drift can be accurately tracked, a tone quality MOS value is improved after compensation, and an over-compensation risk is avoided.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electric vehicle audio systems, and relates to an electrostatic diaphragm automobile audio system, in particular to an electrostatic diaphragm automobile audio system for automatically compensating electromagnetic interference of an electric vehicle. BACKGROUND

[0002] The electrostatic diaphragm automobile audio system is a high-end audio system based on the principle of high-voltage electrostatic field driving sound production. The core of the system is an extremely thin and almost weightless conductive diaphragm suspended between two perforated metal plates (stators) with opposite charges. When an audio signal is loaded onto the stator, the changing electrostatic field drives the charged diaphragm to vibrate at high speed, which squeezes the air to produce sound. This design enables the diaphragm to respond to electrical signals almost instantaneously, freeing it from the inertia of traditional moving coil speakers and heavy paper cones, thus exhibiting significant advantages in sound quality: it can accurately reproduce the subtle details and rich overtones of sound, with extremely smooth, clear and well-extended high-frequency response, natural and transparent mid-frequency, and extremely low sound distortion, presenting a "close to ideal state" listening experience that is pure, transparent and spacious, and is praised by many audiophiles as the closest playback method to the original live sound.

[0003] With the rapid popularization of electric vehicles, the electromagnetic interference generated by the high-voltage electrical system during operation has a growing impact on the audio quality of the electrostatic diaphragm automobile audio system. The pulse width modulation switching process in the motor drive unit and the high-frequency switching action of the inverter invade the audio signal link through conduction coupling and spatial radiation. The electromagnetic environment of traditional fuel vehicles is relatively simple, while the three-electric system of electric vehicles constitutes a complex interference source: the wide frequency harmonics generated by the rotating motor, the transient pulse group caused by the rapid switching action of the power inverter, and the high-precision sampling circuit of the battery management system which is easily disturbed. The dynamic change of motor speed causes the real-time drift of interference frequency, and the frequency shift speed of the interference source is extremely fast under the condition of sudden acceleration, forming time-varying non-stationary noise. These interference energies are mainly concentrated in the frequency band sensitive to human ears, forming persistent high-frequency squeaking and harmonic howling, which seriously degrades the music reproduction and speech clarity of the electrostatic diaphragm automobile audio system.

[0004] Existing solutions such as electromagnetic shielding and grounding can suppress fixed frequency interference, but cannot dynamically track the harmonic frequency shift of the motor. Fixed parameter notch filters can eliminate specific frequency noise, but the phase distortion caused by the center frequency shift leads to under-compensation or over-compensation. The hardware-level filtering scheme increases the complexity of the wiring harness and is difficult to optimize the layout due to space limitations. The industry has tried to use multi-microphone noise reduction technology, but its effective frequency bandwidth is limited and it is almost ineffective for medium and high frequency electromagnetic interference.

[0005] The industry urgently needs to develop an audio compensation technology that can simultaneously solve the three major pain points of interference dynamic tracking, real-time sound field adaptation, and hardware cost control. The ideal solution needs to build a closed-loop control system at the digital signal processing layer, predict the interference spectrum by sensing the motor state, generate a counteracting signal in conjunction with the vehicle dynamics, and achieve collaborative management of electromagnetic pollution and acoustic defects. SUMMARY

[0006] To solve the above problems, the present application provides an electrostatic diaphragm automobile audio system with an electrostatic diaphragm automobile loudspeaker and an audio compensation system for electric vehicles to achieve audio compensation, comprising:

[0007] An interference feature extraction module connected to the output end of the vehicle-mounted power amplifier is used to collect audio signals polluted by electromagnetic interference in real time, extract a set of interference characteristic frequencies dynamically associated with the motor PWM switching frequency and speed through Fast Fourier Transform (FFT) waterfall analysis technology, and the frequency components of this set are composed of motor fundamental frequency and harmonic components;

[0008] An EMI quantification module: based on the set of interference characteristic frequencies, calculate the electromagnetic pollution index, which quantifies the total strength of electromagnetic interference in the 8kHz to 20kHz frequency band through a specific formula, the calculation formula is:

[0009]

[0010] Where |X(f k )| is the amplitude of the characteristic frequency point, V rms is the total effective voltage of the audio signal, w(f k ) is the frequency band weight coefficient; if f k ∈[f s w-200,f s w+200], w=1.2, otherwise w=1.0;

[0011] A dynamic notch compensation module: generates a counteracting compensation signal based on the set of interference characteristic frequencies, and injects it into the pre-stage of the audio processing link through signal superposition, to offset the distortion introduced by electromagnetic interference;

[0012] A dual-loop feedback control module: contains two levels of control, the inner loop adjusts the amplitude of the compensation signal in real time based on the electromagnetic pollution index, and the outer loop dynamically optimizes the phase of the compensation signal according to the total harmonic distortion (THD) measurement value of the audio output end, forming a closed-loop control mechanism.

[0013] The interference feature extraction module contains a speed coupling unit that directly obtains the real-time speed f motor and PWM switching frequency f swand an interference frequency prediction model is established, the model outputs a frequency set {f sw ±n×f motor}, where n is an integer from 1 to 5, for guiding the FFT analysis to focus on the target frequency band.

[0014] The dynamic notch compensation module is implemented by an adjustable parameter filter bank, which is composed of a plurality of parallel notches, the center frequency of each notch is locked at the interference characteristic frequency point, and the notch depth is dynamically controlled by adjusting the filter coefficients; the compensation signal generation principle is:

[0015] For each characteristic frequency f k , an adjustable amplitude and adjustable phase inverse signal is generated, and the mathematical expression satisfies C(f k ) = -γ·X(f k )·e jφ , where γ is the amplitude coefficient output by the inner loop, and φ is the phase offset optimized by the outer loop.

[0016] The inner loop of the double-loop feedback control module is a proportional-integral-derivative (PID) amplitude controller, which activates compensation when the electromagnetic pollution index exceeds -40 dB, and the response time is less than 10 milliseconds; the outer loop is a phase optimizer with the goal of minimizing THD, which dynamically adjusts the phase of the compensation signal using the gradient descent algorithm, and the response time is less than 100 milliseconds.

[0017] The system is deployed in the digital signal processing (DSP) core of the vehicle-mounted audio processor, the compensation signal injection point is located in the analog audio link between the digital-to-analog converter (DAC) and the preamplifier, and the compensation signal is output through a high-precision digital-to-analog conversion channel and superimposed with the original audio signal in the analog domain.

[0018] The adjustable parameter filter bank adopts a second-order infinite impulse response (IIR) notch filter structure, the denominator polynomial contains a frequency tracking term to adapt to the motor speed change in real time, the numerator polynomial is configured with a mirror zero point to eliminate phase distortion caused by frequency offset, and the filter coefficients are updated in real time according to the output of the double-loop feedback control module.

[0019] An electromagnetic interference compensation method for electric vehicle-mounted audio, which is implemented by the system, includes:

[0020] Step S1, signal acquisition: real-time acquisition of the disturbed audio signal output by the power amplifier, and synchronous acquisition of the motor speed f motor and the PWM switching frequency f sw through the vehicle CAN bus;

[0021] Step S2, feature extraction: performing FFT waterfall plot analysis on the audio signal to extract characteristic frequency points related to motor harmonics within the 8kHz-20kHz frequency band, satisfying the relationship fk =f sw ±n×f motor ;

[0022] Step S3, interference quantification: calculation of electromagnetic pollution index When the index exceeds -40dB, the compensation mechanism is triggered;

[0023] Step S4, dynamic compensation: generating an anti-phase compensation signal that matches the characteristic frequency, and adjusting the signal amplitude coefficient γ and phase offset φ in real time through a dual-loop control mechanism;

[0024] Step S5, signal injection: injecting the compensation signal into the front-end stage of the audio processing chain, superimposing it with the original audio and then transmitting it to the power amplifier.

[0025] Step S2 adopts an adaptive window length FFT analysis strategy. When the motor speed change rate exceeds 100 rpm / s, a 10 millisecond short time window is enabled to quickly capture transient interference characteristics; when the speed change rate is lower than 10 rpm / s, a 100 millisecond long time window is used to improve the frequency resolution to accurately separate dense harmonic components.

[0026] The dual-loop control of step S4 includes:

[0027] Inner loop amplitude control: The deviation between the electromagnetic pollution index and the target threshold of -40dB is used as the input of the PID controller, and the output amplitude coefficient γ;

[0028] Outer loop phase optimization: The measured THD value at the power amplifier output is used as the loss function, and the gradient descent algorithm is used to iteratively update the phase offset φ. The step size of each iteration is adaptively adjusted according to the THD change rate.

[0029] The compensation signal generation in step S4 adopts a power limitation strategy, the total power of the compensation signal does not exceed 15% of the original audio signal power, and the compensation amount of a single frequency point |C(f k )|satisfies the constraint|C(f k )|≤0.5×|X(f k )|, to avoid audio distortion caused by overcompensation.

[0030] The beneficial effects of the present invention are:

[0031] The system of the present invention obtains motor operating status parameters in real time through the vehicle bus, constructs an interference frequency prediction model, and combines it with an adaptive time window frequency domain analysis strategy to automatically switch to millisecond-level transient capture mode when the motor speed fluctuates violently, and enables a high-resolution harmonic separation mechanism under steady-state conditions. This design completely solves the phase distortion problem caused by the dynamic drift of motor harmonics in traditional fixed notch filters, improves the interference feature capture accuracy by nearly 90%, and compresses the frequency drift tracking delay to the millisecond level, laying the foundation for dynamic compensation.

[0032] Based on real-time monitoring of electromagnetic pollution index thresholds, a proportional-integral-derivative controller dynamically generates the amplitude of an anti-phase compensation signal within ten milliseconds. Simultaneously, an intelligent optimization algorithm iteratively corrects the phase of the compensation signal, aiming to minimize total harmonic distortion. This dual-loop mechanism reduces electromagnetic interference intensity by over 15 decibels while simultaneously suppressing harmonic distortion to less than 1%, significantly eliminating secondary audio degradation caused by amplitude overshoot and phase mismatch in traditional single-loop control.

[0033] By limiting the total power of the compensation signal to no more than 15 percent of the original audio signal power and strictly limiting the compensation intensity at a single frequency point to less than 50 percent of the interference signal strength, a double safety margin is established. This design effectively blocks audio signal saturation and distortion caused by sudden changes in motor harmonic amplitude or overshoot of compensation parameters, ensuring that the compensation system maintains sound quality integrity and natural listening under all motor operating conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0035] Attachment Figure 1 This is a schematic diagram of the electrostatic diaphragm loudspeaker of the present invention;

[0036] Attachment Figure 2 This is a schematic diagram of the overall architecture of the present invention;

[0037] Attachment Figure 3 This is a compensation flow chart of the present invention. DETAILED DESCRIPTION

[0038] Example 1:

[0039] See attached Figure 1 To the attached Figure 3 , an electrostatic diaphragm car audio system, having an electrostatic diaphragm car speaker, utilizing an electromagnetic interference audio compensation system to achieve audio compensation, comprising:

[0040] The interference feature extraction module is connected with the output end of the vehicle-mounted power amplifier, is used for collecting audio signals polluted by electromagnetic interference in real time, and extracts an interference characteristic frequency set dynamically related to the motor PWM switching frequency and the rotating speed through a fast Fourier transform (FFT) waterfall chart analysis technology. The frequency components of the set are composed of the motor fundamental frequency and harmonic components.

[0041] The EMI quantification module calculates an electromagnetic pollution index based on the interference characteristic frequency set. The index quantifies the total strength of electromagnetic interference in the 8 kHz to 20 kHz frequency band through a specific formula. The calculation formula is:

[0042]

[0043] where |X(f k )| is the amplitude of the characteristic frequency point, V rms is the total effective voltage of the audio signal, and w(f k ) is the frequency band weight coefficient. If f k ∈[f s w-200,f s w+200], w=1.2, otherwise w=1.0.

[0044] The dynamic notch compensation module generates an inverse compensation signal based on the interference characteristic frequency and injects it into the pre-stage of the audio processing chain through signal superposition to offset the distortion introduced by electromagnetic interference.

[0045] The double-loop feedback control module includes an inner loop and an outer loop. The inner loop adjusts the amplitude of the compensation signal in real time based on the electromagnetic pollution index, and the outer loop dynamically optimizes the phase of the compensation signal based on the total harmonic distortion (THD) measurement value of the audio output end to form a closed-loop control mechanism.

[0046] The interference feature extraction module includes a rotating speed coupling unit. The rotating speed coupling unit directly obtains the real-time rotating speed f motor and the PWM switching frequency f sw of the motor through the vehicle CAN bus and establishes an interference frequency prediction model. The model outputs a frequency set {f sw ±n×f motor}, where n is an integer from 1 to 5, to guide the FFT analysis to focus on the target frequency band.

[0047] Implementation of the interference feature extraction module:

[0048] The motor PWM harmonic (8-20 kHz) dynamically drifts with the rotating speed, and the traditional fixed frequency band scanning cannot accurately capture the interference source. The solution is:

[0049] The motor control message is parsed through the CAN bus to obtain the rotating speed f motor(unit: Hz) and PWM switching frequency f sw (typical value 16 kHz), the interference frequency prediction model is established:

[0050]

[0051] wherein F int is the set of interference characteristic frequencies, f sw is the motor PWM switching fundamental frequency (typical value 16 kHz), f motor is the real-time motor speed, and n is the harmonic order (taking 1-5 integers).

[0052] Adaptive FFT waterfall analysis:

[0053] The power amplifier output signal is collected by an ADC (sampling rate 48 kHz) in the F int band, and local amplitude peak values are searched, satisfying ∣X(fk)∣>3σ (σ is the background noise standard deviation);

[0054] The dynamic notch compensation module is implemented by an adjustable parameter filter bank, which is composed of multiple parallel notches. The center frequency of each notch is locked at the interference characteristic frequency point, and the notch depth is dynamically controlled by adjusting the filter coefficients. The compensation signal generation principle is:

[0055] For each characteristic frequency f k , an adjustable amplitude and adjustable phase inverse signal is generated, and its mathematical expression satisfies C(f k ) = -γ·X(f k )·e jφ , wherein γ is the amplitude coefficient output by the inner loop, and φ is the phase offset optimized by the outer loop.

[0056] Dynamic notch compensation implementation:

[0057] Compensation signal generation:

[0058] γ k is the amplitude coefficient output by the inner loop, A k is the interference amplitude A k = X(f k ), and φ k is the phase offset optimized by the outer loop.

[0059] The filter coefficient update algorithm adopts a second-order IIR notch filter, and the transfer function is:

[0060]

[0061] Coefficient update rule:

[0062] df motor=abs(f motor (t)-f motor (t-1));

[0063] If df motor >50, then a1=-1.9*cos(2πf k / fs), otherwise a1=-1.99*cos(2πf k / fs); b0=1, b1=-2cos(2πf k / f s ),b2=1.

[0064] a2=α 2 , where α is the parameter that determines the pole radius (0<α<1). The closer α is to 1, the narrower the notch filter bandwidth.

[0065] The inner loop of the dual-loop feedback control module is an amplitude controller based on proportional-integral-differential (PID). When the electromagnetic pollution index exceeds -40dB, compensation is activated with a response time of less than 10 milliseconds. The outer loop is a phase optimizer aimed at minimizing THD. It uses a gradient descent algorithm to dynamically adjust the phase of the compensation signal with a response time of less than 100 milliseconds.

[0066] Dual-loop feedback control mechanism:

[0067] The inner loop (amplitude control) controller structure adopts digital PID, and the control formula is:

[0068]

[0069] Where error e(t) = EMI Target -EMI Index (t);

[0070] Among them, K p is the proportional gain, the typical value is 0.8, K i is the integral gain, the typical value is 0.05, K d is the differential gain, the typical value is 0.01, EMI Target Target interference threshold, response time: <10ms.

[0071] The outer loop (phase optimization) is implemented using the gradient descent algorithm, and the loss function is defined as:

[0072] L(φ)=THD meas +0.1×|Δφ|;

[0073] THD meas is the total harmonic distortion measured at the output of the power amplifier, and Δφ is the absolute value of the phase change.

[0074] Parameter update:

[0075]

[0076] Gradient calculation uses perturbation observation method: gradient calculation perturbation step size δ = 0.01, THD plus = measure THD (φ k + δ), THD minus = measure THD (φ k - δ);

[0077]

[0078] Learning rate adaptation: learning rate η = 0.1, when Otherwise, learning rate η = 0.01.

[0079] The system is deployed in the digital signal processing (DSP) core of the vehicle-mounted audio processor, the compensation signal injection point is located in the analog audio link between the digital-to-analog converter (DAC) and the preamplifier, and the compensation signal is output through a high-precision digital-to-analog conversion channel and superimposed with the original audio signal in the analog domain.

[0080] The adjustable parameter filter bank adopts a second-order infinite impulse response (IIR) notch filter structure, the denominator polynomial of which contains a frequency tracking term to adapt to the motor speed change in real time, and the numerator polynomial is configured with image zeros to eliminate phase distortion caused by frequency offset, and the filter coefficients are updated in real time according to the output of the double-loop feedback control module.

[0081] An electromagnetic interference compensation method for electric vehicle-mounted audio, which is implemented by using the system, comprises the following steps:

[0082] Step S1, signal acquisition: real-time acquisition of the disturbed audio signal output by the power amplifier, and synchronous acquisition of the motor speed f motor and the PWM switching frequency f sw through the vehicle CAN bus;

[0083] Step S2, feature extraction: FFT waterfall plot analysis is performed on the audio signal, and the feature frequency points related to the motor harmonics in the 8kHz-20kHz frequency band are extracted, which satisfy the relationship f k = f sw ±n×f motor ;

[0084] Step S3, interference quantization: calculation of the electromagnetic pollution index When the index exceeds -40dB, the compensation mechanism is triggered;

[0085] Step S4, dynamic compensation: generate a compensating signal with the same frequency as the characteristic frequency, and adjust the amplitude coefficient γ and the phase offset φ in real time through a double-loop control mechanism;

[0086] Step S5, signal injection: inject the compensating signal into the pre-stage of the audio processing chain, and superimpose it on the original audio before transmitting it to the power amplifier.

[0087] The step S2 adopts an adaptive window length FFT analysis strategy. When the motor speed variation rate exceeds 100 rpm / s, a 10 ms short time window is enabled to quickly capture the transient interference characteristics. When the speed variation rate is lower than 10 rpm / s, a 100 ms long time window is used to improve the frequency resolution to accurately separate the dense harmonic components.

[0088] The double-loop control of the step S4 includes:

[0089] Inner loop amplitude control: take the deviation of the electromagnetic pollution index and the target threshold -40 dB as the input of the PID controller, and output the amplitude coefficient γ;

[0090] Outer loop phase optimization: take the measured THD value at the output end of the power amplifier as the loss function, and use the gradient descent algorithm to iteratively update the phase offset φ. The iteration step size is adaptively adjusted according to the THD variation rate.

[0091] The compensation signal generation of the step S4 adopts a power limiting strategy. The total power of the compensating signal does not exceed 15% of the power of the original audio signal, and the individual frequency compensating amount |C(f k )| satisfies the constraint condition |C(f k )|≤0.5×|X(f k )|, avoiding overcompensation leading to audio distortion.

[0092] The above description of the embodiments is provided for the purpose of illustration and description. It is not intended to be exhaustive or to limit the disclosure. Individual elements or features of a particular embodiment are generally not limited to that particular embodiment, but are interchangeable and can be used in selected embodiments, even if not specifically shown or described. In many aspects, the same elements or features can also be changed. Such changes are not considered to deviate from the disclosure, and all such modifications are intended to be included within the scope of the disclosure.

[0093] Example embodiments are provided so as to be thorough and to convey the full scope of the disclosure. Numerous specific details are set forth such as examples of specific components, devices, and methods, in order to provide a thorough understanding of embodiments of the present disclosure. Clearly, the

[0094] terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. The terms "comprises" and "comprising," when used in this document, are each taken to mean, in the context of the disclosure, that the method includes the recited feature or step, but not excluding the presence or addition of one or more other features or steps. The method steps, processes, and operations described herein are not to be construed as necessarily requiring their performance in the particular order in which they are described. It is also to be understood that additional or alternative steps can be employed.

Claims

1. An electrostatic diaphragm car audio system capable of achieving audio compensation for electromagnetic interference in electric vehicles, comprising an electrostatic diaphragm car speaker, characterized in that Also includes: Interference Feature Extraction Module: This module is connected to the output of the vehicle's power amplifier and is used to collect audio signals contaminated by electromagnetic interference in real time. It uses Fast Fourier Transform (FFT) waterfall analysis technology to extract a set of interference feature frequencies dynamically associated with the motor's PWM switching frequency and speed. The frequency components of this set are composed of the motor's fundamental frequency and harmonic components. EMI quantification module: Calculates the electromagnetic pollution index based on the interference characteristic frequency set. The index quantifies the total intensity of electromagnetic interference in the 8kHz to 20kHz frequency band using a specific formula. The calculation formula is: where |X(f k )| is the amplitude of the characteristic frequency point, V rms is the total effective voltage of the audio signal, w(f k ) is the frequency band weight coefficient; if f k ∈[f s w-200,f s w+200], w=1.2, otherwise w=1.0; Dynamic notch compensation module: generates an anti-phase compensation signal according to the interference characteristic frequency, and injects it into the pre-stage of the audio processing chain through signal superposition to offset the distortion introduced by electromagnetic interference; A dual-loop feedback control module includes two levels of control: an inner loop and an outer loop. The inner loop adjusts the compensation signal amplitude in real time based on the electromagnetic pollution index, and the outer loop dynamically optimizes the compensation signal phase based on the total harmonic distortion (THD) measurement value at the audio output end, forming a closed-loop control mechanism.

2. The sound system according to claim 1, wherein: The interference feature extraction module includes a speed coupling unit, which directly obtains the real-time speed of the motor f through the vehicle CAN bus. motor and PWM switching frequency f sw , and establish an interference frequency prediction model, which outputs the frequency set {f sw ±n×f motor }, where n is an integer from 1 to 5, used to guide the FFT analysis to focus on the target frequency band.

3. The sound system according to claim 1, wherein: The dynamic notch compensation module is implemented using an adjustable parameter filter bank, which consists of multiple parallel notch filters. The center frequency of each notch filter is locked at the interference characteristic frequency point, and the notch depth is dynamically controlled by adjusting the filter coefficient. The compensation signal generation principle is: For each characteristic frequency f k Generate an inverted signal with adjustable amplitude and phase, whose mathematical expression satisfies C(f k )=-γ·X(f k )·e jφ , where γ is the amplitude coefficient of the inner loop output and φ is the phase offset optimized by the outer loop.

4. The sound system according to claim 1, wherein: The inner loop of the dual-loop feedback control module is an amplitude controller based on proportional-integral-differential (PID). When the electromagnetic pollution index exceeds -40dB, compensation is activated with a response time of less than 10 milliseconds. The outer loop is a phase optimizer aimed at minimizing THD. It uses a gradient descent algorithm to dynamically adjust the phase of the compensation signal with a response time of less than 100 milliseconds.

5. The sound system according to claim 1, wherein: The system is deployed in the digital signal processing (DSP) core of the on-board audio processor. The compensation signal injection point is located in the analog audio link between the digital-to-analog converter (DAC) and the preamplifier. The compensation signal is output through a high-precision digital-to-analog conversion channel and superimposed with the original audio signal in the analog domain.

6. The sound system according to claim 3, wherein: The adjustable parameter filter bank adopts a second-order infinite impulse response (IIR) notch filter structure. Its denominator polynomial contains a frequency tracking term to adapt to changes in motor speed in real time. The numerator polynomial is configured with mirror zeros to eliminate phase distortion caused by frequency offset. The filter coefficients are updated in real time according to the output of the dual-loop feedback control module.

7. A method for compensating for electromagnetic interference of audio signals in electric vehicles, using the sound system according to any one of claims 1 to 6 to achieve compensation, characterized in that: include: Step S1, signal acquisition: real-time acquisition of the interfered audio signal output by the power amplifier, and synchronous acquisition of the motor speed f through the vehicle CAN bus motor and PWM switching frequency f sw ; Step S2, feature extraction: Perform FFT waterfall analysis on the audio signal to extract the characteristic frequency points related to the motor harmonics in the 8kHz-20kHz frequency band, satisfying the relationship f k =f sw ±n×f motor ; Step S3, interference quantification: calculation of electromagnetic pollution index When the index exceeds -40dB, the compensation mechanism is triggered; Step S4, dynamic compensation: generating an anti-phase compensation signal that matches the characteristic frequency, and adjusting the signal amplitude coefficient γ and phase offset φ in real time through a dual-loop control mechanism; Step S5, signal injection: injecting the compensation signal into the front-end stage of the audio processing chain, superimposing it with the original audio and then transmitting it to the power amplifier.

8. The method according to claim 7, wherein: Step S2 adopts an adaptive window length FFT analysis strategy. When the motor speed change rate exceeds 100 rpm / s, a 10 millisecond short time window is enabled to quickly capture transient interference characteristics; when the speed change rate is lower than 10 rpm / s, a 100 millisecond long time window is used to improve the frequency resolution to accurately separate dense harmonic components.

9. The method according to claim 7, wherein: The dual-loop control of step S4 includes: Inner loop amplitude control: The deviation between the electromagnetic pollution index and the target threshold of -40dB is used as the input of the PID controller, and the output amplitude coefficient γ; Outer loop phase optimization: The measured THD value at the power amplifier output is used as the loss function, and the gradient descent algorithm is used to iteratively update the phase offset φ. The step size of each iteration is adaptively adjusted according to the THD change rate.

10. The method according to claim 7, wherein: The compensation signal generation in step S4 adopts a power limitation strategy, the total power of the compensation signal does not exceed 15% of the original audio signal power, and the compensation amount of a single frequency point |C(f k )|satisfies the constraint|C(f k )|≤0.5×|X(f k )|, to avoid audio distortion caused by overcompensation.

Images