Vehicle sealing performance detection method and device, vehicle, storage medium and product

Abstract

The application relates to a vehicle sealing performance detection method and device, a vehicle, a storage medium and a product. The method comprises the following steps: acquiring original acoustic signals collected by a main microphone array arranged in a to-be-detected area of a vehicle and reference noise signals collected by a reference microphone array arranged in a reference area of the vehicle; performing background noise filtering processing on the original acoustic signals based on the reference noise signals to obtain target acoustic signals; extracting target acoustic features corresponding to the target acoustic signals; determining a feature similarity between the target acoustic features and reference acoustic features corresponding to a current running state of the vehicle; and in a case where the feature similarity is smaller than a similarity threshold, determining a sealing performance detection result of the to-be-detected area according to the target acoustic features. The method can continuously and comprehensively detect the sealing performance of the vehicle during driving.

Description

Technical Field

[0001] This application relates to the field of automotive electronic control technology, and in particular to a method, apparatus, vehicle, storage medium, and product for testing vehicle sealing performance. Background Technology

[0002] With the continuous improvement of automotive intelligence, a vehicle's self-diagnostic capabilities have become an important indicator of its technological advancement. During vehicle operation, sealing performance is a key factor affecting ride comfort. Aging, deformation, or improper installation of sealing strips in areas such as windows, doors, and sunroofs can all lead to abnormal sealing performance.

[0003] Currently, vehicle sealing performance testing is usually conducted in a laboratory environment, which not only suffers from low testing efficiency and high subjectivity, but also cannot be continuously monitored during vehicle operation, and urgently needs to be addressed. Summary of the Invention

[0004] Based on this, this application addresses the aforementioned technical problems by providing a vehicle sealing performance testing method, apparatus, vehicle, storage medium, and product capable of continuously and comprehensively testing the sealing performance of a vehicle during vehicle operation.

[0005] Firstly, this application provides a method for testing the sealing performance of a vehicle, including:

[0006] Acquire the raw acoustic signal collected by the main microphone array deployed in the area to be detected of the vehicle, and the reference noise signal collected by the reference microphone array deployed in the reference area of ​​the vehicle.

[0007] Based on the reference noise signal, the original acoustic signal is subjected to background noise filtering to obtain the target acoustic signal;

[0008] Extract the target acoustic features corresponding to the target acoustic signal;

[0009] Determine the feature similarity between the target acoustic features and the reference acoustic features corresponding to the current operating state of the vehicle;

[0010] When the feature similarity is less than the similarity threshold, the sealing performance test result of the area to be tested is determined based on the target acoustic features.

[0011] In the aforementioned vehicle sealing performance testing method, the original acoustic signal collected by the main microphone array deployed in the test area of ​​the vehicle, and the reference noise signal collected by the reference microphone array deployed in the reference area of ​​the vehicle are obtained. Based on the reference noise signal, the original acoustic signal is subjected to background noise filtering to obtain the target acoustic signal. The target acoustic features corresponding to the target acoustic signal are extracted, and the feature similarity between the target acoustic features and the baseline acoustic features corresponding to the current operating state of the vehicle is determined. If the feature similarity is less than the similarity threshold, the sealing performance test result of the test area is determined based on the target acoustic features. In this process, on the one hand, filtering out background noise from the original acoustic signal based on the reference noise signal can effectively reduce the influence of the vehicle's inherent noise on the target acoustic features, thereby improving the accuracy of the sealing performance test results. On the other hand, different operating states of the vehicle correspond to different reference acoustic features. By using the relationship between the feature similarity and similarity threshold between the target acoustic features and the reference acoustic features corresponding to the current operating state of the vehicle, and by using the target acoustic features to test the sealing performance, the vehicle sealing performance test is not limited to static testing. Furthermore, the existing microphone array deployed in the vehicle can be fully utilized to collect acoustic signals, reducing the hardware deployment cost and vehicle modification workload caused by adding additional dedicated detection sensors.

[0012] In an optional embodiment of the first aspect, determining the sealing performance test result of the area to be tested based on the target acoustic features includes: determining the sound continuity of the target acoustic signal based on the time-domain features in the target acoustic features; and determining the energy distribution of the target acoustic signal based on the frequency-domain features in the target acoustic features; and determining the sealing performance test result of the area to be tested based on the sound continuity and / or energy distribution.

[0013] In the above optional embodiments, the signal characteristics of the target acoustic signal are analyzed from two dimensions: time domain characteristics and frequency domain characteristics. Then, the sealing performance test results are determined based on the signal characteristics, which can improve the accuracy of the sealing performance test results to a certain extent.

[0014] In an optional embodiment of the first aspect, the sealing performance test result includes a sealing anomaly type; determining the sealing performance test result of the area to be tested based on the continuity of sound and the energy distribution includes: determining the sealing anomaly type as air leakage when the continuity of sound represents a continuous acoustic signal and the energy distribution represents a first signal; determining the sealing anomaly type as liquid leakage when the continuity of sound represents a discontinuous acoustic signal and the energy distribution represents a second signal; wherein the energy of the first signal in a preset frequency band per unit time exceeds an energy threshold; the energy of the second signal in a preset frequency band per unit time does not exceed the energy threshold, and the instantaneous peak value of the energy is greater than the peak value threshold.

[0015] In the above optional embodiments, relying on two indicators, sound continuity and energy distribution, to distinguish specific sealing anomaly types can make the sealing performance test results more precise, facilitating subsequent targeted sealing performance optimization.

[0016] In an optional embodiment of the first aspect, when the sealing performance test result indicates an abnormal sealing in the area to be tested, the method further includes: obtaining the signal arrival time when each microphone in the main microphone array collects the original acoustic signal; constructing a geometric constraint model with the sound source location of the original acoustic signal as the independent variable and the time difference as the dependent variable, based on the time difference between the signal arrival times of different microphones and the deployment positions of different microphones; solving the geometric constraint model to obtain the sound source location of the original acoustic signal; and determining the location of the sealing abnormality based on the sound source location.

[0017] In the above optional embodiments, the sound source location of the original acoustic signal is estimated based on the arrival time of the original acoustic signal collected by each microphone. This enables the location of the sealing abnormality when the sealing performance test results indicate a sealing abnormality, facilitating subsequent fixed-point sealing performance optimization.

[0018] In an optional embodiment of the first aspect, when the sealing performance test result indicates a sealing abnormality in the area to be tested, and the sealing abnormality type is a leakage type, the method further includes: acquiring liquid detection data detected by a liquid sensor in the vehicle; and controlling the window corresponding to the sealing abnormality location to rise when the current driving speed of the vehicle is less than a speed threshold and the liquid detection data exceeds a liquid flow rate threshold.

[0019] In the above optional embodiments, combining liquid detection data and current driving speed to control the sealing of windows in abnormal locations can reduce the probability of liquid seeping into the cabin while ensuring safety.

[0020] In an optional embodiment of the first aspect, the original acoustic signal is subjected to background noise filtering processing based on the reference noise signal to obtain the target acoustic signal, including: performing weighted fusion processing on the reference noise signal to obtain the estimated noise signal of the main microphone array; and filtering out the estimated noise signal from the original acoustic signal to obtain the target acoustic signal.

[0021] In the above optional embodiments, the weighted fusion method is used to generate the estimated noise signal, which can improve the accuracy of the estimated noise signal by combining the distribution pattern of the actual vehicle operating noise, thereby better filtering out the vehicle's own noise in the original acoustic signal, so as to reduce the interference of the vehicle's own noise on the subsequent vehicle sealing performance test.

[0022] In an optional embodiment of the first aspect, extracting target acoustic features corresponding to the target acoustic signal includes: performing time-frequency analysis on the target acoustic signal to obtain a time-frequency spectrogram corresponding to the target acoustic signal; extracting time-domain features and frequency-domain features of the time-frequency spectrogram; and performing weighted fusion processing on the time-domain features and frequency-domain features to obtain the target acoustic features corresponding to the target acoustic signal.

[0023] In the above optional embodiments, determining the time-domain and frequency-domain features of the target acoustic signal and then performing weighted fusion to generate target acoustic features can enrich the signal information contained in the target acoustic features, improve the comprehensiveness of the features, and thus improve the detection accuracy of subsequent vehicle sealing performance testing.

[0024] In an optional embodiment of the first aspect, the method further includes: performing human voice detection on the target acoustic signal to obtain a target signal segment in the target acoustic signal containing human voice; wherein, in the process of weighted fusion processing of time-domain features and frequency-domain features, the feature weight of the target signal segment is lower than the feature weight of other signal segments in the target acoustic signal.

[0025] In the above optional embodiments, by reducing the feature weight of signal segments containing human voices, the interference of in-vehicle conversations on the target acoustic features can be weakened, the proportion of effective information in the target acoustic features can be increased, thereby improving the detection accuracy of subsequent vehicle sealing performance testing.

[0026] Secondly, this application also provides a vehicle sealing performance testing device, comprising:

[0027] The signal acquisition module is used to acquire the raw acoustic signal collected by the main microphone array deployed in the detection area of ​​the vehicle, and the reference noise signal collected by the reference microphone array deployed in the reference area of ​​the vehicle.

[0028] The noise filtering module is used to perform background noise filtering on the original acoustic signal based on the reference noise signal to obtain the target acoustic signal;

[0029] The feature extraction module is used to extract the target acoustic features corresponding to the target acoustic signal;

[0030] The sealing detection module is used to determine the feature similarity between the target acoustic features and the reference acoustic features corresponding to the current operating state of the vehicle, and to determine the sealing performance test result of the area to be tested based on the target acoustic features if the feature similarity is less than the similarity threshold.

[0031] Thirdly, this application also provides a vehicle including a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps of the methods described above.

[0032] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described above.

[0033] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the method described in any of the above aspects.

[0034] Regarding the beneficial effects of any of the technical solutions in the second to fifth aspects mentioned above, refer to the beneficial effects of the corresponding technical solutions in the first aspect; repeated examples will not be listed here. Attached Figure Description

[0035] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0036] Figure 1 This is a schematic diagram of an optional process for testing the sealing performance of a vehicle in one embodiment;

[0037] Figure 2 This is a schematic diagram of an optional main microphone deployment in one embodiment;

[0038] Figure 3 This is a schematic diagram of an optional reference microphone deployment in one embodiment;

[0039] Figure 4 This is a schematic diagram of an optional feature classification model structure in one embodiment;

[0040] Figure 5 This is a schematic diagram of an optional process for determining the location of a sealing anomaly in one embodiment;

[0041] Figure 6 This is a schematic diagram of an optional window control process in one embodiment;

[0042] Figure 7 This is a flowchart illustrating an optional target acoustic signal determination step in one embodiment;

[0043] Figure 8 This is a flowchart illustrating an optional step for extracting target acoustic features in one embodiment.

[0044] Figure 9 This is a flowchart illustrating an optional vehicle sealing performance testing method in another embodiment;

[0045] Figure 10 This is a schematic diagram of an optional vehicle sealing performance testing device in one embodiment;

[0046] Figure 11 This is a schematic diagram of the internal structure of an optional vehicle-mounted terminal in one embodiment. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of this application.

[0048] The terms "first," "second," etc., used in this application may be used to describe various elements, but these elements are not limited by these terms. These terms are used only to distinguish the first element from the second element. The terms "comprising" and "having," and any variations thereof, used in this application, are intended to cover non-exclusive inclusion. The term "multiple" used in this application refers to two or more. The term "and / or" used in this application refers to one of the embodiments, or any combination of multiple embodiments.

[0049] Currently, vehicle sealing performance testing has become an important means of improving driving comfort and ensuring overall vehicle quality. However, related technologies typically employ the following methods for testing vehicle sealing performance: First, manual inspection, where technicians visually inspect or conduct smoke tests in specific environments (such as rain chambers) to identify leaks; second, pressure decay testing, which assesses vehicle sealing performance by pressurizing the cabin and monitoring the rate of pressure change; third, sensor array testing, which detects leaks by deploying humidity or pressure sensors at key locations on the vehicle; and fourth, visual inspection, which uses cameras and image processing technology to detect traces of leakage. However, all of these methods have significant limitations: they either only allow testing when the vehicle is stationary, or only after a leak has occurred, or require the installation of sensors on the vehicle itself, failing to simultaneously achieve dynamic testing of vehicle sealing performance and improve the versatility of the testing.

[0050] In the field of automotive noise, vibration, and harshness (NVH) research, active noise cancellation and active road noise control technologies are relatively mature, and vehicles are typically equipped with acoustic monitoring devices (such as microphone arrays) for noise acquisition and control. Therefore, how to fully utilize these acoustic monitoring devices to achieve intelligent detection of sealing performance has become a research focus. The embodiments of this application are based on the acoustic signals collected by these acoustic monitoring devices to detect vehicle sealing performance.

[0051] In one exemplary embodiment, such as Figure 1 As shown, a method for testing the sealing performance of a vehicle is provided. Taking the application of this method to a vehicle, specifically to an on-board terminal in a vehicle, as an example, the method includes the following steps:

[0052] S101, acquire the raw acoustic signal collected by the main microphone array deployed in the detection area of ​​the vehicle, and the reference noise signal collected by the reference microphone array deployed in the reference area of ​​the vehicle.

[0053] The area to be tested can be understood as any area on the vehicle where poor sealing performance may affect the driving and riding experience. For example, the area to be tested is prone to air or liquid leakage after a seal failure; common areas to be tested include the vehicle cabin and trunk. The reference area can be understood as any area in the vehicle close to a noise source; for example, the reference area is the area where the noise source is located. Noise sources can include the main noise-generating vehicle components (such as the engine, tires, and brake pads), as well as the external environment in which the vehicle operates.

[0054] In some embodiments, a main microphone array is used to acquire acoustic signals within a detection area, and a reference microphone array is used to acquire acoustic signals within a reference area. The main microphone array includes multiple main microphone sensors; the reference microphone array includes multiple reference microphone sensors. This embodiment does not limit the model or specific deployment location of each main microphone and each reference microphone.

[0055] For example, see Figure 2 The diagram shown illustrates the deployment of the main microphone in a vehicle, illustrating its location. The main microphone can be deployed in locations including, but not limited to: near the A-pillars (including the left and right A-pillars), near the B-pillars (including the left and right B-pillars), near the C-pillars (including the left and right C-pillars), the center of the roof, the inner door panels (including the left front door, left rear door, right front door, and right rear door), the trunk, and around the sunroof—areas near potential leakage points.

[0056] For example, see Figure 3 The diagram shown illustrates the deployment of a reference microphone in a vehicle, illustrating the possible locations of the reference microphone. These locations may include, but are not limited to, areas near noise sources such as the vehicle's engine, tires, and the outer edge of the vehicle's roof.

[0057] Because the reference area is close to the vehicle's noise source, the noise level in the reference area is relatively high. Therefore, the corresponding reference noise signal can effectively characterize the vehicle's own baseline noise signal. The reference noise signal can include ambient noise signals unrelated to the leak, such as the vehicle's ambient noise, road noise, and engine noise during operation.

[0058] In one alternative implementation, the raw acoustic signal and reference noise signal can be acquired in real time or periodically after vehicle startup is detected. Exemplarily, signal feedback commands are sent to the main microphone array and reference microphone array in real time or periodically to acquire the raw acoustic signal and reference noise signal. The sending period of the signal feedback commands can be determined based on human experience or through extensive experimentation; this application does not impose any limitations on this.

[0059] In one embodiment, to ensure that the signal acquired by the main microphone sensor covers the audible range of the human ear and guarantees sufficient frequency resolution, the sampling frequency of the main microphone sensor can be set to 16kHz, which can completely reproduce acoustic signals with frequencies from 0 to 8kHz. That is, the effective acquisition bandwidth of the main microphone sensor is 0-8kHz. Correspondingly, to facilitate the subsequent step of filtering background noise from the original acoustic signal based on the reference noise signal to obtain the target acoustic signal, the acquisition frequency of the reference microphone sensor can also be configured to 16kHz.

[0060] Optionally, since the vehicle's own reference noise signal is usually concentrated in the 0-4kHz range, in order to reduce the amount of data in the reference noise signal, the sampling frequency of the reference microphone sensor can also be configured to 0-8kHz, which can completely reproduce the acoustic signal in the 0-4kHz frequency range.

[0061] S102, based on the reference noise signal, performs background noise filtering on the original acoustic signal to obtain the target acoustic signal.

[0062] The target acoustic signal can be understood as the acoustic signal obtained after filtering out the noise signal in the original acoustic signal, which can reduce the interference of the noise signal on the subsequent sealing performance test.

[0063] In a real cockpit environment, the raw acoustic signal acquired by the main microphone array typically contains background noise in addition to the target acoustic signal required for sealing performance testing. Similarly, the reference noise signal acquired by the reference microphone array typically contains both background noise and the target acoustic signal. However, because the reference microphone array is deployed near the noise source, the energy of the background noise signal it acquires is much higher than the target acoustic signal, thus the target acoustic signal acquired by the reference microphone array can be ignored. Based on this, the target acoustic signal can be obtained by filtering out background noise from the raw acoustic signal using subsequent steps based on the reference noise signal.

[0064] In one alternative implementation, frequency domain spectral subtraction can be used to filter background noise from the original acoustic signal based on a reference noise signal. For example, short-time Fourier transforms can be performed on the original acoustic signal and the reference noise signal respectively to obtain a first signal spectrum corresponding to the original acoustic signal and a second signal spectrum corresponding to the reference noise signal; the noise power spectral density is estimated based on the second signal spectrum, and the noise amplitude corresponding to each frequency point is determined using this noise spectral density as a reference; the estimated noise amplitude is subtracted from the first signal spectrum frequency by frequency point, and then subjected to an inverse short-time Fourier transform to generate the target acoustic signal after removing the vehicle body reference noise.

[0065] In another alternative implementation, an adaptive coherent filtering method can be used to filter background noise from the original acoustic signal based on a reference noise signal. For example, the reference noise signal and the original acoustic signal can be used as filter inputs. The correlation between the reference noise signal and the original acoustic signal is calculated frame by frame based on the filter, and the filter coefficients are updated in real time. This results in the filter outputting a signal component in the original acoustic signal that contains the vehicle reference noise. The original acoustic signal is then subtracted from this signal component to obtain the target acoustic signal after noise filtering.

[0066] To improve the accuracy of the target acoustic signal and the efficiency of the background noise filtering process, in some embodiments, the acquired raw acoustic signal can be preprocessed. The preprocessing method includes at least one of the following:

[0067] For example, the original acoustic signal is bandpass filtered to remove irrelevant frequency components below a first preset frequency and above a second preset frequency. Considering that the audible frequency range of the human ear is 20Hz-8kHz, and that frequencies below 20Hz are usually ultra-low frequency noises caused by vehicle resonance or road bumps, which do not carry leakage characteristics, the first preset frequency can be set to 20Hz and the second preset frequency can be set to 8kHz.

[0068] For example, voice activity detection is performed on the raw acoustic signal: when conversations between people inside the vehicle are detected, the acoustic signal during that period is labeled to facilitate subsequent analysis and processing.

[0069] For example, the raw acoustic signal is segmented: the continuous raw acoustic signal is divided into analysis frames of fixed duration according to a preset fixed duration. During subsequent sealing performance testing, each analysis frame can be analyzed sequentially. The fixed duration can be determined based on manual experience or through extensive experimentation; this application does not impose any limitations on it. For example, the fixed duration can be set to 20-40 ms.

[0070] S103, extract the target acoustic features corresponding to the target acoustic signal.

[0071] In one alternative implementation, a feature extraction model for extracting acoustic features can be pre-trained, and the target acoustic signal can be input into the feature extraction model to obtain the target acoustic features corresponding to the target acoustic signal.

[0072] The feature extraction model can be built based on common neural networks, which will not be elaborated here. When training the feature extraction model, sample acoustic signals can be input into the model to predict the sample acoustic features. Based on the difference between the sample acoustic features and the corresponding acoustic feature labels, the model parameters of the feature extraction model are adjusted to improve the model accuracy.

[0073] S104, determine the feature similarity between the target acoustic features and the reference acoustic features corresponding to the current operating state of the vehicle.

[0074] The vehicle's current operating status may include at least one of the following: vehicle speed, road conditions, and external weather conditions. Feature similarity is used to characterize feature differences; higher feature similarity indicates smaller feature differences, and lower feature similarity indicates larger feature differences.

[0075] In one embodiment, operating condition templates for the same model of vehicle under different reference operating states can be pre-constructed, with each template corresponding to the baseline acoustic features under the corresponding operating condition. For example, reference acoustic signals of the same model of vehicle under normal, leak-free conditions such as different vehicle speeds, different road conditions, and different external weather conditions can be collected. These reference acoustic signals can then be pre-processed to extract acoustic features, or statistical parameters such as feature vectors and covariance corresponding to the baseline acoustic features can be extracted as the baseline acoustic features for the corresponding operating state. For instance, before the vehicle leaves the factory or during the initial user period, reference acoustic signals of the vehicle under various normal operating conditions can be collected. Normal operating conditions include different vehicle speeds (0-120 km / h), different road surfaces (asphalt, cement, gravel), different weather conditions (sunny, cloudy), and different window opening / closing states. Statistical analysis of the acoustic features of the reference acoustic signals under each operating condition can then be performed to obtain the baseline acoustic features for that specific operating condition.

[0076] In one alternative implementation, a feature vector corresponding to the target acoustic feature and a feature vector corresponding to the reference acoustic feature can be constructed, and then the feature similarity between the target acoustic feature and the reference acoustic feature can be determined based on the distance between the two feature vectors. For example, the closer the two are, the higher the feature similarity; the farther the two are, the lower the feature similarity.

[0077] In some embodiments, the target acoustic features and the reference acoustic features can be input into the feature similarity determination model to obtain the feature similarity between the two.

[0078] The feature similarity determination model can be built based on common neural networks, which will not be elaborated here. When training the feature similarity determination model, the acoustic features of each group of samples can be input into the model to obtain the feature similarity prediction results between them. Based on the difference between the feature similarity prediction results and the similarity labels, the feature similarity determination model is trained to improve its accuracy.

[0079] S105, when the feature similarity is less than the similarity threshold, determine the sealing performance test result of the area to be tested based on the target acoustic features.

[0080] For example, if the feature similarity is less than a similarity threshold, an abnormal acoustic feature is identified, and the step of determining the sealing performance test result of the area to be tested based on the target acoustic feature is performed. The similarity threshold can be determined based on human experience or through extensive experimentation; this application does not impose any limitations on it.

[0081] Optionally, if the feature similarity is not less than the similarity threshold, it is determined that there are no abnormal acoustic features, and accordingly, the sealing performance test result of the area to be tested is determined to be without abnormality.

[0082] In one optional implementation, the sealing performance test result includes the type of sealing anomaly; the target acoustic features include time-domain features and frequency-domain features. Accordingly, the sound continuity of the target acoustic signal can be determined based on the time-domain features of the target acoustic features; and the energy distribution of the target acoustic signal can be determined based on the frequency-domain features of the target acoustic features; the sealing performance test result of the area to be tested can be determined based on the sound continuity and / or energy distribution.

[0083] Optionally, when determining the sealing anomaly type of the area to be detected solely based on time-domain characteristics, if the sound continuity characterizes the target acoustic signal as a continuous acoustic signal, then the sealing anomaly type is determined to be air leakage; if the sound continuity characterizes the target acoustic signal as a discontinuous acoustic signal, then the sealing anomaly type is determined to be liquid leakage. Liquid leakage can encompass rainwater leakage, water leakage, and leakage of other liquids that can flow into the vehicle.

[0084] Optionally, when determining the sealing anomaly type of the area to be detected solely based on frequency domain characteristics, if the energy distribution characterizes the target acoustic signal as a first signal, then the sealing anomaly type is determined to be an air leakage type; if the energy distribution characterizes the target acoustic signal as a second signal, then the sealing anomaly type is determined to be a liquid leakage type.

[0085] Optionally, when determining the sealing anomaly type of the area to be detected based on both time-domain and frequency-domain characteristics, if the target acoustic signal is a continuous acoustic signal and the energy distribution represents the target acoustic signal as a first signal, the sealing anomaly type is determined to be an air leakage type; if the target acoustic signal is a discontinuous acoustic signal and the energy distribution represents the target acoustic signal as a second signal, the sealing anomaly type is determined to be a liquid leakage type.

[0086] The first signal's energy per unit time in the preset frequency band exceeds the energy threshold; the second signal's energy per unit time in the preset frequency band does not exceed the energy threshold, and its instantaneous peak energy is greater than the peak value threshold. The preset frequency can be selected as the concentrated frequency of leaking airflow noise, 2000Hz-6000Hz. The energy threshold and peak value threshold can be calibrated using sample data from well-sealed vehicles of the same model under different driving conditions. For example, the average energy plus a margin in the 2000Hz-6000Hz frequency band under normal operating conditions can be used as the energy threshold; the maximum instantaneous energy value in this frequency band under normal operating conditions plus a safety margin can be used as the peak value threshold. For example, the energy threshold can be selected as 55dB; the peak value threshold can be selected as 68dB.

[0087] In the above process, on the one hand, the signal characteristics of the target acoustic signal are analyzed from the two dimensions of time domain characteristics and frequency domain characteristics, and the sealing performance test results are determined based on the signal characteristics, which can improve the accuracy of the sealing performance test results to a certain extent; on the other hand, the specific sealing anomaly types are distinguished by relying on the two indicators of sound continuity and energy distribution, which can make the sealing performance test results more refined and facilitate subsequent targeted sealing performance optimization.

[0088] In another alternative implementation, the target acoustic signal can be input into a pre-trained feature classification model to obtain a signal classification result. Based on the signal classification result, the type of sealing anomaly can be determined. The feature classification model can be a pre-trained convolutional neural network (CNN) or recurrent neural network (RNN) model. The feature classification model has pre-learned the sound characteristics of the air leakage sound (high-frequency hissing, with broadband noise in the spectral characteristics, concentrated energy in the 2000Hz-6000Hz range) and the sound characteristics of the raindrop impact sound (pulsating impact sound, with transient high-frequency components in the spectral characteristics), enabling it to perform signal classification based on the target acoustic signal.

[0089] See Figure 4The schematic diagram of the feature classification model shown above includes, from top to bottom, an input layer, a feature extraction layer, a temporal modeling layer, a fully connected classification layer, and an output layer. The system consists of several layers: an input layer to receive the target acoustic signal and convert it into a 128×100 Mel spectrogram; a feature extraction layer consisting of four cascaded convolutional blocks with 3×3 kernels of 32, 64, 128, and 256 kernels respectively, used to extract frequency domain features of anomalous acoustic characteristics from the Mel spectrogram for anomaly classification; a temporal modeling layer consisting of two bidirectional Long Short-Term Memory (LSTM) networks, each with 128 units, used to analyze the changes in sound over time for anomaly classification; a fully connected layer consisting of two layers with 512 and 256 units respectively, fusing all extracted features and then outputting the probabilities of each classification result through a Softmax activation function; and an output layer determining the classification result based on the probabilities of each classification result and outputting the sealing anomaly type.

[0090] In the aforementioned vehicle sealing performance testing method, the original acoustic signal collected by the main microphone array deployed in the test area of ​​the vehicle, and the reference noise signal collected by the reference microphone array deployed in the reference area of ​​the vehicle are obtained. Based on the reference noise signal, the original acoustic signal is subjected to background noise filtering to obtain the target acoustic signal. The target acoustic features corresponding to the target acoustic signal are extracted, and the feature similarity between the target acoustic features and the baseline acoustic features corresponding to the current operating state of the vehicle is determined. If the feature similarity is less than the similarity threshold, the sealing performance test result of the test area is determined based on the target acoustic features. In this process, on the one hand, filtering out background noise from the original acoustic signal based on the reference noise signal can effectively reduce the influence of the vehicle's inherent noise on the target acoustic features, thereby improving the accuracy of the sealing performance test results. On the other hand, different operating states of the vehicle correspond to different reference acoustic features. By using the relationship between the feature similarity and similarity threshold between the target acoustic features and the reference acoustic features corresponding to the current operating state of the vehicle, and by using the target acoustic features to test the sealing performance, the vehicle sealing performance test is not limited to static testing. Furthermore, the existing microphone array deployed in the vehicle can be fully utilized to collect acoustic signals, reducing the hardware deployment cost and vehicle modification workload caused by adding additional dedicated detection sensors.

[0091] In some embodiments, the reference acoustic features corresponding to the current operating state of the vehicle can also be obtained, and the feature differences between the target acoustic features and the reference acoustic features can be determined. Then, the sealing performance test results of the area to be tested can be obtained based on the feature difference results.

[0092] Optionally, feature vectors corresponding to the target acoustic features and reference acoustic features can be constructed, and then the feature differences between the target acoustic features and reference acoustic features can be determined based on the two feature vectors. For example, the Euclidean distance or cosine distance between the two can be determined, thereby quantifying the feature differences between the target acoustic features and reference acoustic features based on the distance magnitude.

[0093] Optionally, if the distance between the two exceeds a distance threshold, the sealing performance test result of the area to be tested is determined to be abnormal; if the distance between the two does not exceed the distance threshold, the sealing performance test result of the area to be tested is determined to be normal.

[0094] Based on the above embodiments, this application provides an optional method for determining the location of sealing abnormalities, such as... Figure 5 As shown, it includes the following steps:

[0095] S501, obtain the signal arrival time when each microphone in the main microphone array collects the original acoustic signal.

[0096] For example, for any microphone, the signal acquisition of the microphone can be absolutely clock-synchronized. At the instant the microphone begins to acquire the raw acoustic signal, an absolute timestamp is added, and the timestamp is used as the signal arrival time of the corresponding raw acoustic signal.

[0097] S502, based on the time difference between the arrival times of signals from different microphones and the deployment locations of different microphones, constructs a geometric constraint model with the sound source location of the original acoustic signal as the independent variable and the time difference as the dependent variable.

[0098] In some embodiments, since the microphones in the main microphone array are positioned differently, the arrival times of the original acoustic signal collected by different microphones for the original acoustic signal emitted from a sound source location are different. Therefore, the time difference between the arrival times of the corresponding signals when two main microphones collect the same original acoustic signal can be determined separately.

[0099] For example, the main microphone array can be configured to include M microphones, and the spatial coordinates of the i-th microphone are m. i =(x i y i , z i Let i = 1, 2, ..., M; and let the source location of the original acoustic signal be P = (x, y, z). Then, a geometric constraint model can be constructed with the source location of the original acoustic signal as the independent variable and the time difference as the dependent variable. By solving this geometric constraint model, the source location of the original acoustic signal can be determined.

[0100] ;

[0101] In the formula, P represents the location of the sound source; m i The m represents the spatial coordinates of the i-th microphone; j The j-th microphone represents its spatial coordinates; c represents the speed of light. This represents the signal arrival time difference between the i-th microphone and the j-th microphone.

[0102] S503 solves the geometric constraint model to obtain the sound source location of the original acoustic signal.

[0103] For example, the source location of the original acoustic signal can be determined using the least squares method or spherical interpolation. For instance, in solving a geometrically constrained model, the following solution method can be adopted:

[0104] Choosing any microphone as the first microphone i=1, for each second microphone j=2, 3, ..., M, we can construct M-1 hyperbolic equations:

[0105] ;

[0106] In the formula, This indicates the distance between the sound source location and the first microphone; c represents the distance between the sound source and the second microphone j; c represents the speed of light. This represents the signal arrival time difference between the first microphone and the j-th second microphone.

[0107] S504, determine the location of the sealing abnormality based on the location of the sound source.

[0108] In one alternative implementation, the coordinates of the sound source location can be directly used as the location of the sealing anomaly.

[0109] To facilitate users in identifying the location of sealing abnormalities, in another optional implementation, a mapping table between vehicle spatial coordinates and body parts can be pre-established. The obtained sound source location coordinates can be mapped to preset body parts of the vehicle to obtain a concrete location of the sealing abnormality, such as the upper edge of the left front window, the front edge of the sunroof, and the upper edge of the door frame.

[0110] In the above embodiments, the sound source location of the original acoustic signal is estimated based on the arrival time of the original acoustic signal collected by each microphone. This enables the location of the sealing abnormality when the sealing performance test results indicate a sealing abnormality, facilitating subsequent fixed-point sealing performance optimization.

[0111] To facilitate targeted maintenance of sealing performance abnormalities, some embodiments can comprehensively analyze the location and type of sealing abnormality to generate an anomaly diagnosis conclusion. For example, the anomaly level of the corresponding sealing abnormality location can be determined based on the number of anomaly diagnoses and the duration of the anomaly within a preset historical time period. For example, the possible causes and maintenance recommendations for the sealing abnormality at the corresponding location can be determined by comparing the sealing abnormality location with a pre-defined location-sealing abnormality cause comparison table. For instance, when the sealing abnormality location is the upper edge of the left front window and the sealing abnormality type is air leakage, a diagnosis conclusion can be generated: the sealing strip at the upper edge of the left front window may have a sealing failure, and it is recommended to check whether the sealing strip at this location is aged or deformed; as another example, when the sealing abnormality location is the front edge of the sunroof and the sealing abnormality type is liquid leakage, a diagnosis conclusion can be generated: the sealing strip at the front edge of the sunroof may have insufficient sealing performance, and it is recommended to check and replace the sealing strip at this location.

[0112] To facilitate users' understanding of sealing abnormalities, in some embodiments, the location of the sealing abnormality and the diagnostic conclusion can be displayed on the vehicle's central control screen, or the location and type of sealing abnormality can be prompted to the user through voice broadcast; or the diagnostic conclusion can be transmitted from the camera to the user's mobile terminal via the Internet of Vehicles.

[0113] Based on the above embodiments, this application provides an optional method for controlling vehicle windows, such as... Figure 6 As shown, it includes the following steps:

[0114] S601: When the sealing performance test results indicate an abnormality in the area to be tested, and the abnormality type is leakage, acquire the liquid detection data detected by the liquid sensor in the vehicle.

[0115] The liquid sensor can be a sensor used to detect liquid flow rate, such as, but not limited to, a rain sensor. The liquid detection data is used to characterize the real-time liquid flow rate in the vehicle's environment; for example, in a leak detection scenario, the liquid detection data can characterize the real-time rainfall intensity.

[0116] In one alternative implementation, when the sealing performance test results indicate a sealing abnormality in the area to be tested, and the sealing abnormality type is a leakage type, the liquid detection data detected by the liquid sensor in the vehicle can be directly obtained.

[0117] To improve vehicle control safety and prevent unauthorized activation of onboard accessories, in another optional implementation, if the sealing performance test results indicate a sealing abnormality in the area to be tested and the sealing abnormality type is leakage, it can be determined whether the user has authorized active control, and if the user has authorized it, the liquid detection data detected by the liquid sensor in the vehicle can be obtained.

[0118] S602, when the vehicle's current speed is less than the speed threshold and the liquid detection data exceeds the liquid flow threshold, controls the window corresponding to the abnormal sealing location to rise.

[0119] The velocity threshold and the liquid flow rate threshold can be determined based on human experience or through a large number of experiments. This application does not impose any restrictions on them.

[0120] In some embodiments, if the vehicle's current speed is less than a speed threshold and the liquid detection data exceeds a liquid flow threshold, it is considered that the vehicle may have a significant risk of leakage. In this case, a window control command is generated and sent to the window lift controller to control the window at the abnormal sealing position to rise, thereby mitigating the leakage.

[0121] In the above embodiments, combining liquid detection data and current driving speed to control the sealing of windows in abnormal locations can reduce the probability of liquid seeping into the cabin while ensuring safety.

[0122] Based on the above embodiments, this application provides a step for determining a target acoustic signal, such as... Figure 7 As shown, it includes the following steps:

[0123] S701 performs weighted fusion processing on the reference noise signal to obtain the estimated noise signal collected by the main microphone array.

[0124] S702 filters out the estimated noise signal from the original acoustic signal to obtain the target acoustic signal.

[0125] The predicted noise signal can be understood as the prediction result of the noise signal collected by the main microphone array.

[0126] The reference noise signal acquired by the reference microphone array is the original ambient noise. The original acoustic signal acquired by the main microphone array includes both the target acoustic signal and the noise signal. However, due to the different placement positions of the main and reference microphone arrays, the volume and phase of the noise change as it propagates to the main microphone array. In other words, the noise signal acquired by the main microphone array is different from the reference noise signal acquired by the reference microphone array. Under these circumstances, it is unreasonable to simply use the difference between the original acoustic signal and the reference noise signal as the target acoustic signal.

[0127] Based on the above reasons, in some embodiments, the propagation change can be simulated by weighting based on the acoustic propagation characteristics of adaptive noise reduction, thereby obtaining the estimated noise signal of the main microphone array. Subsequently, the estimated noise signal is removed from the original acoustic signal to obtain a relatively pure target acoustic signal.

[0128] For example, the original acoustic signal is composed of the following:

[0129] d(n) = s(n) + n(n);

[0130] In the formula, d(n) represents the original acoustic signal; s(n) represents the target acoustic signal; n(n) represents the actual noise signal after propagation path H; H is the unknown transmission path from the noise source to the main microphone array, including inversion, attenuation, and time delay, etc., and is the estimated noise signal in this embodiment. This is the estimated value of the actual noise signal n(n). Ideally, The value of n(n) is equal to that of n.

[0131] In one embodiment, the prediction of the noise signal can be based on an adaptive filtering algorithm. For example, the prediction method for the noise signal is as follows:

[0132] The reference noise signal is organized into a signal sequence according to the sampling time sequence. Where L is the length of the noise data involved in the weighted calculation, ranging from 64 to 256, with 128 selected as an example; based on a pre-determined set of weighting coefficients. By performing weighted summation on each reference noise signal, the following predicted noise signal is obtained:

[0133] ;

[0134] In the formula, Indicates the estimated noise signal; This represents the weighting coefficient corresponding to the i-th segment of the reference noise signal; This represents the reference noise signal obtained by downsampling i sampling times prior to the current sampling time.

[0135] Using the original acoustic signal d(n) and the predicted noise signal Perform the difference operation to obtain the error signal. This error signal is the target acoustic signal after initial background noise filtering. To reduce the error and improve the accuracy of the target acoustic signal, the Normalized Least Mean Square (NLMS) algorithm can be used to correct the entire set of weighting coefficients in real time.

[0136] ;

[0137] In the formula, This represents the weighting coefficients after one round of adjustments; This represents the weighting coefficient corresponding to the reference noise signal at time n; This indicates the step size for weight adjustment, ranging from 0.1 to 0.5. Indicates the error signal; The time-series vector representing the reference noise signal; This represents the minimum regularity coefficient, with a value ranging from 0.001 to 0.01.

[0138] Continue the above weight coefficient update operation in a loop until the weight coefficients can no longer be optimized, and then reset the corresponding weight coefficients. As the predicted noise signal for the main microphone array, the corresponding weighting coefficients are... As the target acoustic signal.

[0139] In the above embodiments, the weighted fusion method is used to generate the estimated noise signal, which can improve the accuracy of the estimated noise signal by combining the distribution pattern of the actual vehicle operating noise, thereby better filtering out the vehicle's own noise in the original acoustic signal, so as to reduce the interference of the vehicle's own noise on the subsequent vehicle sealing performance test.

[0140] Based on the above embodiments, this application provides a step for extracting target acoustic features, such as... Figure 8 As shown, it includes the following steps:

[0141] S801 performs time-frequency analysis on the target acoustic signal to obtain the time-frequency acoustic spectrum corresponding to the target acoustic signal.

[0142] For example, the Short-Time Fourier Transform (STFT) can be used to perform frame-by-frame windowing and time-frequency conversion on the target acoustic signal, converting the one-dimensional time-series target acoustic signal into a two-dimensional time-frequency spectrogram distributed in the time and frequency domains.

[0143] S802 extracts the time-domain and frequency-domain features of the time-frequency spectrogram.

[0144] Optionally, Mel frequency cepstral coefficients (MFCCs) can be calculated based on the time-frequency spectrogram to extract features related to human hearing perception. At the same time, parameters such as the spectral centroid, spectral flux, and spectral attenuation of the time-frequency spectrogram can be extracted and these parameters can be arranged and summarized in order to obtain the frequency domain features corresponding to the target acoustic signal.

[0145] Optionally, the temporal characteristics of the target acoustic signal can be combined to extract temporal features such as the duration of abnormal noise, frequency of sound emission, and changes in signal pulses. These temporal features can then be arranged and summarized in chronological order to obtain the temporal features corresponding to the target acoustic signal.

[0146] S803 performs weighted fusion processing on the time-domain features and frequency-domain features to obtain the target acoustic features corresponding to the target acoustic signal.

[0147] The weighting coefficients corresponding to the time-domain features and frequency-domain features can be determined based on human experience or through a large number of experiments. This application does not impose any restrictions on this.

[0148] To reduce the impact of human voices within a vehicle on the target acoustic features, in one embodiment, human voice detection can be performed on the target acoustic signal to obtain target signal segments containing human voices. During the weighted fusion processing of time-domain and frequency-domain features, the feature weight of the target signal segment is lower than the feature weight of other signal segments in the target acoustic signal. By reducing the feature weight of the signal segment containing human voices, the interference of in-vehicle conversations on the target acoustic features can be weakened, increasing the proportion of effective information in the target acoustic features, thereby improving the detection accuracy of subsequent vehicle sealing performance testing.

[0149] In the above embodiments, determining the time-domain and frequency-domain features of the target acoustic signal and then performing weighted fusion to generate target acoustic features can enrich the signal information contained in the target acoustic features, improve the comprehensiveness of the features, and thus improve the detection accuracy of subsequent vehicle sealing performance testing.

[0150] Based on the above embodiments, this application provides a detailed description of the vehicle sealing performance testing method provided in this application, such as... Figure 9 As shown, it includes the following steps:

[0151] S901, acquire the raw acoustic signal collected by the main microphone array deployed in the area to be detected of the vehicle, and the reference noise signal collected by the reference microphone array deployed in the reference area of ​​the vehicle.

[0152] S902 performs weighted fusion processing on the reference noise signal to obtain the estimated noise signal of the main microphone array.

[0153] S903 filters out the predicted noise signal from the original acoustic signal to obtain the target acoustic signal.

[0154] S904 performs time-frequency analysis on the target acoustic signal to obtain the time-frequency spectrogram corresponding to the target acoustic signal, and extracts the time-domain and frequency-domain features of the time-frequency spectrogram.

[0155] S905 performs weighted fusion processing on the time-domain features and frequency-domain features to obtain the target acoustic features corresponding to the target acoustic signal.

[0156] In some embodiments, human voice detection can also be performed on the target acoustic signal to obtain a target signal segment in the target acoustic signal containing human voice; wherein, in the process of weighted fusion processing of time domain features and frequency domain features, the feature weight of the target signal segment is lower than the feature weight of other signal segments in the target acoustic signal.

[0157] S906, determine the feature similarity between the target acoustic features and the reference acoustic features corresponding to the current operating state of the vehicle.

[0158] Feature similarity is used to characterize feature differences.

[0159] S907, when the feature similarity is less than the similarity threshold, determine the sound continuity of the target acoustic signal based on the time domain features in the target acoustic features; and determine the energy distribution of the target acoustic signal based on the frequency domain features in the target acoustic features.

[0160] S908 determines the sealing performance test result of the area to be tested based on the continuity of sound and / or energy distribution.

[0161] For example, when the sound continuity characterizes the target acoustic signal as a continuous acoustic signal and the energy distribution characterizes the target acoustic signal as a first signal, the sealing anomaly type is determined to be an air leakage type; when the sound continuity characterizes the target acoustic signal as a discontinuous acoustic signal and the energy distribution characterizes the target acoustic signal as a second signal, the sealing anomaly type is determined to be a liquid leakage type; wherein, the energy of the first signal in the preset frequency band per unit time exceeds the energy threshold; the energy of the second signal in the preset frequency band per unit time does not exceed the energy threshold, and the instantaneous peak value of the energy is greater than the peak value threshold.

[0162] S909, when the sealing performance test results indicate that the sealing of the area under test is abnormal, obtains the signal arrival time when each microphone in the main microphone array collects the original acoustic signal.

[0163] S910 constructs a geometric constraint model with the sound source location of the original acoustic signal as the independent variable and the time difference as the dependent variable, based on the time difference between the arrival times of signals corresponding to different microphones and the deployment locations of different microphones.

[0164] S911 solves the geometric constraint model to obtain the sound source location of the original acoustic signal, and determines the location of the sealing anomaly based on the sound source location.

[0165] To facilitate understanding, the following examples in several specific scenarios will be used to illustrate the above-mentioned vehicle sealing performance testing method:

[0166] In one embodiment, the vehicle is traveling at 80 km / h on a highway in clear weather. The system continuously monitors the acoustic data collected by the main microphone array and the reference microphone array. When the vehicle passes a road seam, the main microphone array on the left front side captures an abnormal high-frequency hissing sound (spectral analysis shows an abnormal increase in energy in the 3000-5000Hz range). Comparing the real-time acoustic spectrum with a normal template for 80 km / h in clear weather, the similarity is only 0.62 (the similarity threshold is 0.80), indicating an abnormal sealing performance. The feature classification model identifies the target acoustic signal and determines it to be an air leakage type. Based on the localization algorithm, the sound source is calculated to be located in the upper edge area of ​​the left front window. The system generates a diagnostic report: a slight air leakage was detected in the left front window, suggesting checking whether the sealing strip in this area is aging or has foreign objects stuck. At the same time, a schematic diagram of the location of the left front window is displayed on the central control screen.

[0167] In one embodiment, the vehicle is parked in an outdoor parking lot, and it begins to rain heavily. The right rear microphone captures regular high-frequency pulse sounds, and spectrum analysis shows typical raindrop impact sound characteristics. The external liquid sensor detects a heavy rain level. The system identifies a leak in the right rear corner of the sunroof, assessing its severity as moderate. Since the active control conditions are met, the system automatically sends a command to the sunroof control unit to increase the sunroof pressure by approximately 5%, effectively reducing the leakage. Simultaneously, the system pushes a notification to the user's mobile phone: "Leakage detected in the right rear corner of the sunroof; sunroof pressure automatically increased to alleviate leakage." It is recommended to check the sunroof sealing strip as soon as possible and replace it if necessary.

[0168] In one instance, a user reported excessive wind noise when driving at high speeds. System analysis revealed slight air leaks in both the right front door and the left rear door. The diagnostic report showed the locations of both leaks and assessed them as multiple minor leaks. The system recommended that, given the detected minor air leaks in the right front door and left rear door, which may affect high-speed driving comfort, an appointment should be scheduled to inspect the relevant door seals and assess their overall sealing performance.

[0169] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0170] Based on the same inventive concept, this application also provides a vehicle sealing performance testing device for implementing the vehicle sealing performance testing method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more vehicle sealing performance testing device embodiments provided below can be found in the limitations of the vehicle sealing performance testing method described above, and will not be repeated here.

[0171] In one exemplary embodiment, such as Figure 10 As shown, a vehicle sealing performance testing device 1 is provided, comprising: a signal acquisition module 10, a noise filtering module 20, a feature extraction module 30, and a sealing detection module 40, wherein:

[0172] The signal acquisition module 10 is used to acquire the raw acoustic signal collected by the main microphone array deployed in the detection area of ​​the vehicle, and the reference noise signal collected by the reference microphone array deployed in the reference area of ​​the vehicle.

[0173] The noise filtering module 20 is used to perform background noise filtering on the original acoustic signal based on the reference noise signal to obtain the target acoustic signal;

[0174] Feature extraction module 30 is used to extract target acoustic features corresponding to the target acoustic signal;

[0175] The sealing detection module 40 is used to determine the feature similarity between the target acoustic features and the reference acoustic features corresponding to the current operating state of the vehicle, and to determine the sealing performance detection result of the area to be detected based on the target acoustic features if the feature similarity is less than the similarity threshold.

[0176] In one exemplary embodiment, the seal detection module 40 is specifically used for:

[0177] Based on the time-domain characteristics of the target acoustic features, determine the sound continuity of the target acoustic signal; and based on the frequency-domain characteristics of the target acoustic features, determine the energy distribution of the target acoustic signal; and based on the sound continuity and / or energy distribution, determine the sealing performance test result of the area to be tested.

[0178] In one exemplary embodiment, the sealing performance test result includes the sealing anomaly type; the sealing detection module 40 is specifically used for:

[0179] When the sound continuity characterizes the target acoustic signal as a continuous acoustic signal and the energy distribution characterizes the target acoustic signal as a first signal, the sealing anomaly type is determined to be air leakage; when the sound continuity characterizes the target acoustic signal as a discontinuous acoustic signal and the energy distribution characterizes the target acoustic signal as a second signal, the sealing anomaly type is determined to be liquid leakage; wherein, the energy of the first signal in the preset frequency band per unit time exceeds the energy threshold; the energy of the second signal in the preset frequency band per unit time does not exceed the energy threshold, and the instantaneous peak value of the energy is greater than the peak value threshold.

[0180] In an exemplary embodiment, when the sealing performance test result indicates an abnormal sealing in the area to be tested, the vehicle sealing performance testing device 1 further includes an abnormal location detection module 50, used for:

[0181] Obtain the arrival time of the original acoustic signal when each microphone in the main microphone array collects the original acoustic signal; based on the time difference between the arrival times of the signals corresponding to different microphones and the deployment positions of different microphones, construct a geometric constraint model with the sound source position of the original acoustic signal as the independent variable and the time difference as the dependent variable; solve the geometric constraint model to obtain the sound source position of the original acoustic signal; determine the sealing anomaly position based on the sound source position.

[0182] In an exemplary embodiment, when the sealing performance test result indicates an abnormality in the area to be tested, and the type of sealing abnormality is leakage, the vehicle sealing performance testing device 1 further includes a window control module 60, used for:

[0183] Acquire liquid detection data from liquid sensors in the vehicle; when the vehicle's current speed is less than a speed threshold and the liquid detection data exceeds a liquid flow threshold, control the window corresponding to the location of the sealing abnormality to rise.

[0184] In one exemplary embodiment, the noise filtering module 20 is specifically used for:

[0185] The reference noise signal is weighted and fused to obtain the estimated noise signal of the main microphone array; the estimated noise signal is then filtered out from the original acoustic signal to obtain the target acoustic signal.

[0186] In one exemplary embodiment, the feature extraction module 30 is specifically used for:

[0187] Time-frequency analysis is performed on the target acoustic signal to obtain the corresponding time-frequency spectrogram; time-domain and frequency-domain features of the time-frequency spectrogram are extracted; weighted fusion processing of the time-domain and frequency-domain features is performed to obtain the target acoustic features corresponding to the target acoustic signal.

[0188] In one exemplary embodiment, the feature extraction module 30 is further configured to:

[0189] Human voice detection is performed on the target acoustic signal to obtain the target signal segment containing human voice. In the process of weighted fusion of time domain features and frequency domain features, the feature weight of the target signal segment is lower than the feature weight of other signal segments in the target acoustic signal.

[0190] Each module in the aforementioned vehicle sealing performance testing device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of the vehicle terminal as software, so that the processor can call and execute the corresponding operations of each module.

[0191] In one exemplary embodiment, a vehicle is provided, including a memory and a processor, the memory storing a computer program, which, when executed by the processor, implements the steps of the vehicle sealing performance testing method described above.

[0192] In some embodiments, the main body performing the above-described vehicle sealing performance testing method may be an on-board terminal in the vehicle, and its internal structure diagram may be as follows: Figure 11As shown, the vehicle-mounted terminal includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage medium. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, Near Field Communication (NFC), or other technologies. When the computer program is executed by the processor, it implements a method for detecting vehicle sealing performance. The display unit of the vehicle-mounted terminal is used to form a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the vehicle terminal can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the vehicle terminal, or external keyboards, touchpads, or mice, etc.

[0193] Those skilled in the art will understand that Figure 11 The structure shown is a block diagram of a portion of the structure related to the solution of this application, and does not constitute a limitation on the vehicle terminal to which the solution of this application is applied. A specific vehicle terminal may include more or fewer components than shown in the figure, or combine certain components, or have different component arrangements.

[0194] In one exemplary embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above-described method embodiments.

[0195] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0196] The user information (including but not limited to user device information, user personal information, and user voice information) and data (including but not limited to data used for analysis, stored data, and displayed data) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use, and processing of the relevant data must comply with relevant regulations.

[0197] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program mentioned can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0198] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0199] The above embodiments are merely illustrative of several implementation methods of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for testing the sealing performance of a vehicle, characterized in that, The method includes: Acquire the raw acoustic signal collected by the main microphone array deployed in the area to be detected of the vehicle, and the reference noise signal collected by the reference microphone array deployed in the reference area of ​​the vehicle. Based on the reference noise signal, the original acoustic signal is subjected to background noise filtering to obtain the target acoustic signal; Extract the target acoustic features corresponding to the target acoustic signal; Determine the feature similarity between the target acoustic features and the reference acoustic features corresponding to the current operating state of the vehicle; If the feature similarity is less than the similarity threshold, the sealing performance test result of the area to be tested is determined based on the target acoustic features.

2. The method according to claim 1, characterized in that, The step of determining the sealing performance test result of the area to be tested based on the target acoustic characteristics includes: Based on the temporal characteristics of the target acoustic features, determine the sound continuity of the target acoustic signal; and, The energy distribution of the target acoustic signal is determined based on the frequency domain characteristics of the target acoustic features. The sealing performance test result of the area to be tested is determined based on the continuity of the sound and / or the energy distribution.

3. The method according to claim 2, characterized in that, The sealing performance test results include the type of sealing anomaly; based on the sound continuity and energy distribution, the sealing performance test results of the area to be tested are determined, including: If the sound continuity indicates that the target acoustic signal is a continuous acoustic signal, and the energy distribution indicates that the target acoustic signal is a first signal, then the sealing anomaly type is determined to be an air leakage type. If the sound continuity indicates that the target acoustic signal is a discontinuous acoustic signal, and the energy distribution indicates that the target acoustic signal is a second signal, then the sealing anomaly type is determined to be a leakage type. Wherein, the energy of the first signal in the preset frequency band exceeds the energy threshold per unit time; the energy of the second signal in the preset frequency band does not exceed the energy threshold per unit time, and the instantaneous peak energy is greater than the peak threshold.

4. The method according to any one of claims 1-3, characterized in that, When the sealing performance test results indicate an abnormal sealing in the area to be tested, the method further includes: Obtain the signal arrival time when each microphone in the main microphone array acquires the original acoustic signal; Based on the time difference between the arrival times of signals from different microphones and the deployment locations of different microphones, a geometric constraint model is constructed with the sound source location of the original acoustic signal as the independent variable and the time difference as the dependent variable. The geometric constraint model is solved to obtain the sound source location of the original acoustic signal; Based on the location of the sound source, determine the location of the sealing abnormality.

5. The method according to any one of claims 1-3, characterized in that, When the sealing performance test result indicates an abnormality in the area to be tested, and the abnormality type is leakage, the method further includes: Acquire liquid detection data detected by the liquid sensor in the vehicle; If the vehicle's current speed is less than a speed threshold and the liquid detection data exceeds a liquid flow threshold, the window corresponding to the abnormal sealing location will be raised.

6. The method according to any one of claims 1-3, characterized in that, The step of performing background noise filtering on the original acoustic signal based on the reference noise signal to obtain the target acoustic signal includes: The reference noise signal is weighted and fused to obtain the estimated noise signal of the main microphone array; The predicted noise signal is filtered out from the original acoustic signal to obtain the target acoustic signal.

7. The method according to any one of claims 1-3, characterized in that, The extraction of the target acoustic features corresponding to the target acoustic signal includes: Time-frequency analysis is performed on the target acoustic signal to obtain the time-frequency acoustic spectrum corresponding to the target acoustic signal; Extract the time-domain and frequency-domain features of the time-frequency spectrogram; The time-domain features and frequency-domain features are weighted and fused to obtain the target acoustic features corresponding to the target acoustic signal.

8. The method according to claim 7, characterized in that, The method further includes: Human voice detection is performed on the target acoustic signal to obtain a target signal segment in the target acoustic signal that contains human voice; In the process of weighted fusion of the time-domain features and frequency-domain features, the feature weight of the target signal segment is lower than the feature weight of other signal segments in the target acoustic signal.

9. A vehicle sealing performance testing device, characterized in that, The device includes: The signal acquisition module is used to acquire the raw acoustic signal collected by the main microphone array deployed in the detection area of ​​the vehicle, and the reference noise signal collected by the reference microphone array deployed in the reference area of ​​the vehicle; wherein, the distance between the reference area and the noise source of the vehicle is smaller than the distance between the detection area and the noise source. The noise filtering module is used to perform background noise filtering processing on the original acoustic signal based on the reference noise signal to obtain the target acoustic signal; The feature extraction module is used to extract the target acoustic features corresponding to the target acoustic signal; The sealing detection module is used to determine the feature similarity between the target acoustic feature and the reference acoustic feature corresponding to the current operating state of the vehicle, and if the feature similarity is less than a similarity threshold, to determine the sealing performance detection result of the area to be detected based on the target acoustic feature.

10. A vehicle comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1-8.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1-8.

12. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1-8.

Images