Automatic evaluation system and method for vehicle-mounted sound reinforcement system

Abstract

The invention provides an automatic evaluation system and method for a vehicle-mounted sound reinforcement system. Efficient and objective whole vehicle acoustic performance evaluation is realized through binaural microphone multi-dimensional acquisition, pulse response analysis and a dynamic weighted scoring model. According to the system, a binaural microphone is fixed by a multi-dimensional bracket with a height and distance adjusting mechanism, and the head posture of a passenger is accurately simulated; a 20Hz-20kHz sweep frequency signal and an 18-24kHz near ultrasonic signal are driven in a time-sharing manner through a loudspeaker controller, and crosstalk is suppressed in combination with a time division multiplexing technology; six indexes such as frequency response flatness, reverberation time, total harmonic distortion and a speech transmission index are extracted based on an impulse response file, a centesimal system score is generated by adopting reverse normalization and dynamic weight adjustment, and the problems of spatial distortion, high-frequency detection blind areas, subjective and objective disjunction and the like of traditional single-point measurement are effectively solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of audio analysis, in particular to an automatic evaluation system and method for a vehicle-mounted sound reinforcement system. BACKGROUND

[0002] With the rapid development of new energy vehicle technology and the continuous improvement of consumers' demand for vehicle-mounted sound quality, the vehicle-mounted audio system has evolved from a basic sound playback device to a professional audio system with high fidelity characteristics. "Sound quality" is gradually becoming the focus of differentiation competition for vehicle manufacturers. However, the existing evaluation technology for vehicle-mounted sound reinforcement systems still has significant limitations.

[0003] Traditional evaluation methods rely on single-point microphone measurements (such as sound level meters and spectrum analyzers), which can only obtain local sound pressure levels and frequency response curves, and cannot simulate the real listening state of the human ear. Especially in a small and complexly reflective car cabin, the superposition of sound waves leads to standing waves and phase interference, making it difficult for single-point measurements to capture the sound pressure balance and sound image positioning deviation of multiple seats. Industry standards such as "Objective Test Method for Vehicle-mounted Audio Passenger Car Sound Reinforcement System" define indicators such as in-cabin sound field non-uniformity, but the testing efficiency is low and relies on manual point placement, making it difficult to achieve dynamic spatial mapping.

[0004] Sound quality evaluation has long relied on subjective scoring by expert listening teams, but individual preferences result in significant evaluation dispersion. Despite the introduction of psychoacoustic parameters, there is still a lack of dynamic mapping models with subjective listening. Traditional acoustic evaluation methods provide calculation methods for each indicator, but fail to distinguish the importance of each indicator and provide overall evaluation.

[0005] To break through the above bottlenecks, the industry urgently needs an automated evaluation solution that integrates high-precision spatial sound field reconstruction, full-band defect diagnosis, and subjective and objective dynamic mapping. SUMMARY

[0006] To solve the above problems, the present application provides an automatic evaluation system for a vehicle-mounted sound reinforcement system, comprising a microphone, a loudspeaker, a control module, a storage module, an audio analysis module, an audio acquisition module, and an audio processing module; characterized in that:

[0007] The control module controls the loudspeaker to emit audio, the audio emitted by the loudspeaker is collected by the microphone and sent to the audio processing module through the audio acquisition module; the audio processing module performs preprocessing and utilizes the audio analysis module to perform automatic evaluation of the vehicle-mounted sound reinforcement system, obtaining a comprehensive evaluation index of the sound reinforcement quality of the vehicle-mounted sound reinforcement system.

[0008] The microphone is a binaural microphone; the binaural microphone is arranged on a multi-dimensional adjustment support; the multi-dimensional adjustment support is arranged on a seat in a passenger compartment of a vehicle; the multi-dimensional adjustment support is provided with a height adjustment mechanism and a spacing adjustment mechanism for adjusting the spacing between the left and right ear microphones of the binaural microphone, so that the binaural microphone can simulate the listening state of a passenger at different positions in the passenger compartment.

[0009] The loudspeaker is controlled by a loudspeaker controller; the loudspeaker controller is communicatively connected to the control module and receives audio signal output instructions and loudspeaker control parameters from the control module;

[0010] The loudspeaker controller is used to control the amplitude and frequency characteristics of the audio signal input to the loudspeaker, control the switching between the working state, standby state and mute state of the loudspeaker, control the output sound field mode of the loudspeaker to select single-channel, stereo or multi-channel audio signal input, monitor the key working parameters of the loudspeaker, and perform loudspeaker protection actions in response to control module instructions or detected abnormal states;

[0011] The loudspeaker controller feeds back the running state and diagnostic data of the loudspeaker to the control module in real time.

[0012] The loudspeaker controller and the audio processing module cooperatively realize impulse response (IR) based sound reinforcement quality evaluation, specifically including:

[0013] The loudspeaker controller is used to generate high-precision test signals and monitor hardware states, specifically including:

[0014] Test signal generation: receiving control module instructions, driving the loudspeaker to emit full-band test signals, including 20Hz-20kHz logarithmic sweep signals or 18-24kHz band-limited near-ultrasonic signals, dynamic range ≤100dB, total harmonic distortion (THD) <0.1%;

[0015] Multi-channel synchronous control: time division multiplexing technology is used to drive each loudspeaker in time division, avoiding inter-channel crosstalk;

[0016] State protection mechanism: real-time monitoring of loudspeaker current, temperature and impedance parameters, prediction of voice coil overheating risk through a thermal power consumption model, triggering of dynamic compression or mute protection when THD>5% or temperature>90℃, and feedback of diagnostic data to the control module.

[0017] The audio acquisition module is used to send the audio data collected by the microphone to the audio analysis module, to perform cross-correlation operation and Sine Sweep deconvolution algorithm through binaural microphone collected test signal response, and to generate 48kHz / 24bit precision IR files;

[0018] The audio processing module performs acoustic parameter extraction, and calculates four core indicators based on an IR file:

[0019] Frequency response flatness: 1 / 3 octave smoothed FFT amplitude spectrum within the 300Hz-10kHz frequency band, target tolerance ±3dB;

[0020] Reverberation time (RT60): linear fitting of the energy decay curve (EDC), target value <0.4s;

[0021] Total harmonic distortion (THD): fundamental and harmonic energy ratio, threshold value <1% @1kHz;

[0022] Speech transmission index (STIPA): modulation analysis, target value >0.6.

[0023] The audio analysis module obtains indicators including: volume balance and sound separation degree;

[0024] The volume balance is a parameter representing the difference in sound size and frequency response measured at four positions in the passenger cabin; the sound separation degree is calculated according to the degree of broadening of the measured audio frequency response curve relative to the played frequency response curve;

[0025] The audio analysis module generates the loudspeaker system quality comprehensive evaluation index according to the frequency response flatness, reverberation time, total harmonic distortion, speech transmission index, volume balance and sound separation degree.

[0026] The application also provides a method for automatically evaluating a loudspeaker system, comprising the following steps:

[0027] (a) Adjusting the posture of the binaural microphone at four positions in the passenger cabin through a multi-dimensional adjustment support;

[0028] (b) The loudspeaker controller drives the loudspeaker to emit full-band sweep signals and near-ultrasonic signals in time-sharing manner, and the binaural microphone synchronously collects acoustic responses;

[0029] (c) Extracting six indicators including frequency response flatness, reverberation time, total harmonic distortion, speech transmission index, volume balance and sound separation degree based on the impulse response file;

[0030] (d) Normalizing and dynamically weighting the six indicators, mapping them into a percentage score and outputting.

[0031] Step (b) specifically includes:

[0032] The loudspeaker controller generates a 20Hz-20kHz logarithmic sweep signal, the sweep frequency points are distributed according to 1 / 3 octave, and fine tuning is performed to ensure that the second and third harmonics of each frequency point do not overlap; at the same time, an 18-24kHz band-limited near-ultrasonic signal is generated for high-frequency sound field characteristic analysis;

[0033] Time-division multiplexing is used to control the time-sharing work of each speaker. The left front channel transmits the sweep signal, and the right front channel transmits in the silent period. The process is repeated in sequence to all channels. The silent period is used to avoid inter-channel crosstalk, and the independent channel components in the mixed response are separated by the cross-correlation algorithm.

[0034] A high-precision voltage-controlled crystal oscillator generates a synchronous clock signal, and a synchronous instruction is sent to all binaural microphones. The crystal oscillator waveform is periodically reset to eliminate time accumulation error.

[0035] The Sine Sweep deconvolution algorithm is performed on the collected sweep signal response to convert the time-domain signal into a 48kHz / 24bit precision impulse response file.

[0036] Step (c) is specifically:

[0037] Four core indicators are calculated based on the IR file:

[0038] Frequency response flatness: 1 / 3 octave smoothing FFT amplitude spectrum in the 300Hz-10kHz frequency band, target tolerance ±3dB;

[0039] Reverberation time (RT60): Linear fitting of energy decay curve (EDC), target value <0.4s;

[0040] Total harmonic distortion (THD): Fundamental and harmonic energy ratio, threshold value <1% @1kHz;

[0041] Speech transmission index (STIPA): Modulation and demodulation analysis, target value >0.6;

[0042] Binaural microphones are placed at the left front, right front, left rear, and right rear positions of the passenger compartment to measure the 1kHz sound pressure level at each position. The standard deviation of the sound pressure level at the four positions is calculated and normalized to the balance score.

[0043] The correlation coefficient ρ of the original sweep signal frequency response curve and the microphone collected frequency response curve in the 500Hz-5kHz frequency band is calculated, and the separation degree is defined as S=1-ρ.

[0044] Step (d) is specifically:

[0045] The maximum-minimum normalization method is used to map the six indicators to the [0,1] interval to eliminate dimensional differences. The total harmonic distortion (THD) and reverberation time (RT60) are normalized in reverse.

[0046] The weight coefficient is dynamically adjusted according to the vehicle-mounted acoustic priority model: The initial weight is preset according to the influence of the indicator on subjective listening, and if the normalized value of a certain indicator is lower than the safety threshold, the weight is automatically increased by 20% to highlight the impact of the defect.

[0047] The comprehensive score F is linearly converted into a percentage score.

[0048] The weight of each index is: frequency response flatness 0.15, reverberation time 0.10, total harmonic distortion 0.20, speech transmission index 0.25, volume balance 0.15, and sound separation degree 0.15.

[0049] According to the percentage score, four quality grades are divided:

[0050] 90-100 points: all six indexes are better than the target value, and the subjective listening feeling is transparent and hierarchical;

[0051] 75-89 points: the core index meets the standard, and the secondary index slightly deviates;

[0052] 60-74 points: part of the index does not meet the standard and needs to be optimized;

[0053] <60 points: multiple indexes deviate seriously, and the acoustic performance is unqualified.

[0054] The beneficial effects of the present application are:

[0055] The multi-dimensional support with height and spacing adjustment mechanism is adopted to fix the binaural microphone on the seat, the head posture and ear spacing of passengers in different positions are accurately simulated, the traditional vehicle-mounted acoustic test relies on a single-point microphone, and the real spatial listening feeling of the human ear cannot be reflected. The scheme realizes the full-seat sound field consistency mapping in the car for the first time by physically simulating the position of the human ear and combining binaural recording technology. The test shows that the method reduces the sound field positioning error to ≤3°, and can capture the subtle influence of seat material on high-frequency reflection, and provides accurate data of human ear scale for sound field optimization.

[0056] The loudspeaker controller integrates high-precision sweep signal generation and multi-channel time-sharing driving, combines harmonic distortion analysis of impulse response, exposes design defects of frequency divider through near-ultrasonic band-limited signal, and suppresses channel crosstalk through time-sharing driving technology. IR analysis is combined to obtain evaluation indexes.

[0057] Based on the IR file, six indexes of frequency response flatness, RT60, THD, STIPA, volume balance and sound separation degree are extracted, percentage scores are generated through reverse normalization and defect triggering type dynamic weight, weight adjustment is performed, the correlation between the score and the subjective evaluation of professional sound engineers reaches 0.91. BRIEF DESCRIPTION OF DRAWINGS

[0058] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0059] Appendix Fig. 1 This is a schematic diagram showing the arrangement of the microphone and speaker in this invention;

[0060] Appendix Fig. 2 This is a schematic diagram of the shape of the binaural microphone used in this invention.

[0061] Appendix Fig. 3 This is a schematic diagram of the evaluation system of the present invention. Detailed Implementation

[0062] Example 1:

[0063] See Figs. 1 to 3 This invention provides an automatic evaluation system for vehicle-mounted sound reinforcement systems, comprising a microphone, a speaker, a control module, a storage module, an audio analysis module, an audio acquisition module, and an audio processing module; characterized in that:

[0064] The control module controls the speaker to emit audio, which is then picked up by the microphone and sent to the audio processing module via the audio acquisition module. After preprocessing, the audio processing module uses the audio analysis module to automatically evaluate the vehicle-mounted sound reinforcement system and obtain a comprehensive evaluation index of the sound reinforcement quality of the vehicle-mounted sound reinforcement system.

[0065] The microphone is a binaural microphone; the binaural microphone is mounted on a multi-dimensional adjustment bracket; the multi-dimensional adjustment bracket is mounted on the passenger seat in the vehicle's passenger compartment; the multi-dimensional adjustment bracket is equipped with a height adjustment mechanism and a spacing adjustment mechanism for adjusting the distance between the left and right earpieces of the binaural microphone, so that the binaural microphone can simulate the listening state of passengers in different positions in the passenger compartment.

[0066] The loudspeaker is controlled by a loudspeaker controller; the loudspeaker controller is communicatively connected to the control module and receives audio signal output commands and loudspeaker control parameters from the control module.

[0067] The speaker controller is used to control the amplitude and frequency characteristics of the audio signal input to the speaker, control the switching of the speaker between working state, standby state, and mute state, control the output sound field mode of the speaker to select mono, stereo, or multi-channel audio signal input, monitor the key operating parameters of the speaker, and perform speaker protection actions in response to control module commands or detected abnormal states.

[0068] The loudspeaker controller feeds the loudspeaker operating status and diagnostic data to the control module in real time.

[0069] The loudspeaker controller cooperates with the audio processing module to realize impulse response (IR) based sound reinforcement quality evaluation, specifically including:

[0070] The loudspeaker controller is used to generate high-precision test signals and monitor hardware status, specifically including:

[0071] Test signal generation: receiving control module instructions, driving the loudspeaker to emit full-band test signals, including 20Hz-20kHz logarithmic sweep signals or 18-24kHz band-limited near-ultrasonic signals, dynamic range ≤100dB, total harmonic distortion (THD)<0.1%;

[0072] Multi-channel synchronous control: using time division multiplexing technology to drive each loudspeaker in time, avoiding inter-channel crosstalk;

[0073] State protection mechanism: real-time monitoring of loudspeaker current, temperature and impedance parameters, predicting voice coil overheating risk through a thermal power consumption model, triggering dynamic compression or mute protection when THD>5% or temperature>90℃, and feeding diagnostic data to the control module.

[0074] The audio acquisition module is used to send audio data collected by the microphone to the audio analysis module, to collect test signal responses through binaural microphones, to perform cross-correlation operation and Sine Sweep deconvolution algorithm, and to generate 48kHz / 24bit precision IR files;

[0075] The audio processing module performs acoustic parameter extraction and calculates four core indicators based on the IR files:

[0076] Flatness of frequency response: 1 / 3 octave smoothed FFT amplitude spectrum in the 300Hz-10kHz frequency band, target tolerance ±3dB;

[0077] Reverberation time (RT60): linear fitting of energy decay curve (EDC), target value <0.4s;

[0078] Total harmonic distortion (THD): fundamental and harmonic energy ratio, threshold value <1% @1kHz;

[0079] Speech transmission index (STIPA): modulation and demodulation analysis, target value >0.6.

[0080] The audio analysis module obtains indicators including volume balance and sound separation degree;

[0081] wherein the volume balance is a parameter representing the difference in the magnitude and frequency response of the sound measured at the four positions in the passenger cabin; the sound separation is calculated according to the extent of the spread of the measured audio frequency response curve relative to the played frequency response curve;

[0082] The audio analysis module generates the comprehensive evaluation index of the sound reinforcement quality according to the frequency response flatness, the reverberation time, the total harmonic distortion, the speech transmission index, the volume balance and the sound separation.

[0083] Embodiment 2:

[0084] The application also provides a method for automatically evaluating a sound reinforcement system, comprising the following steps:

[0085] (a) calibrating the posture of the binaural microphone at the four positions in the passenger cabin through the multi-dimensional adjustment support;

[0086] (b) driving the loudspeakers to emit full-band sweep signals and near-ultrasonic signals in time division, and synchronously collecting the acoustic responses by the binaural microphone;

[0087] (c) extracting the six indexes of the frequency response flatness, the reverberation time, the total harmonic distortion, the speech transmission index, the volume balance and the sound separation based on the impulse response file;

[0088] (d) dynamically weighting and summing the six indexes after normalization, mapping into a percentage score and outputting.

[0089] Step (b) specifically comprises:

[0090] The loudspeaker controller generates a 20Hz-20kHz logarithmic sweep signal, the sweep frequency points are distributed according to 1 / 3 octave, and the second and third harmonics of each frequency point are ensured not to overlap through fine tuning; at the same time, an 18-24kHz band-limited near-ultrasonic signal is generated for high-frequency sound field characteristic analysis;

[0091] Time division multiplexing is used to control the time-division work of each loudspeaker, the left front channel emits the sweep signal, the right front channel emits in the silent period, and the cycle is sequentially repeated to all channels; the silent period is used to avoid inter-channel crosstalk, and the independent channel components in the mixed response are separated through the cross-correlation algorithm;

[0092] A high-precision voltage-controlled crystal oscillator generates a synchronous clock signal, and the synchronous instruction is sent to all binaural microphones, and the crystal oscillator waveform is periodically reset to eliminate the time accumulation error;

[0093] The Sine Sweep deconvolution algorithm is performed on the collected sweep signal response, and the time-domain signal is converted into a 48kHz / 24bit precision impulse response file.

[0094] Step (c) specifically comprises:

[0095] Four core indicators are calculated based on the IR file:

[0096] Frequency response flatness: 1 / 3 octave smoothed FFT amplitude spectrum within 300Hz-10kHz band, target tolerance ±3dB;

[0097] Reverberation time (RT60): linear fitting of energy decay curve (EDC), target value <0.4s;

[0098] Total harmonic distortion (THD): fundamental to harmonic energy ratio, threshold value <1% @1kHz;

[0099] Speech transmission index (STIPA): modulation analysis, target value >0.6;

[0100] Dual-ear microphones are placed at the left front, right front, left rear, and right rear positions in the passenger cabin to measure the 1kHz sound pressure level at each position; the standard deviation of the sound pressure levels at the four positions is calculated and normalized as the balance score;

[0101] The correlation coefficient ρ of the original sweep signal frequency response curve and the microphone-acquired frequency response curve is calculated within the 500Hz-5kHz band, and the separation degree is defined as S=1-ρ.

[0102] Step (d) is specifically:

[0103] The maximum-minimum normalization method is used to map the six indicators to the [0,1] interval to eliminate dimensional differences; among them, the total harmonic distortion (THD) and the reverberation time (RT60) are normalized in reverse;

[0104] The weight coefficients are dynamically adjusted according to the vehicle-mounted acoustic priority model: the initial weight is preset according to the influence of the indicators on subjective listening, and if the normalized value of an indicator is lower than the safety threshold, its weight is automatically increased by 20% to highlight the impact of the defect;

[0105] The comprehensive score F is linearly converted to a percentage score.

[0106] The weights of each indicator are: frequency response flatness 0.15, reverberation time 0.10, total harmonic distortion 0.20, speech transmission index 0.25, volume balance 0.15, and sound separation degree 0.15;

[0107] According to the percentage score, four quality levels are divided:

[0108] 90-100 points: all six indicators are better than the target value, the subjective listening is transparent and the levels are clear;

[0109] 75-89 points: core indicators meet the standards, and secondary indicators are slightly deviated;

[0110] 60-74 points: some indicators do not meet the standards and need to be optimized;

[0111] <60 points: multiple indicators deviate seriously, acoustic performance unqualified.

[0112] The specific methods of the six indicators of frequency response flatness, reverberation time, total harmonic distortion, speech transmission index, volume balance, and sound separation are as follows:

[0113] Frequency response flatness:

[0114] The amplitude spectrum of the 300 Hz-10 kHz frequency band is extracted by performing 1 / 3 octave smoothing on the IR file, and the arithmetic mean of the amplitudes of each frequency point is calculated as the reference value.

[0115] The frequency response flatness is defined as the absolute deviation mean of the amplitudes of each frequency point from the reference value, and the formula is:

[0116]

[0117] where yREF is the average amplitude of the 1 kHz-3 kHz reference frequency band, y bandn is the average amplitude of the nth octave band, and N is the total number of frequency bands

[0118] Reverberation time (RT60):

[0119] The Schroeder inverse integral is performed on the IR file to generate the sound energy decay curve (EDC), and the formula is

[0120] E(t) = ∫ t ∞ h 2 (τ)d;

[0121] where h(τ) is the impulse response signal, and the slope absolute value inverse in the interval where the decay curve drops by 30 dB is the RT60 value.

[0122] Total harmonic distortion (THD):

[0123] The amplitude A1 of the 1 kHz fundamental signal, and the second and third harmonic amplitudes A2 and A3 are extracted from the IR file; the total harmonic distortion calculation formula is:

[0124]

[0125] Speech transmission index:

[0126] The modulation transfer function (MTF) is generated based on the IR file, and the modulation degree attenuation value of 14 modulation frequencies (0.5 Hz-12.5 Hz) is calculated for 7 carrier frequencies (0.63 kHz-8 kHz); the STIPA value is calculated by integrating the results of each frequency band, and the formula is:

[0127]

[0128] where w i are frequency band weight coefficients.

[0129] The above description of the embodiments has been provided for the purpose of illustration and description. It is not intended to be exhaustive or to limit the disclosure to the precise forms disclosed. Individual elements or features of a particular embodiment are generally not limited to that particular embodiment, but are interchangeable with other

[0130] The example embodiments are provided so that this disclosure will be thorough, and will fully convey the scope to those who are skilled in the art. Numerous specific details are set forth such as examples of specific parts, devices, and methods, to provide a thorough understanding of the embodiments of the present disclosure. It will be apparent to those skilled in the art that specific details need not be employed, that the embodiments can be practiced in many different

[0131] terminology is used herein for the purpose of describing particular embodiments and is not intended to be limiting as such terms can encompass multiple meanings that are appreciated by those skilled in the art. Whenever a conditional language is used, e.g., "can," "could," "might," "may," "e.g.," "for example," "such as," it means that there are useful alternatives. The use of such conditional language is not to be construed to mean that there are not also many other alternatives. The terms "comprise," "comprising," "include," "including," and "includes" are inclusive and therefore specify the presence of stated features, integers, steps, operations, elements, and / or components but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. The mere fact that measures, steps or features are recited in

[0132] The order of execution or performance of the operations in the examples illustrated is not essential, unless otherwise specified. Further, the examples are not limited to the specific

[0133] It is understood that additional or alternative steps can be employed.

Claims

1. An automatic evaluation system for a vehicle-mounted sound reinforcement system, comprising a microphone, a speaker, a control module, a storage module, an audio analysis module, an audio acquisition module, and an audio processing module; characterized in that: The control module controls the speaker to emit audio, which is then picked up by the microphone and sent to the audio processing module via the audio acquisition module. After preprocessing, the audio processing module uses the audio analysis module to automatically evaluate the vehicle-mounted sound reinforcement system and obtain a comprehensive evaluation index of the sound reinforcement quality of the vehicle-mounted sound reinforcement system.

2. The automatic evaluation system for vehicle-mounted sound reinforcement systems according to claim 1, characterized in that: The microphone is a binaural microphone; the binaural microphone is mounted on a multi-dimensional adjustment bracket; the multi-dimensional adjustment bracket is mounted on the passenger seat in the vehicle's passenger compartment; the multi-dimensional adjustment bracket is equipped with a height adjustment mechanism and a spacing adjustment mechanism for adjusting the distance between the left and right earpieces of the binaural microphone, so that the binaural microphone can simulate the listening state of passengers in different positions in the passenger compartment.

3. The automatic evaluation system for vehicle-mounted sound reinforcement systems according to claim 1 or 2, characterized in that: The loudspeaker is controlled by a loudspeaker controller; the loudspeaker controller is communicatively connected to the control module and receives audio signal output commands and loudspeaker control parameters from the control module. The speaker controller is used to control the amplitude and frequency characteristics of the audio signal input to the speaker, control the switching of the speaker between working state, standby state, and mute state, control the output sound field mode of the speaker to select mono, stereo, or multi-channel audio signal input, monitor the key operating parameters of the speaker, and perform speaker protection actions in response to control module commands or detected abnormal states. The speaker controller feeds back the speaker's operating status and diagnostic data to the control module in real time.

4. The automatic evaluation system for vehicle-mounted sound reinforcement systems according to claim 3, characterized in that: The speaker controller and audio processing module work together to achieve sound reinforcement quality evaluation based on impulse response, or IR, specifically including: The speaker controller is used to generate high-precision test signals and monitor hardware status, specifically including: Test signal generation: Receives instructions from the control module and drives the speaker to emit a full-band test signal, including a 20Hz–20kHz logarithmic sweep signal or an 18–24kHz band-limited near-sonic signal, with a dynamic range ≤100dB and a total harmonic distortion (THD) <0.1%. Multi-channel synchronous control: Time-division multiplexing technology is used to drive each speaker in a time-division manner to avoid crosstalk between channels; Status protection mechanism: Real-time monitoring of speaker current, temperature and impedance parameters, prediction of voice coil overheating risk through thermal power consumption model, triggering dynamic compression or mute protection when THD>5% or temperature>90℃, and feeding back diagnostic data to the control module; The audio acquisition module is used to send the audio data collected by the microphone to the audio analysis module. It collects test signal responses through binaural microphones, performs cross-correlation calculations and Sine Sweep deconvolution algorithm, and generates an IR file with 48kHz / 24bit precision. The audio processing module extracts acoustic parameters and calculates four core metrics based on the IR file: Frequency response flatness: 1 / 3 octave smoothed FFT amplitude spectrum within the 300Hz–10kHz frequency band, with a target tolerance of ±3dB; Reverberation time (RT60): Linear fitting of the energy decay curve (EDC), target value <0.4s; Total Harmonic Distortion (THD): The ratio of fundamental frequency to harmonic energy, threshold <1%@1kHz; Speech Transmission Index (STIPA): Modulation and demodulation analysis, target value >0.

6.

5. The automatic evaluation system for vehicle-mounted sound reinforcement systems according to claim 4, characterized in that: The audio analysis module acquires metrics including volume balance and sound separation. Volume balance is a parameter characterizing the differences in the volume and frequency response of the sound measured at four locations within the passenger cabin; sound separation is calculated based on the degree of broadening of the measured audio frequency response curve relative to the played frequency response curve. The audio analysis module generates the comprehensive evaluation index of sound reinforcement quality based on frequency response flatness, reverberation time, total harmonic distortion, speech transmission index, volume balance, and sound separation.

6. The method for automatically evaluating a sound reinforcement system according to any one of claims 1-5, characterized in that... Includes the following steps: (a) Calibrate the attitude of the binaural microphones in four positions in the crew cabin by adjusting the multidimensional adjustment bracket; (b) The loudspeaker controller drives the loudspeaker to emit full-band sweep frequency signals and near-ultrasonic signals in a time-division manner, and the binaural microphones simultaneously collect the acoustic response; (c) Extract six indicators based on impulse response files: frequency response flatness, reverberation time, total harmonic distortion, speech transmission index, volume balance, and sound separation. (d) After normalizing the six indicators, dynamically weight and sum them to map them into a percentage score and output it.

7. The automatic evaluation method according to claim 6, characterized in that: Step (b) specifically includes: The loudspeaker controller generates a 20Hz–20kHz logarithmic sweep signal with the sweep frequency points distributed in 1 / 3 octave bands. Fine-tuning is used to ensure that the second and third harmonics of each frequency point do not overlap. At the same time, it generates a 18–24kHz band-limited near-ultrasonic signal for high-frequency sound field characteristic analysis. Time-division multiplexing is used to control each speaker to work in a time-division manner. The left front channel emits a sweep frequency signal, and during the silence period, the right front channel emits the signal, and so on, cycling through all channels in sequence. The silence period is used to avoid crosstalk between channels, and the independent channel components in the mixed response are separated by a cross-correlation algorithm. A high-precision voltage-controlled crystal oscillator generates a synchronous clock signal, which is then sent to all binaural microphones. The crystal oscillator waveform is periodically reset to eliminate accumulated time errors. The Sine Sweep deconvolution algorithm is applied to the acquired swept frequency signal response to convert the time-domain signal into a pulse response file with a precision of 48kHz / 24bit.

8. The automatic evaluation method according to claim 6, characterized in that: Step (c) specifically involves: Calculate four core metrics based on IR files: Frequency response flatness: 1 / 3 octave smoothed FFT amplitude spectrum within the 300Hz–10kHz frequency band, with a target tolerance of ±3dB; Reverberation time (RT60): Linear fitting of the energy decay curve (EDC), target value <0.4s; Total Harmonic Distortion (THD): The ratio of fundamental frequency to harmonic energy, threshold <1%@1kHz; Speech Transmission Index (STIPA): Modulation and demodulation analysis, target value >0.6; Binocular microphones were placed at four positions in the crew cabin: left front, right front, left rear, and right rear. The sound pressure level at 1kHz was measured at each position. The standard deviation of the sound pressure level at the four positions was calculated and normalized to a balance score. By comparing the frequency response curve of the original swept frequency signal with the frequency response curve acquired by the microphone, the correlation coefficient ρ of the two curves in the 500Hz–5kHz frequency band is calculated, and the separation degree is defined as: S=1-ρ.

9. The automatic evaluation method according to claim 8, characterized in that: Step (d) specifically involves: The maximum-minimum normalization method was used to map the six indicators to the [0,1] interval to eliminate dimensional differences; in particular, the total harmonic distortion (THD) and reverberation time (RT60) were normalized by inverse normalization. The weighting coefficients are dynamically adjusted based on the vehicle acoustic priority model: the initial weights are preset according to the degree of influence of the indicators on subjective hearing. If the normalized value of a certain indicator is lower than the safety threshold, its weight is automatically increased by 20% to highlight the impact of the defective item. The overall score F is linearly converted to a percentage score.

10. The automatic evaluation method according to claim 9, characterized in that: The weights of each indicator are as follows: frequency response flatness 0.15, reverberation time 0.10, total harmonic distortion 0.20, speech transmission index 0.25, volume balance 0.15, and sound separation 0.15; The quality is divided into four levels based on a percentage score: 90–100 points: All six indicators are better than the target value, and the subjective listening experience is clear and distinct. 75–89 points: Core indicators meet the standards, secondary indicators deviate slightly; 60–74 points: Some indicators did not meet the standards and need to be optimized accordingly; <60 points: Multiple indicators deviate significantly, and the acoustic performance is unqualified.

Images