Vehicle tire detection method, vehicle and computer readable storage medium

Abstract

The embodiment of the invention provides a tire detection method for a vehicle, the vehicle and a computer readable storage medium, and the method comprises the steps: obtaining an acoustic signal generated when a tire of the vehicle is in contact with the ground, and a rotating speed signal of the tire in a vehicle driving process; determining a target phase angle window corresponding to the acoustic signal based on the rotating speed signal; mapping the acoustic signal to an acoustic feature substrate corresponding to the target phase angle window to obtain a target feature vector corresponding to the acoustic signal; determining an initial detection result of the tire based on an initial feature vector and a target feature vector corresponding to the acoustic signal; and based on the initial detection result and the tire pressure change trend information of the tire, a detection result of the tire is determined, and the detection result is used for representing whether the tire is abnormal or not. The technical problem that the abnormal phenomenon of the vehicle tire cannot be detected in time is solved.

Description

Technical Field

[0001] This application relates to the field of vehicle technology, and more specifically, to a tire inspection method for a vehicle, a vehicle, and a computer-readable storage medium. Background Technology

[0002] The health of tires is crucial during vehicle operation, especially when a tire is punctured by a sharp object such as a nail. This can cause a rapid drop in tire pressure and even lead to a tire blowout, threatening driving safety.

[0003] In related technologies, traditional tire damage detection techniques mainly rely on direct tire pressure monitoring systems (TPMS). These systems use built-in tire pressure sensors to monitor tire pressure in real time, triggering an alarm only when the internal tire pressure falls below a safe threshold, thus alerting the driver. This method depends on a drop in tire pressure; an alarm is only triggered after a sharp object punctures the tire, causing a substantial decrease in internal pressure. It cannot provide early warnings when a foreign object has just entered the tire and before rapid air leakage begins, potentially missing the optimal intervention window to prevent further tire damage.

[0004] There is currently no good solution to the technical problem of not being able to detect abnormalities in vehicle tires in a timely manner. Summary of the Invention

[0005] This application provides a tire inspection method for a vehicle, a vehicle, and a computer-readable storage medium to at least solve the technical problem of the inability to detect abnormal phenomena in vehicle tires in a timely manner.

[0006] According to one aspect of the embodiments of this application, a tire detection method for a vehicle is provided. The method includes: acquiring acoustic signals generated by the contact between the vehicle's tires and the ground, and tire rotation speed signals, during vehicle operation; determining a target phase angle window corresponding to the acoustic signals based on the rotation speed signals, wherein the target phase angle window is used to characterize the range of tire rotation angles during the contact between the tires and the ground; mapping the acoustic signals onto an acoustic feature basis corresponding to the target phase angle window to obtain a target feature vector corresponding to the acoustic signals, wherein the acoustic feature basis is used to characterize the acoustic features of the target phase angle window when the tire is in normal operating condition, and the target feature vector is used to characterize the expected acoustic features of the acoustic signals on the acoustic feature basis; determining an initial detection result of the tire based on the initial feature vector and the target feature vector corresponding to the acoustic signals, wherein the initial feature vector is used to characterize the initial acoustic features corresponding to the acoustic signals within the target phase angle window, and the initial detection result is used to characterize the degree to which the initial acoustic features deviate from the expected acoustic features; and determining a detection result of the tire based on the initial detection result and tire pressure change trend information, wherein the detection result is used to characterize whether the tire is abnormal.

[0007] Optionally, determining the target phase angle window corresponding to the acoustic signal based on the rotational speed signal includes: aligning the acoustic signal and the rotational speed signal on the time axis based on a first timestamp corresponding to the rotational speed signal and a second timestamp corresponding to the acoustic signal, wherein the first timestamp is used to characterize the generation time of the rotational speed signal and the second timestamp is used to characterize the generation time of the acoustic signal; determining the rotational phase angle corresponding to the tire based on the aligned rotational speed signal; and determining the phase angle window corresponding to the rotational phase angle as the target phase angle window corresponding to the aligned acoustic signal.

[0008] Optionally, mapping the acoustic signal onto the acoustic feature basis corresponding to the target phase angle window to obtain the target feature vector corresponding to the acoustic signal includes: using a fast Fourier transform to convert the acoustic signal from the time domain to the frequency domain to obtain the frequency domain signal corresponding to the acoustic signal; extracting the Mel-spectrum feature vector from the frequency domain signal, wherein the Mel-spectrum feature vector is used to characterize the key frequencies in the frequency domain signal; and mapping the Mel-spectrum feature vector onto the acoustic feature basis corresponding to the target phase angle window to obtain the target feature vector corresponding to the acoustic signal.

[0009] Optionally, the initial detection result of the tire is determined based on the initial feature vector and the target feature vector corresponding to the acoustic signal, including: determining the difference vector between the initial feature vector and the target feature vector corresponding to the acoustic signal; and determining the initial detection result of the tire based on the Euclidean norm of the difference vector, wherein the Euclidean norm is used to quantify the distance between the initial feature vector and the target feature vector, and the Euclidean norm is positively correlated with the degree to which the initial acoustic features represented by the initial detection result deviate from the expected acoustic features.

[0010] Optionally, the tire detection result is determined based on the initial detection result and the tire pressure change trend information, including: periodically verifying the initial detection result to obtain a verification result, wherein the verification result is used to indicate whether there is a periodic abnormal event in the tire within the target phase angle window; in response to the verification result indicating that there is a periodic abnormal event in the tire within the target phase angle window, determining the tire pressure change trend information based on the tire pressure data within a first preset time window; determining the acoustic anomaly confidence level based on the degree to which the initial acoustic characteristics represented by the initial detection result deviate from the expected acoustic characteristics, and determining the tire pressure anomaly confidence level based on the tire pressure change trend information, wherein the acoustic anomaly confidence level is used to characterize the degree of certainty of the acoustic anomaly of the tire, and the tire pressure anomaly confidence level is used to characterize the degree of certainty of the tire pressure anomaly; and determining the tire detection result based on the acoustic anomaly confidence level and the tire pressure anomaly confidence level.

[0011] Optionally, the initial detection results are periodically verified to obtain verification results, including: acquiring the Euclidean mean values ​​corresponding to multiple phase angle windows of the tire, wherein the Euclidean mean value is used to characterize the average value of the Euclidean norms corresponding to multiple phase angle windows within a second preset time window, the multiple phase angle windows are obtained by equally dividing the tire's rotation period, and the multiple phase angle windows include a target phase angle window; comparing the first Euclidean mean value corresponding to the target phase angle window with the second Euclidean mean value corresponding to the phase angle windows other than the target phase angle window within the multiple phase angle windows to obtain a comparison result; in response to the comparison result indicating that the first Euclidean mean value is greater than the second Euclidean mean value, determining a periodic index corresponding to the target phase angle window based on the first Euclidean mean value corresponding to the target phase angle window and the rotation angle range corresponding to the target phase angle window, wherein the periodic index is used to characterize the linear correlation between the first Euclidean mean value and the rotation angle range; and determining the verification result based on the periodic index and a periodic index threshold.

[0012] Optionally, the verification result is determined based on the periodicity index and the periodicity index threshold, including: in response to the periodicity index being less than or equal to the periodicity index threshold, determining that the verification result indicates that there is a periodic abnormal event in the tire within the target phase angle window; in response to the periodicity index being greater than the periodicity index threshold, determining that the verification result indicates that there is no periodic abnormal event in the tire within the target phase angle window.

[0013] Optionally, the tire detection result is determined based on the acoustic anomaly confidence level and the tire pressure anomaly confidence level, including: fusing the acoustic anomaly confidence level and the tire pressure anomaly confidence level to obtain the tire's comprehensive confidence level, wherein the comprehensive confidence level is used to characterize the degree of certainty that the tire has an anomaly; in response to the comprehensive confidence level being greater than a confidence level threshold, the tire detection result is determined to indicate that the tire has an anomaly; the method further includes: in response to the tire having an anomaly, triggering an alarm for the tire.

[0014] According to another aspect of the embodiments of this application, a tire detection device for a vehicle is also provided. The device includes: an acquisition unit for acquiring acoustic signals generated by the tires contacting the ground and tire rotation speed signals during vehicle operation; a first determination unit for determining a target phase angle window corresponding to the acoustic signals based on the rotation speed signals, wherein the target phase angle window characterizes the range of tire rotation angles during tire-ground contact; a mapping unit for mapping the acoustic signals onto an acoustic feature base corresponding to the target phase angle window to obtain a target feature vector corresponding to the acoustic signals, wherein the acoustic feature base characterizes the acoustic features of the target phase angle window under normal tire operation, and the target feature vector characterizes the expected acoustic features of the acoustic signals on the acoustic feature base; a second determination unit for determining an initial detection result of the tire based on the initial feature vector and the target feature vector corresponding to the acoustic signals, wherein the initial feature vector characterizes the initial acoustic features of the acoustic signals within the target phase angle window, and the initial detection result characterizes the degree to which the initial acoustic features deviate from the expected acoustic features; and a third determination unit for determining a detection result of the tire based on the initial detection result and tire pressure change trend information, wherein the detection result characterizes whether the tire is abnormal.

[0015] According to another aspect of the embodiments of this application, a vehicle is also provided, including: a memory storing an executable program; and a processor for running the program, wherein the program executes the methods in various embodiments of this application when it runs.

[0016] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored executable program, wherein, when the executable program is running, it controls the device where the computer-readable storage medium is located to perform the methods of various embodiments of this application.

[0017] According to another aspect of the embodiments of this application, a computer program product is also provided, including a computer program that, when executed by a processor, implements the methods of various embodiments of this application.

[0018] According to another aspect of the embodiments of this application, a computer program product is also provided, including a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, implements the methods in various embodiments of this application.

[0019] According to another aspect of the embodiments of this application, a computer program is also provided, which, when executed by a processor, implements the methods of the various embodiments of this application.

[0020] In this embodiment, the acoustic signal generated when the tire contacts the ground is used, combined with a real-time rotation speed signal, to locate the target phase angle window corresponding to the tire. The acquired acoustic signal is then mapped onto the acoustic feature substrate corresponding to the target phase angle window to obtain the expected acoustic feature (i.e., the normal acoustic feature corresponding to the acoustic signal). By comparing this expected acoustic feature with the initial acoustic feature (i.e., the acoustic feature corresponding to the acoustic signal itself), it can be determined whether the tire is abnormal. This can be verified by combining tire pressure change trend information, accurately identifying tire abnormalities and reducing false alarm rates. In other words, this application can determine whether a tire is abnormal by analyzing the tire's acoustic signal in real time. It can issue an early warning through acoustic anomaly detection in the early stages of tire damage, before the air pressure drops significantly, providing drivers with sufficient time to address tire problems, avoid potential safety hazards, improve driving safety, and thus solve the technical problem of not being able to detect abnormal phenomena in vehicle tires in a timely manner. Attached Figure Description

[0021] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0022] Figure 1 This is a flowchart of a tire inspection method for a vehicle according to an embodiment of this application;

[0023] Figure 2 This is a schematic diagram of a hardware structure according to an embodiment of this application;

[0024] Figure 3 This is a flowchart of a tire inspection method for a vehicle according to an embodiment of this application;

[0025] Figure 4 This is a flowchart of a multimodal confidence fusion and alarm decision method according to an embodiment of this application;

[0026] Figure 5 This is a schematic diagram of a tire detection device for a vehicle according to an embodiment of this application. Detailed Implementation

[0027] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0028] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0029] According to an embodiment of this application, a method for detecting tires of a vehicle is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0030] This embodiment provides a method for tire inspection of a vehicle. Figure 1 This is a flowchart of a tire inspection method for a vehicle according to an embodiment of this application, such as... Figure 1 As shown, the process includes the following steps:

[0031] Step S101: During vehicle operation, acquire the acoustic signal generated by the contact between the vehicle's tires and the ground, as well as the tire rotation speed signal.

[0032] In the technical solution provided in step S101 of this application, the acoustic signal is used to characterize the vibration and noise characteristics generated when the tire contacts the ground. The rotational speed signal is used to characterize the real-time rotational speed and angular position of the tire. The tire can be any tire in a vehicle.

[0033] In this embodiment, during vehicle operation, acoustic signals generated by the contact between the vehicle's tires and the ground, as well as tire rotation speed signals, can be collected in real time. Regarding the acoustic signals, when the tire is operating normally, its contact with the ground produces a series of stable acoustic patterns. These patterns reflect the combined effects of multiple factors, including tire material, tread pattern, tire pressure, and road surface conditions. However, once the tire is damaged, such as by a nail or other foreign object, local structural changes can significantly alter the contact acoustic characteristics, producing acoustic anomalies that differ from the normal state. For example, after a tire is punctured by a nail, the damaged area undergoes different physical deformations each time it contacts the ground, generating unique vibration frequencies and intensities, and causing additional noise. The spectral characteristics of this noise differ from the normal operating sound of the tire. Due to the tire's rotational characteristics, acoustic anomalies typically exhibit periodic repetition synchronized with the tire's rotation cycle. Regarding the rotation speed signal, the phase angle during tire rotation can be identified. By detecting the tire's rotation speed signal, it can be determined whether the acoustic anomalies exhibit a regularity consistent with the tire's rotation cycle.

[0034] Optionally, acoustic signals can be acquired using a waterproof microphone array installed within the vehicle's wheel arches. These microphones capture the sounds generated when the tire contacts the ground, including acoustic changes caused by tire wear, road surface variations, and punctures. The waterproof microphone array is positioned 15-20 cm vertically from the tire tread and horizontally pointed towards the tire's contact patch to ensure the captured sound signals directly reflect the tire's condition. Simultaneously, the acquired acoustic signals can be pre-processed using an Infinite Impulse Response (IIR) low-pass filter to remove high-frequency wind noise and environmental noise (e.g., high-frequency components from vehicle horns).

[0035] Optionally, the speed signal is acquired in real time via a sensor connected to the vehicle's Controller Area Network (CAN) bus or Anti-lock Braking System (ABS). The speed signal not only provides information on the tire's rotational speed but also indirectly reflects the vehicle's driving status and the tire's rotational position.

[0036] In this step, acoustic signals can sensitively capture changes in vibration modes and noise characteristics when the tire contacts the ground. The rotational speed signal provides precise information on the tire's rotational angle and speed. This rotational speed signal allows for accurate marking of the acoustic signal's rotational phase angle, ensuring a precise correspondence between abnormal sounds and tire position. Therefore, by acquiring the acoustic signals generated by the tire's contact with the ground and the tire's rotational speed signal in real time during vehicle operation, strong data support can be provided for subsequent tire anomaly detection.

[0037] Step S102: Based on the rotational speed signal, determine the target phase angle window corresponding to the acoustic signal.

[0038] In the technical solution provided in step S102 of this application, a complete rotation cycle of the tire (e.g., 360 degrees) is pre-divided into multiple phase angle windows. Each phase angle window represents a specific area where the tire contacts the ground during rotation. For example, by setting the starting rotation position of the tire and marking it as 0°, and assigning a tire rotation angle of 10° to each phase angle window, the complete rotation cycle of the tire (360°) can be divided into 36 phase angle windows. The center phase angle of each phase angle window is 5°, 15°…355°. The aforementioned target phase angle window is used to characterize the range of tire rotation angles during the tire's contact with the ground. For example, the range of tire rotation angles corresponding to this target phase angle window can be 20° to 30°.

[0039] In this embodiment, after acquiring the acoustic signal and the rotational speed signal, the acoustic signal and the rotational speed signal can be aligned on the time axis according to their acquisition times to establish a correlation between them. Based on the aligned rotational speed signal, the corresponding rotational angular velocity of the tire is determined, and based on the acquisition time of the rotational speed signal, the cumulative rotational angle of the tire starting from a certain reference point is determined. Then, the cumulative rotational angle of the tire is converted into the instantaneous phase angle of the tire to ensure that the tire's phase angle is within a complete rotation cycle.

[0040] For example, the cumulative rotation angle of the tire can be converted into the instantaneous phase angle of the tire according to the following formula.

[0041]

[0042] in, This is the cumulative rotation angle of the tire. This is the instantaneous phase angle corresponding to the cumulative rotation angle of the tire.

[0043] Optionally, after determining the instantaneous phase angle of the tire, the rotational phase angle window corresponding to the acoustic signal can be determined based on the phase angle window in which the instantaneous phase angle of the tire is located.

[0044] For example, suppose the complete rotation cycle of a tire is divided into 36 phase angle windows, each corresponding to an angle range of 10°. Based on this, if the calculated instantaneous phase angle of the tire is 7.3°, then the phase angle window corresponding to this instantaneous phase angle can be determined to be from 0° to 10°. Based on this, the target phase angle window corresponding to the acoustic signal can be determined to be a phase angle window of [0°, 10°].

[0045] In this step, the target phase angle window corresponding to the acoustic signal is determined by the rotation speed signal. The acoustic signal can be associated with a specific phase angle window, providing key time and space information for subsequent acoustic feature analysis and tire damage localization.

[0046] Step S103: Map the acoustic signal onto the acoustic feature basis corresponding to the target phase angle window to obtain the target feature vector corresponding to the acoustic signal.

[0047] In the technical solution provided by step S103 of this application, the aforementioned acoustic feature basis is used to characterize the acoustic features of the target phase angle window when the tire is in normal operating condition. For example, when the tire is in a stable operating state, the acoustic data within each target phase angle window is subjected to unsupervised learning using the Slow Feature Analysis (SFA) algorithm to extract the most stable and representative slow feature basis of the tire's normal acoustic characteristics. This slow feature basis is then determined as the acoustic feature basis corresponding to each phase angle window. This acoustic feature basis contains the main frequency components and energy distribution characteristics of the sound under each phase angle window when the tire is operating normally, providing a comparison benchmark for acoustic signal analysis.

[0048] In this embodiment, after determining the target phase angle window corresponding to the acoustic signal, the acoustic signal can be mapped onto the acoustic feature basis corresponding to the target phase angle window to obtain the target feature vector corresponding to the acoustic signal. The target feature vector is used to characterize the expected acoustic features of the acoustic signal on the acoustic feature basis.

[0049] For example, before mapping an acoustic signal, it can be preprocessed to extract Mel-frequency cepstral coefficients (MFCCs) to form a high-dimensional feature vector representing the signal. Then, this high-dimensional feature vector is projected onto the acoustic feature basis corresponding to the target phase angle window to reconstruct the corresponding feature vector, obtaining the target feature vector. This target feature vector can be understood as the expression of the acoustic signal in the coordinate system of the acoustic feature basis, used to characterize the expected acoustic features, i.e., the normal acoustic features, of the signal.

[0050] In this step, by mapping the acoustic signal onto the acoustic feature basis, the feature vector corresponding to the acoustic signal can be reconstructed to obtain the expected acoustic features (e.g., normal acoustic features) corresponding to the acoustic signal, providing a basis for subsequent acoustic signal anomaly detection.

[0051] Step S104: Determine the initial detection result of the tire based on the initial feature vector and the target feature vector corresponding to the acoustic signal.

[0052] In the technical solution provided by step S104 of this application, the initial feature vector is used to characterize the initial acoustic features corresponding to the acoustic signal within the target phase angle window, the target feature vector is used to characterize the target acoustic features corresponding to the acoustic signal within the target phase angle window, and the initial detection result is used to characterize the degree to which the initial acoustic features deviate from the expected acoustic features.

[0053] In this embodiment, after obtaining the target feature vector corresponding to the acoustic signal, the initial detection result of the tire can be determined based on the initial feature vector corresponding to the acoustic signal and the target feature vector.

[0054] For example, after obtaining the initial feature vector and the target feature vector corresponding to the acoustic signal, the degree of deviation between the initial feature vector and the target feature vector can be determined based on the Euclidean distance between them, and then the initial detection result of the tire can be determined based on this degree of deviation.

[0055] In this step, after obtaining the initial feature vector and the target feature vector corresponding to the acoustic signal, the initial detection result of the tire can be determined based on the deviation between the initial feature vector and the target feature vector.

[0056] Step S105: Based on the initial detection results and the tire pressure change trend information, determine the tire detection results.

[0057] In the technical solution provided by step S105 of this application, the tire pressure change trend information is used to characterize the change of tire internal air pressure over a period of time.

[0058] In this embodiment, after obtaining the initial detection result, if the initial detection result indicates a deviation between the tire's target feature vector and the initial feature vector, the tire pressure change trend information within a preset time window can be obtained. Then, based on the tire pressure change trend information, the tire pressure change trend can be determined. If the tire pressure change trend within the preset time window indicates a continuous decrease in tire pressure, then the tire pressure is considered abnormal.

[0059] Optionally, if the initial test results indicate that the tire's acoustic characteristics are abnormal, and the tire pressure change trend information indicates that the tire pressure change trend is abnormal, then it can be determined that the tire's test results indicate that the tire is abnormal.

[0060] Optionally, after determining that there is an abnormality in the tire, a warning message can be generated and a clear warning signal can be issued to the driver of the vehicle through the in-vehicle multimedia system and the instrument panel to remind the driver that there is an abnormality in the tire.

[0061] Optionally, by inspecting each tire in the vehicle using the above method, abnormal tire conditions can be detected in a timely manner, providing effective protection for driving safety and avoiding warnings only after substantial tire failure has occurred, thus effectively grasping the optimal time for tire repair.

[0062] In steps S101 to S105 above, the acoustic signal generated when the tire contacts the ground is used, combined with the real-time rotation speed signal, to locate the target phase angle window corresponding to the tire. The acquired acoustic signal is then mapped onto the acoustic feature substrate corresponding to the target phase angle window to obtain the expected acoustic feature (i.e., the normal acoustic feature corresponding to the acoustic signal). This expected acoustic feature is compared with the initial acoustic feature (i.e., the acoustic feature corresponding to the acoustic signal itself) to determine if there is an abnormality in the tire. This is further verified by combining tire pressure change trend information, which can accurately determine the tire's abnormality and reduce the false alarm rate. In other words, this application can determine if there is an abnormality in the tire by analyzing the tire's acoustic signal in real time. It can issue an early warning through acoustic anomaly detection in the early stages of tire damage, before the air pressure drops significantly, providing drivers with sufficient time to address tire problems, avoid potential safety hazards, improve driving safety, and thus solve the technical problem of not being able to detect abnormal phenomena in vehicle tires in a timely manner.

[0063] The tire testing method for the aforementioned vehicle described in this application will be further described below.

[0064] As an optional implementation, step S102, determining the target phase angle window corresponding to the acoustic signal based on the rotational speed signal, includes: aligning the acoustic signal and the rotational speed signal on the time axis based on a first timestamp corresponding to the rotational speed signal and a second timestamp corresponding to the acoustic signal, wherein the first timestamp is used to characterize the generation time of the rotational speed signal and the second timestamp is used to characterize the generation time of the acoustic signal; determining the rotational phase angle corresponding to the tire based on the aligned rotational speed signal; and determining the phase angle window corresponding to the rotational phase angle as the target phase angle window corresponding to the aligned acoustic signal.

[0065] In this embodiment, since there is a direct correlation between the generation of acoustic signals (vibration caused by tire contact with the ground) and the actual rotation state of the tire (reflected by the rotation speed signal), time alignment is a prerequisite for analyzing the synchronization between acoustic signals and tire rotation cycle.

[0066] Optionally, a timestamp is appended to each data set when sampling the acoustic and rotational speed signals. The first timestamp of the rotational speed signal reflects the output time of the wheel speed sensor, and the second timestamp of the acoustic signal corresponds to the acquisition time of that signal frame. Based on this, the rotational speed signal and the acoustic signal can be aligned on the time axis according to the first timestamp corresponding to the rotational speed signal and the second timestamp corresponding to the acoustic signal.

[0067] Optionally, after obtaining the aligned acoustic signal and rotational speed signal, the rotational angular velocity of the tire can be calculated using the aligned rotational speed signal according to the following formula.

[0068]

[0069] in, Used to characterize tire rotation speed, Used to characterize tire radius.

[0070] Optionally, after obtaining the rotational angular velocity of the tire, the cumulative rotational angle of the tire can be further determined, and then the target phase angle window corresponding to the acoustic signal can be determined based on the cumulative rotational angle of the tire.

[0071] For example, the cumulative rotation angle of a tire can be determined using the following formula.

[0072]

[0073] in, ω represents the angular velocity of the tire, and t represents the cumulative rotation time of the tire at the moment the speed signal is acquired.

[0074] Optionally, after determining the cumulative rotational angular velocity of the tire, the phase angle corresponding to each frame of the acoustic signal can be determined by the following formula based on the acquisition time of each frame of the acoustic signal.

[0075]

[0076] in, This is the cumulative rotation angle of the tire. This is the instantaneous phase angle corresponding to the cumulative rotation angle of the tire.

[0077] Optionally, after determining the instantaneous phase angle of the tire, the rotational phase angle window corresponding to the acoustic signal can be determined based on the phase angle window in which the instantaneous phase angle of the tire is located.

[0078] For example, suppose the complete rotation cycle of a tire is divided into 36 phase angle windows, each corresponding to an angle range of 10°. Based on this, if the calculated instantaneous phase angle of the tire is 7.3°, then the phase angle window corresponding to this instantaneous phase angle can be determined to be from 0° to 10°. Based on this, the target phase angle window corresponding to the acoustic signal can be determined to be a phase angle window of [0°, 10°].

[0079] In this step, by aligning the rotational speed signal with the acoustic signal, the target phase angle window corresponding to the acoustic signal can be determined from the aligned rotational speed signal. The acoustic signal can be associated with a specific phase angle window, providing crucial temporal and spatial information for subsequent acoustic feature analysis and tire damage localization.

[0080] As an optional implementation, step S103, mapping the acoustic signal onto the acoustic feature basis corresponding to the target phase angle window to obtain the target feature vector corresponding to the acoustic signal, includes: using a fast Fourier transform to convert the acoustic signal from the time domain to the frequency domain to obtain the frequency domain signal corresponding to the acoustic signal; extracting the Mel spectrum feature vector from the frequency domain signal, wherein the Mel spectrum feature vector is used to characterize the key frequencies in the frequency domain signal; and mapping the Mel spectrum feature vector onto the acoustic feature basis corresponding to the target phase angle window to obtain the target feature vector corresponding to the acoustic signal.

[0081] In this embodiment, the acquired acoustic signal exists in the form of a time series. Based on this, when mapping the acoustic signal onto the acoustic feature basis corresponding to the target phase angle window, the acoustic signal can be converted from the time domain to the frequency domain using Fast Fourier Transform (FFT) to obtain the frequency domain signal corresponding to the acoustic signal.

[0082] For example, Hamming windowing can be applied to each frame of the acoustic signal to reduce spectral leakage and Gibbs free energy. Then, a Fast Fourier Transform (FFT) is performed on the processed acoustic signal to obtain the corresponding frequency domain signal. This frequency domain signal contains amplitude information for different frequency components.

[0083] Optionally, after obtaining the frequency domain signal corresponding to the acoustic signal, a multi-channel Mel filter bank (e.g., 26 channels, with a frequency range set to 20Hz-8kHz) can be used to extract Mel frequency spectral coefficients (MFCCs) from the frequency domain signal of the acoustic signal. The Mel frequency spectral coefficients are used to characterize the key frequencies in the frequency domain signal.

[0084] Optionally, after extracting the Mel spectrum feature vector from the frequency domain signal of the acoustic signal, the Mel spectrum feature vector can be mapped onto the acoustic feature basis corresponding to the target phase angle window to obtain the target feature vector corresponding to the acoustic signal.

[0085] For example, the target feature vector can be obtained by projecting the Mel spectrum feature vector onto the acoustic feature basis corresponding to the target phase angle window according to the projection coefficient using the following formula.

[0086]

[0087] in, Used to characterize the target feature vector Used to represent the acoustic feature basis corresponding to the target phase angle window Used to represent the projection coefficient, and the projection coefficient ,in, Used to represent the eigenvectors of the Mel spectrum.

[0088] Optionally, after obtaining the target feature vector, the target feature vector can be regarded as the reconstructed feature vector corresponding to the acoustic signal. The target feature vector represents the expected acoustic feature of the tire within the target phase angle window, that is, the acoustic feature of the tire in the target phase angle window under normal conditions.

[0089] As an optional implementation, step S104, based on the initial feature vector and the target feature vector corresponding to the acoustic signal, determines the initial detection result of the tire, including: determining the difference vector between the initial feature vector and the target feature vector corresponding to the acoustic signal; and determining the initial detection result of the tire based on the Euclidean norm of the difference vector, wherein the Euclidean norm is used to quantify the distance between the initial feature vector and the target feature vector, and the Euclidean norm is positively correlated with the degree to which the initial acoustic features represented by the initial detection result deviate from the expected acoustic features.

[0090] In this embodiment, the initial feature vector corresponding to the acoustic signal is used to indicate the acoustic characteristics of the tire within the target phase angle window under the current state. This initial feature vector can be a feature vector obtained after preprocessing the acoustic signal (e.g., noise reduction, framing, feature extraction, etc.), typically a Mel-spectrum feature vector or other acoustic feature indices. The target feature vector corresponding to the acoustic signal is used to represent the acoustic characteristics of the tire within the target phase angle window under normal conditions. After obtaining the initial and target feature vectors corresponding to the acoustic signal, the difference vector between the initial and target feature vectors can be calculated.

[0091] For example, the difference vector between the initial feature vector and the target feature vector of the acoustic signal can be determined by the following formula.

[0092]

[0093] in, Used to represent the difference vector. Used to represent the initial eigenvector (Mel spectrum eigenvector). Used to represent the target feature vector.

[0094] Optionally, after obtaining the difference vector between the initial feature vector and the target feature vector of the acoustic signal, the Euclidean norm (e.g., L2 norm) corresponding to the difference vector can be determined. This Euclidean norm is used to quantize the distance between the initial feature vector and the target feature vector of the acoustic signal.

[0095] For example, the Euclidean norm corresponding to the difference vector can be determined by the following formula.

[0096]

[0097] in, The Euclidean norm is used to represent the eigenvalues, and N is used to represent the dimensions of the initial eigenvectors and the target eigenvectors.

[0098] Optionally, after obtaining the Euclidean norm corresponding to the difference vector, the initial detection result of the tire can be determined based on the Euclidean norm. The larger the Euclidean norm, the greater the deviation between the initial feature vector and the target feature vector of the acoustic signal, that is, the more the features of the acoustic signal deviate from the normal acoustic feature basis.

[0099] Optionally, after obtaining the Euclidean norm, it can be compared with a preset threshold to determine whether the current acoustic characteristics of the tire are normal. If the Euclidean norm exceeds the preset threshold, it indicates that the tire has an abnormality such as a nail puncture. The detection result can be marked as abnormal and incorporated into the subsequent periodic abnormal event detection and multimodal fusion decision-making process as part of the tire condition assessment.

[0100] As an optional implementation, step S105, based on the initial detection result and the tire pressure change trend information, determines the tire detection result, including: periodically verifying the initial detection result to obtain a verification result, wherein the verification result is used to indicate whether there is a periodic abnormal event in the tire within the target phase angle window; in response to the verification result indicating that there is a periodic abnormal event in the tire within the target phase angle window, determining the tire pressure change trend information based on the tire pressure data within a first preset time window; determining the acoustic anomaly confidence level based on the degree to which the initial acoustic characteristics represented by the initial detection result deviate from the expected acoustic characteristics, and determining the tire pressure anomaly confidence level based on the tire pressure change trend information, wherein the acoustic anomaly confidence level is used to characterize the degree of certainty of the tire's acoustic anomaly, and the tire pressure anomaly confidence level is used to characterize the degree of certainty of the tire's tire pressure anomaly; and determining the tire detection result based on the acoustic anomaly confidence level and the tire pressure anomaly confidence level.

[0101] In this embodiment, after obtaining the initial detection results, in order to avoid false alarms caused by noise or other transient interference, the initial detection results can be periodically verified after obtaining the initial detection results to determine whether the phenomenon of the initial acoustic characteristics of the acoustic signals deviating from the expected acoustic characteristics has a regularity corresponding to the tire rotation cycle.

[0102] Optionally, when periodically verifying the initial detection results, the Euclidean norm between the initial eigenvector and the target eigenvector of the acoustic signal within the target phase angle window can be continuously detected, and this Euclidean norm can be identified as an anomaly. This score reflects the degree of deviation of the acoustic signal from the normal baseline. If, over multiple rotation cycles, the anomaly score of the target phase angle window is consistently significantly higher than the average of other windows, and this increase is consistent with the tire rotation cycle, then a periodic anomaly event is determined to exist within the target phase angle window. That is, the verification result indicates the presence of a periodic anomaly event within the target phase angle window; otherwise, a periodic anomaly event is determined to exist within the target phase angle window.

[0103] Optionally, if the verification result indicates the presence of a periodic abnormal event within the target phase angle window, the tire pressure data of the tires associated with the periodic abnormal event can be further analyzed to assess the tire pressure change trend.

[0104] For example, tire pressure data can be acquired within a first preset time window, and then the tire pressure change trend information can be determined based on the tire pressure data. The tire pressure change trend information is used to characterize the tire pressure change trend within the first preset time window. The selection of the first preset time window takes into account the balance between the rate of tire pressure drop after tire leakage and the response time of the detection system.

[0105] Optionally, after obtaining the tire pressure change trend information, the acoustic anomaly confidence level can be determined based on the degree to which the initial acoustic characteristics characterized by the initial detection results deviate from the expected acoustic characteristics, and the tire pressure anomaly confidence level can be determined based on the tire pressure change trend information.

[0106] Optionally, when determining the confidence level of acoustic anomalies, the Euclidean norm corresponding to the target phase angle window can be normalized first to obtain the normalized Euclidean norm. Then, based on the historical distribution of the Euclidean norm corresponding to the target phase angle window, the probability that the normalized Euclidean norm exceeds N times the distribution mean is calculated, where N can be set according to requirements. After obtaining the probability, the acoustic confidence level can be determined based on this probability. When this probability is greater than the target probability threshold, the degree of certainty regarding the acoustic anomaly of the tire is considered to be...

[0107] Optionally, when determining the confidence level of tire pressure anomalies, real-time tire pressure data can be continuously collected, and linear regression can be performed on the tire pressure data within a sliding time window to obtain the slope k of the tire pressure decrease trend. The confidence level of tire pressure anomalies can then be determined based on the slope of the tire pressure decrease trend and a preset confidence assessment standard for the tire pressure decrease trend. For example, taking a confidence level of 0.85 when k = -0.03 bar / h as an example, this indicates that if the absolute value of the tire pressure decrease rate reaches or exceeds 0.03 bar / h, the tire pressure change trend is considered to be highly correlated with its potential tire damage (e.g., tire puncture), and the confidence level of the tire pressure trend is high.

[0108] Optionally, after obtaining the acoustic anomaly confidence level and the tire pressure anomaly confidence level, the acoustic anomaly confidence level and the tire pressure anomaly confidence level can be fused and analyzed to determine the tire detection result. This fusion decision-making method can significantly reduce the false alarm rate because even if a false alarm occurs in a single detection mode, it is difficult for the confidence levels of both modes to exceed the threshold at the same time.

[0109] As an optional implementation, the initial detection results are periodically verified to obtain verification results, including: acquiring the Euclidean mean values ​​corresponding to multiple phase angle windows of the tire, wherein the Euclidean mean value is used to characterize the average value of the Euclidean norms corresponding to multiple phase angle windows within a second preset time window, the multiple phase angle windows are obtained by equally dividing the tire's rotation period, and the multiple phase angle windows include a target phase angle window; comparing the first Euclidean mean value corresponding to the target phase angle window with the second Euclidean mean values ​​corresponding to the phase angle windows other than the target phase angle window within the multiple phase angle windows to obtain a comparison result; in response to the comparison result indicating that the first Euclidean mean value is greater than the second Euclidean mean value, determining a periodic index corresponding to the target phase angle window based on the first Euclidean mean value corresponding to the target phase angle window and the rotation angle range corresponding to the target phase angle window, wherein the periodic index is used to characterize the linear correlation between the first Euclidean mean value and the rotation angle range; and determining the verification result based on the periodic index and a periodic index threshold.

[0110] In this embodiment, when periodically verifying the initial detection results, the average value of the Euclidean norm (i.e., anomaly score) corresponding to the tire in multiple phase angle windows can be determined first to evaluate the overall anomaly level of the tire acoustic characteristics at different rotation positions.

[0111] For example, the method for determining the Euclidean norm described above can be used to determine the Euclidean norm (i.e., the outlier score) corresponding to each phase angle window of the tire, and within the second preset time window, the average value of the Euclidean norm in each phase angle window can be calculated to obtain the Euclidean mean value corresponding to each of the multiple phase angle windows of the tire.

[0112] Optionally, after obtaining the Euclidean means corresponding to multiple phase angle windows of the tire, the first Euclidean mean corresponding to the target phase angle window can be compared with the second Euclidean means corresponding to the phase angle windows other than the target phase angle window. Based on the comparison result, it can be determined whether there is a significant abnormal signal in the target window. For example, if the comparison result indicates that the first Euclidean mean of the target phase angle window is greater than the second Euclidean mean, and this difference is statistically significant (e.g., the p-value obtained through a T-test or U-test is less than 0.05), then a significant abnormal signal is determined to exist within the target phase angle window.

[0113] For example, assuming the complete rotation cycle of a tire is divided into 36 phase angle windows, the first Euclidean mean corresponding to the target phase angle window can be compared with the second Euclidean mean corresponding to the Euclidean mean of the remaining 35 phase angle windows to obtain the comparison result.

[0114] Optionally, if the comparison results indicate that the first Euclidean mean corresponding to the target phase angle window is significantly higher than the second Euclidean mean corresponding to other phase angle windows, it indicates that the tire has a significant abnormal signal within the target phase angle window. In this case, the synchronicity between the first Euclidean mean corresponding to the target phase angle window and the tire rotation cycle can be further verified. For example, the Pearson correlation coefficient can be calculated based on the first Euclidean mean corresponding to the target phase angle window and the rotation angle range corresponding to the target phase angle window (e.g., a specific angle segment in a specific tire rotation cycle), and this Pearson correlation coefficient can be determined as the periodicity index corresponding to the target phase angle window.

[0115] Optionally, by calculating the Pearson correlation coefficient, it is possible to distinguish between acoustic anomalies caused by tire damage and random environmental noise or short-term interference events. This is because random environmental noise or short-term interference events do not exhibit a periodic variation pattern with tire rotation. Therefore, by defining the Pearson correlation coefficient as the periodicity index corresponding to the target phase angle window, it is possible to accurately determine whether periodic abnormal events exist within the tire's target phase angle window.

[0116] Optionally, after obtaining the periodicity index corresponding to the target phase angle window, the verification result can be determined based on this periodicity index and its threshold. The process of determining the verification result based on the periodicity index and its threshold will be further described below.

[0117] As an optional implementation, the verification result is determined based on the periodicity index and the periodicity index threshold, including: in response to the periodicity index being less than or equal to the periodicity index threshold, determining that the verification result indicates that there is a periodic abnormal event in the tire within the target phase angle window; and in response to the periodicity index being greater than the periodicity index threshold, determining that the verification result indicates that there is no periodic abnormal event in the tire within the target phase angle window.

[0118] In this embodiment, if the periodicity index corresponding to the target phase angle window is greater than the periodicity index threshold, it indicates that the verification result indicates that the tire has a periodic abnormal event within the target phase angle window. Conversely, if the periodicity index corresponding to the target phase angle window is less than or equal to the periodicity index threshold, it indicates that the verification result indicates that the tire does not have a periodic abnormal event within the target phase angle window.

[0119] For example, assuming the periodic anomaly index is 0.8 and the periodicity index is r, if |r|>0.8, it means that the verification result indicates that there is a periodic anomaly event in the tire within the target phase angle window; otherwise, it means that the verification result indicates that there is no periodic anomaly event in the tire within the target phase angle window.

[0120] In this step, by comparing the periodic index within the target phase angle window with the periodic index threshold, it can be determined whether there are periodic abnormal events within the tire's target phase angle window, thereby improving the accuracy of tire abnormality detection.

[0121] As an optional implementation, the tire detection result is determined based on the acoustic anomaly confidence level and the tire pressure anomaly confidence level, including: fusing the acoustic anomaly confidence level and the tire pressure anomaly confidence level to obtain the tire's comprehensive confidence level, wherein the comprehensive confidence level is used to characterize the degree of certainty that the tire has an anomaly; and in response to the comprehensive confidence level being greater than a confidence level threshold, determining that the tire detection result indicates that the tire has an anomaly.

[0122] In this embodiment, after obtaining the acoustic anomaly confidence level and the tire pressure anomaly confidence level of the tire, in order to comprehensively evaluate the overall condition of the tire, the acoustic anomaly confidence level and the tire pressure anomaly confidence level can be fused to determine the overall confidence level of the tire.

[0123] Optionally, the overall confidence level of the tire can be determined by fusing acoustic anomaly confidence levels and tire pressure anomaly confidence levels using the Dempster-Shafer (DS) synthesis rule. The DS synthesis rule allows information from different sources to be synthesized in a probabilistic manner, thus yielding more reliable conclusions.

[0124] For example, the overall confidence level of a tire can be obtained by fusing the acoustic anomaly confidence level and the tire pressure anomaly confidence level using the following formula.

[0125]

[0126] in, Used to represent the overall confidence level Used to express the confidence level of acoustic anomalies. Used to indicate the confidence level of abnormal tire pressure.

[0127] Optionally, after obtaining the overall confidence level of the tire, the overall confidence level can be compared with a confidence threshold to obtain a comparison result. If the comparison result indicates that the overall confidence level of the tire is greater than the confidence threshold, then the tire detection result indicates that the tire is abnormal; if the comparison result indicates that the overall confidence level of the tire is less than or equal to the confidence threshold, then the tire detection result indicates that the tire is abnormal. This ensures that an alarm is triggered only when both types of anomaly confidence levels reach a certain level, which greatly reduces the probability of false alarms and improves the reliability of the system.

[0128] In this step, by fusing the acoustic anomaly confidence level with the tire pressure trend confidence level, a more comprehensive and accurate tire condition assessment can be made. This multimodal fusion decision-making method overcomes the limitations of a single sensor in complex environments and improves the detection accuracy of tire anomalies.

[0129] As an optional implementation, the method further includes: triggering an alarm for the tire in response to an abnormality in the tire.

[0130] In this embodiment, once a tire abnormality is detected, a tire alarm can be triggered. For example, a voice warning can be issued through the in-vehicle multimedia system, and a red light on the dashboard can be illuminated to clearly indicate the specific tire affected (e.g., the left front tire, the right rear tire, etc.), reminding the driver to take immediate action, such as slowing down, pulling over, and checking the tire condition, so as to repair the damaged tire in a timely manner, avoid greater damage, and ensure driving safety.

[0131] The above technical solutions of the embodiments of this application will be further illustrated below with reference to preferred embodiments.

[0132] In related technologies, tire puncture detection typically involves directly measuring the tire's inflation pressure using pressure sensors installed inside each tire. When the tire pressure falls below a certain threshold, an alarm is triggered, indicating a potential leak or damage. This detection method only activates the alarm after the tire inflation pressure has dropped to a certain level (i.e., the impact of a substantial leak or damage is sufficient for the sensor to detect). This means that initial damage, such as a nail puncture, often doesn't receive an immediate response because the tire pressure change may not yet be within the warning range. Due to this delayed warning, drivers often miss the initial stages of damage, potentially missing the ideal repair window when the damage is minor and easily repairable. Over time, even a small puncture can lead to rapid tire deterioration and even a sudden tire blowout while driving, jeopardizing driving safety.

[0133] Related technologies also utilize portable tire inspection devices to perform a comprehensive scan of tires when the vehicle is stationary, primarily for detecting external damage such as punctures and cracks. The biggest limitation of these portable devices lies in their "parking inspection" mode, meaning they can only perform inspections when the vehicle is stationary and cannot provide real-time monitoring. This means that any sudden tire damage while the vehicle is in motion cannot be detected in time. If a tire is damaged while driving and the driver fails to notice it immediately, it can lead to serious consequences, such as complete tire failure or a traffic accident.

[0134] However, this application provides a tire detection method for vehicles. By acquiring the acoustic signal generated when the tire contacts the ground and combining it with a real-time rotational speed signal, a target phase angle window corresponding to the tire is located. The acquired acoustic signal is then mapped onto an acoustic feature substrate corresponding to the target phase angle window to obtain the expected acoustic feature (i.e., the normal acoustic feature corresponding to the acoustic signal). By comparing this expected acoustic feature with the initial acoustic feature (i.e., the acoustic feature corresponding to the acoustic signal itself), it is possible to determine whether the tire is abnormal. This can be verified by combining tire pressure change trend information, which can accurately determine the tire's abnormality and reduce the false alarm rate. In other words, this application can determine whether the tire is abnormal by analyzing the tire's acoustic signal in real time. It can issue an early warning through acoustic anomaly detection in the early stages of tire damage, before the air pressure drops significantly, providing drivers with sufficient time to address tire problems, avoid potential safety hazards, improve driving safety, and thus solve the technical problem of not being able to detect abnormal phenomena in vehicle tires in a timely manner.

[0135] Figure 2 This is a schematic diagram of a hardware structure according to an embodiment of this application. For example... Figure 2As shown, the core hardware deployment and data interaction process of the tire puncture detection system based on unsupervised acoustic feature learning are demonstrated, clearly depicting the system composition and the data exchange paths between various components.

[0136] like Figure 2 As shown, the hardware architecture includes an acoustic sensing module 201, a signal conditioning circuit 202, a speed signal module 203, a tire pressure monitoring module 204, a central processing unit 205, an alarm output module 206, and a model update module 207. The central processing unit 205 includes a digital signal processor (DSP) 2051, a microcontroller unit (MCU) 2052, and a neural processing unit (NPU) 2053.

[0137] The acoustic sensing module 201 consists of four omnidirectional waterproof microphones, each located inside one of the four wheel arches (left front, right front, left rear, and right rear) of the vehicle, to capture acoustic signals generated by the tires' contact with the ground. The microphones are designed with an IP67 protection rating to ensure stable operation even in harsh environments.

[0138] The signal conditioning circuit 202 is used to amplify and filter the acoustic signal collected by the acoustic sensor in order to remove noise signals.

[0139] The speed signal module 203 is used to access the vehicle's CAN bus and mainly utilizes the wheel speed data from the ABS sensor to accurately track the rotational state of each tire.

[0140] The tire pressure monitoring module 204 communicates with the vehicle's TPMS system via a Local Interconnect Network (LIN) bus to acquire real-time pressure data for all four tires. Tire pressure information forms the basis for trend verification in this detection system, ultimately confirming whether abnormal acoustic signals are accompanied by a decrease in tire pressure.

[0141] The central processing unit 205, serving as the control unit for the entire process, includes a DSP 2051, an MCU 2052, and an NPU 2053. The DSP 2051 handles real-time acoustic signal processing, including complex calculations such as noise reduction and feature extraction. The MCU 2052 undertakes logical decision-making tasks, such as feature learning, anomaly detection, and alarm triggering. Furthermore, the NPU 2053 improves the efficiency of complex feature extraction, particularly the computational speed of unsupervised learning algorithms.

[0142] The alarm output module 206 is used to send voice prompts to the driver through the vehicle multimedia system and illuminate the red warning light on the instrument panel to indicate the location of the specific tires, so as to remind the driver to take timely action.

[0143] The model update module 207 is used to periodically or according to specific conditions to retrain the acoustic feature model to adapt to changes in vehicle and tire conditions.

[0144] Figure 3 This is a flowchart of a vehicle tire inspection method according to an embodiment of this application, such as... Figure 3 As shown, the method includes the following steps.

[0145] Step S301: Synchronize signal acquisition.

[0146] In this embodiment, a waterproof microphone array installed in the vehicle's wheel arches is used to collect sound signals generated by the tires contacting the road surface; real-time wheel speed signals are received via the vehicle's CAN bus or ABS sensors; and tire pressure data for all four tires are collected in real time for cross-modal information fusion in subsequent analysis.

[0147] For example, for audio signals, the microphone can collect audio signals at a sampling frequency of 44.1kHz and a quantization bit depth of 16 bits, extracting 2048 sampling points per frame (frame length ≈ 46ms), with a frame shift of 1024 points (frame shift ≈ 23ms) to ensure temporal continuity.

[0148] Step S302, signal preprocessing.

[0149] In this embodiment, for the sound signal, high-frequency wind noise and environmental noise (such as the high-frequency components of vehicle horns) are filtered out using an IIR low-pass filter during acquisition. The cutoff frequency of this IIR low-pass filter is 8kHz. For the wheel speed signal, the tire's rotational angular velocity and cumulative rotation angle can be calculated from the wheel speed signal. The timestamp of each frame of the sound signal is associated with the cumulative rotation angle of the tire. Based on this, the rotational phase angle corresponding to each frame of the sound signal can be calculated, and the phase angle window corresponding to each frame of the acoustic signal can be determined by linear interpolation. For the tire pressure signal, the tire pressure signal and the acoustic signal can be aligned to form a timestamp-synchronized multimodal dataset.

[0150] Step S303, Feature extraction.

[0151] In this embodiment, a Fast Fourier Transform (FFT) is performed on the preprocessed acoustic frame to convert the time-domain signal into a frequency-domain signal, obtaining the signal's spectrum. The spectrum obtained from the FFT is then passed through a set of Mel filter banks. These filters are designed according to the perceptual characteristics of the human auditory system at different frequency bands, enabling them to capture frequency components that are more sensitive to tire anomaly detection. The energy output from each Mel filter is logarithmically compressed to reflect the human auditory system's greater sensitivity to low frequencies. Subsequently, the logarithmic energy is subjected to a Discrete Cosine Transform (DCT) to obtain a set of Mel spectral cepstral coefficients. Typically, the first 13 coefficients are selected as the Mel spectral feature vector, containing the main characteristic information of the tire sound signal.

[0152] Optionally, in the final stage of feature extraction, the Slow Feature Analysis (SFA) algorithm can be used to further extract the features that change most slowly over time. These features are more likely to reflect the health status of the tire. For example, a time delay operation can be performed on the MFCC feature vector to generate a series of delayed features, which helps to capture acoustic change patterns within the tire's rotation cycle. The feature dimension can be increased through polynomial expansion to enhance the model's ability to capture nonlinear relationships. In the expanded feature space, the SFA algorithm can be applied to extract the most stable features, i.e., those that change slowly over time.

[0153] Step S304, anomaly detection.

[0154] In this embodiment, after obtaining the Mel spectrum feature vector, it can be projected onto the slow feature basis of the corresponding rotating phase angle window to calculate the reconstruction error of the current acoustic state.

[0155] Optionally, this reconstruction error reflects the degree of difference between the current signal and the normal model, and can be regarded as an anomaly score. To make the anomaly scores more comparable and meaningful, they are usually normalized to calculate a normalized anomaly score.

[0156] Step S305, periodic verification.

[0157] In this embodiment, after anomaly detection, the normalized anomaly score for each rotation phase angle window can be continuously recorded. This monitoring process is continuous, covering the entire vehicle driving process, ensuring comprehensive tracking of anomaly events. A sufficiently long time period can be selected for statistical analysis, such as 30 minutes. During this period, a large amount of anomaly score information for each phase angle window will be collected, providing a data foundation for subsequent statistical verification.

[0158] Optionally, for each phase angle window, hypothesis testing methods, such as the T-test or the Mann-Whitney U test, are used to compare the mean outlier scores of that specific window with those of other windows. The purpose of hypothesis testing is to determine whether the outlier scores within a particular phase angle window are significantly higher than the average level of other windows, thereby determining whether the outliers are concentrated at a specific tire rotation position.

[0159] Optionally, a significance level α (typically 0.05) is set as the criterion for judging the significance of the results. For phase angle windows suspected of having anomalies, the mean of their anomaly scores is calculated, along with the sum of the mean anomaly scores of other phase angle windows. An independent samples t-test is performed to compare the mean anomaly score of the suspected anomaly phase angle window with the sum of the mean anomaly scores of other phase angle windows, obtaining a significance score p-value. If the obtained p-value is less than α, it can be inferred that the anomaly score of this phase angle window is significantly higher than that of other windows, and there may be periodic anomalies.

[0160] Optionally, if the abnormal score of a suspected abnormal phase angle window is confirmed by statistical analysis to be significantly higher than that of other phase angle windows, it is further verified whether these abnormal scores are highly correlated with the tire rotation cycle.

[0161] For example, the anomaly score of a suspected anomaly phase angle window can be calculated, along with the Pearson correlation coefficient r between the tire rotation angle θ. If |r| is higher than a preset threshold (e.g., 0.8), it indicates a strong positive correlation between the anomaly score and the tire rotation cycle. This confirms that the tire corresponding to the suspected anomaly phase angle window has a significant periodic anomaly, and the tire is marked as a suspect tire.

[0162] Step S306: Calculate the linear regression slope based on the tire pressure change trend.

[0163] In this embodiment, after determining that the suspected tire exhibits a significant periodic anomaly, the tire pressure changes of the suspected tire can be further monitored. A sliding time window is selected after the warning. Calculate the slope of linear regression based on tire pressure data in seconds. (formula: , This represents the change in tire pressure. =60 seconds). Set preset threshold. (That is, a decrease of approximately 0.0033 bar per minute), if k < -0.02 bar / h and continues If the tire pressure is measured within seconds, it is determined that the tire pressure is showing a significant downward trend.

[0164] Step S307, fusion decision.

[0165] In this embodiment, after obtaining the acoustic signal detection results and tire pressure detection results of the suspected tire, the probability that the abnormal score exceeds the normal range can be calculated based on the statistical characteristics of the abnormal score of the suspected tire, thus obtaining the acoustic anomaly confidence level of the suspected tire. Simultaneously, the reliability of the tire pressure decrease trend of the suspected tire can be judged based on the linear regression slope of the suspected tire, thus obtaining the tire pressure anomaly confidence level of the suspected tire.

[0166] Optionally, after obtaining the acoustic anomaly confidence level and the tire pressure anomaly confidence level of the suspected tire, a decision fusion algorithm (such as Dempster-Shafer evidence theory or Bayesian fusion) can be used to comprehensively calculate the acoustic anomaly confidence level and the tire pressure anomaly confidence level to obtain the final fused decision confidence level.

[0167] Step S308, continue monitoring.

[0168] In this embodiment, if the linear regression slope of the suspect tire is stable, the suspect tire is monitored.

[0169] Step S309: Trigger an alarm.

[0170] In this embodiment, an alarm is triggered if the linear regression slope of the suspected tire is less than 0.02 bar / h. For example, a specific alarm message is generated, such as "Suspected nail puncture in the left front tire, please check," and this alarm message is conveyed to the driver through the instrument panel and multimedia system.

[0171] In steps S301 to S309 above, by integrating unsupervised acoustic feature learning, rotational phase synchronization analysis and tire pressure trend verification, a high-precision, high-reliability and low-false-alarm early warning for tire puncture events is achieved. It can not only construct an acoustic feature model of the tire under normal conditions and accurately locate anomalies, but also effectively avoid false alarms through comprehensive analysis of multimodal information, ensuring the accuracy and timeliness of the warning, thereby significantly improving driving safety and reducing maintenance costs.

[0172] Figure 4 This is a flowchart of a multimodal confidence fusion and alarm decision-making method according to an embodiment of this application, such as... Figure 4 As shown, the method includes the following steps.

[0173] Step S401: Receive the acoustic anomaly confidence level and the tire pressure anomaly confidence level.

[0174] In this embodiment, after the system detects an abnormal sound signal within a specific phase angle window during the tire rotation cycle, it calculates the probability of the deviation between the current abnormal event and the normal state based on statistical analysis of historical data. For example, if the abnormal score of the current phase angle window consistently exceeds the historical mean by 3 standard deviations, the acoustic anomaly confidence level may be set to 0.9, indicating that the confidence level of the acoustic anomaly event is very high.

[0175] Optionally, based on time series analysis of tire pressure data, if the calculated linear regression slope k (tire pressure decrease rate) is consistently negative and its absolute value exceeds a preset threshold T, the following conditions may be considered. If the pressure drops to -0.02 bar / h, it is considered that the tire pressure is showing an abnormal downward trend. The confidence level of the abnormal tire pressure is based on the absolute value of k and the duration, which is used to quantify the confidence level of the downward trend in tire pressure.

[0176] Step S402, DS evidence synthesis.

[0177] In this embodiment, DS evidence is a mathematical framework for decision fusion involving uncertainty and multiple evidence sources, capable of effectively handling confidence information from different sensors. Acoustic anomaly confidence and tire pressure trend confidence are considered as evidence from different sources. DS evidence synthesis calculates the synthesis probability of these pieces of evidence, fusing uncertainty information from different modalities into a more comprehensive decision basis.

[0178] Step S403: Calculate the overall confidence level.

[0179] In this embodiment, the overall confidence level calculated using the DS synthesis rule, based on the acoustic anomaly confidence level and the tire pressure anomaly confidence level, reflects the system's overall judgment of the current tire damage state. The specific calculation formula is shown below.

[0180]

[0181] in, Used to represent the overall confidence level Used to express the confidence level of acoustic anomalies. Used to indicate the confidence level of abnormal tire pressure.

[0182] Step S404: Determine whether the overall confidence level is greater than the preset confidence threshold.

[0183] In this embodiment, a comprehensive confidence threshold is set to determine whether an alarm is triggered. This threshold is determined based on the system's security requirements and false alarm rate control objectives, and is typically set to a high value to ensure alarm accuracy (e.g., 0.8).

[0184] Optionally, the overall confidence level is compared with a preset confidence threshold. If the overall confidence level is greater than the preset confidence threshold, the tire damage is confirmed as a high-confidence event, triggering a final alarm and executing step S405. If the overall confidence level is less than or equal to the preset confidence threshold, the tire damage is considered a low-confidence event, and in this case, step S406 is executed.

[0185] Step S405: Trigger the final alarm.

[0186] In this embodiment, the alarm message should clearly indicate the specific location of the damaged tire to help the driver quickly pinpoint the problem. The alarm can be implemented in various ways, including visual warnings on the dashboard and voice prompts from the multimedia system, ensuring the driver notices the issue promptly and takes appropriate action.

[0187] Step S406: Return to monitoring status.

[0188] In this embodiment, acoustic and tire pressure data continue to be collected and analyzed.

[0189] In steps S401 to S406 above, by synthesizing DS evidence, the confidence information of acoustic anomalies and tire pressure drop trends can be effectively fused to calculate a comprehensive confidence level, which is used to determine whether an alarm is triggered. This component improves the accuracy of the alarm and also ensures the reliability and practicality of the system.

[0190] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0191] According to an embodiment of this application, a tire detection device for a vehicle is provided. It should be noted that the tire detection device for the vehicle can be used to perform the tire detection method for the vehicle described above.

[0192] Figure 5 This is a schematic diagram of a tire detection device for a vehicle according to an embodiment of this application. Figure 5 As shown, the tire detection device 500 for the vehicle may include: an acquisition unit 501, a first determination unit 502, a mapping unit 503, a second determination unit 504, and a third determination unit 505.

[0193] The acquisition unit 501 is used to acquire the acoustic signal generated by the contact between the vehicle's tires and the ground, as well as the tire rotation speed signal, during the vehicle's operation.

[0194] The first determining unit 502 is used to determine the target phase angle window corresponding to the acoustic signal based on the rotation speed signal, wherein the target phase angle window is used to characterize the range of the tire's rotation angle during the tire's contact with the ground.

[0195] The mapping unit 503 is used to map the acoustic signal onto the acoustic feature basis corresponding to the target phase angle window to obtain the target feature vector corresponding to the acoustic signal. The acoustic feature basis is used to characterize the acoustic features of the target phase angle window when the tire is in normal operating condition, and the target feature vector is used to characterize the expected acoustic features of the acoustic signal on the acoustic feature basis.

[0196] The second determining unit 504 is used to determine the initial detection result of the tire based on the initial feature vector and the target feature vector corresponding to the acoustic signal. The initial feature vector is used to characterize the initial acoustic features corresponding to the acoustic signal within the target phase angle window, and the initial detection result is used to characterize the degree to which the initial acoustic features deviate from the expected acoustic features.

[0197] The third determining unit 505 is used to determine the tire detection result based on the initial detection result and the tire pressure change trend information, wherein the detection result is used to characterize whether the tire is abnormal.

[0198] Optionally, the first determining unit 502 is further configured to: align the acoustic signal and the rotational speed signal on the time axis based on a first timestamp corresponding to the rotational speed signal and a second timestamp corresponding to the acoustic signal, wherein the first timestamp is used to characterize the generation time of the rotational speed signal and the second timestamp is used to characterize the generation time of the acoustic signal; determine the rotational phase angle corresponding to the tire based on the aligned rotational speed signal; and determine the phase angle window corresponding to the rotational phase angle as the target phase angle window corresponding to the aligned acoustic signal.

[0199] Optionally, the mapping unit 503 is further configured to: convert the acoustic signal from the time domain to the frequency domain using a fast Fourier transform to obtain the frequency domain signal corresponding to the acoustic signal; extract the Mel spectrum feature vector from the frequency domain signal, wherein the Mel spectrum feature vector is used to characterize the key frequencies in the frequency domain signal; and map the Mel spectrum feature vector onto the acoustic feature basis corresponding to the target phase angle window to obtain the target feature vector corresponding to the acoustic signal.

[0200] Optionally, the second determining unit 504 is further configured to: determine the difference vector between the initial feature vector and the target feature vector corresponding to the acoustic signal; and determine the initial detection result of the tire based on the Euclidean norm of the difference vector, wherein the Euclidean norm is used to quantify the distance between the initial feature vector and the target feature vector, and the Euclidean norm is positively correlated with the degree to which the initial acoustic features represented by the initial detection result deviate from the expected acoustic features.

[0201] Optionally, the third determining unit 505 is further configured to: periodically verify the initial detection result to obtain a verification result, wherein the verification result is used to indicate whether there is a periodic abnormal event in the tire within the target phase angle window; in response to the verification result indicating that there is a periodic abnormal event in the tire within the target phase angle window, determine the tire pressure change trend information of the tire based on the tire pressure data of the tire within a first preset time window; determine the acoustic anomaly confidence level based on the degree to which the initial acoustic features characterized by the initial detection result deviate from the expected acoustic features, and determine the tire pressure anomaly confidence level based on the tire pressure change trend information, wherein the acoustic anomaly confidence level is used to characterize the degree of certainty of the acoustic anomaly of the tire, and the tire pressure anomaly confidence level is used to characterize the degree of certainty of the tire pressure anomaly; and determine the detection result of the tire based on the acoustic anomaly confidence level and the tire pressure anomaly confidence level.

[0202] Optionally, the third determining unit 505 is further configured to: obtain the Euclidean mean values ​​corresponding to multiple phase angle windows of the tire, wherein the Euclidean mean value is used to characterize the average value of the Euclidean norms corresponding to multiple phase angle windows within a second preset time window, the multiple phase angle windows are obtained by equally dividing the rotation period of the tire, and the multiple phase angle windows include a target phase angle window; compare the first Euclidean mean value corresponding to the target phase angle window with the second Euclidean mean value corresponding to the phase angle windows other than the target phase angle window within the multiple phase angle windows to obtain a comparison result; in response to the comparison result indicating that the first Euclidean mean value is greater than the second Euclidean mean value, determine the periodicity index corresponding to the target phase angle window based on the first Euclidean mean value corresponding to the target phase angle window and the rotation angle range corresponding to the target phase angle window, wherein the periodicity index is used to characterize the linear correlation between the first Euclidean mean value and the rotation angle range; and determine the verification result based on the periodicity index and the periodicity index threshold.

[0203] Optionally, the third determining unit 505 is further configured to: determine that the verification result indicates that a periodic abnormal event exists in the tire within the target phase angle window in response to the periodic index being less than or equal to the periodic index threshold; and determine that the verification result indicates that no periodic abnormal event exists in the tire within the target phase angle window in response to the periodic index being greater than the periodic index threshold.

[0204] Optionally, the third determining unit 505 is further configured to: fuse the acoustic anomaly confidence level with the tire pressure anomaly confidence level to obtain the tire's comprehensive confidence level, wherein the comprehensive confidence level is used to characterize the degree of certainty that the tire has an anomaly; and in response to the comprehensive confidence level being greater than the confidence level threshold, determine that the tire's detection result indicates that the tire has an anomaly.

[0205] Optionally, the device 500 is also used to: trigger a tire alarm in response to a tire malfunction.

[0206] In the aforementioned tire detection device for vehicles, the acoustic signal generated when the tire contacts the ground is used, combined with a real-time rotational speed signal, to locate the target phase angle window corresponding to the tire. The acquired acoustic signal is then mapped onto the acoustic feature substrate corresponding to the target phase angle window to obtain the expected acoustic feature (i.e., the normal acoustic feature corresponding to the acoustic signal). By comparing this expected acoustic feature with the initial acoustic feature (i.e., the acoustic feature corresponding to the acoustic signal itself), it is possible to determine whether there is an abnormality in the tire. This can be verified by combining tire pressure change trend information, accurately identifying tire abnormalities and reducing false alarm rates. In other words, this application can determine whether there is an abnormality in the tire by analyzing the tire's acoustic signal in real time. It can issue an early warning through acoustic anomaly detection in the early stages of tire damage, before the air pressure drops significantly, providing drivers with sufficient time to address tire problems, avoid potential safety hazards, improve driving safety, and thus solve the technical problem of not being able to detect abnormal phenomena in vehicle tires in a timely manner.

[0207] Embodiments of this application also provide a vehicle, including: a memory storing an executable program; and a processor for running the program, wherein the program executes the methods described in various embodiments of this application when it runs.

[0208] Embodiments of this application also provide a computer-readable storage medium including a stored executable program, wherein, when the executable program is running, it controls the device where the computer-readable storage medium is located to perform the methods of various embodiments of this application.

[0209] Embodiments of this application also provide a computer program product, including a computer program that, when executed by a processor, implements the methods of various embodiments of this application.

[0210] Embodiments of this application also provide a computer program product, including a non-volatile computer-readable storage medium for storing a computer program that, when executed by a processor, implements the methods in various embodiments of this application.

[0211] Embodiments of this application also provide a computer program that, when executed by a processor, implements the methods described in the various embodiments of this application.

[0212] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0213] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0214] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0215] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0216] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0217] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for inspecting vehicle tires, characterized in that, include: During vehicle operation, acoustic signals generated by the contact between the vehicle's tires and the ground, as well as the tire rotation speed signal, are acquired. Based on the rotational speed signal, a target phase angle window corresponding to the acoustic signal is determined, wherein the target phase angle window is used to characterize the range of the tire's rotation angle during the contact process between the tire and the ground; The acoustic signal is mapped onto the acoustic feature basis corresponding to the target phase angle window to obtain the target feature vector corresponding to the acoustic signal. The acoustic feature basis is used to characterize the acoustic features of the target phase angle window when the tire is in normal operating condition, and the target feature vector is used to characterize the expected acoustic features of the acoustic signal on the acoustic feature basis. Based on the initial feature vector corresponding to the acoustic signal and the target feature vector, the initial detection result of the tire is determined, wherein the initial feature vector is used to characterize the initial acoustic feature corresponding to the acoustic signal within the target phase angle window, and the initial detection result is used to characterize the degree to which the initial acoustic feature deviates from the expected acoustic feature; Based on the initial detection results and the tire pressure change trend information of the tire, the detection result of the tire is determined, wherein the detection result is used to characterize whether the tire is abnormal.

2. The method according to claim 1, characterized in that, Based on the rotational speed signal, determining the target phase angle window corresponding to the acoustic signal includes: Based on the first timestamp corresponding to the rotational speed signal and the second timestamp corresponding to the acoustic signal, the acoustic signal and the rotational speed signal are aligned on the time axis, wherein the first timestamp is used to characterize the generation time of the rotational speed signal and the second timestamp is used to characterize the generation time of the acoustic signal; Based on the aligned rotational speed signal, the rotational phase angle corresponding to the tire is determined; The phase angle window corresponding to the rotated phase angle is determined as the target phase angle window corresponding to the aligned acoustic signal.

3. The method according to claim 1, characterized in that, Mapping the acoustic signal onto the acoustic feature basis corresponding to the target phase angle window to obtain the target feature vector corresponding to the acoustic signal includes: The acoustic signal is converted from the time domain to the frequency domain using a fast Fourier transform to obtain the frequency domain signal corresponding to the acoustic signal. Mel-spectral feature vectors are extracted from the frequency domain signal, wherein the Mel-spectral feature vectors are used to characterize key frequencies in the frequency domain signal; The Mel spectrum feature vector is mapped onto the acoustic feature basis corresponding to the target phase angle window to obtain the target feature vector corresponding to the acoustic signal.

4. The method according to claim 1, characterized in that, Based on the initial feature vector corresponding to the acoustic signal and the target feature vector, the initial detection result of the tire is determined, including: Determine the difference vector between the initial feature vector corresponding to the acoustic signal and the target feature vector; The initial detection result of the tire is determined based on the Euclidean norm of the difference vector, wherein the Euclidean norm is used to quantify the distance between the initial feature vector and the target feature vector, and the Euclidean norm is positively correlated with the degree to which the initial acoustic features represented by the initial detection result deviate from the expected acoustic features.

5. The method according to claim 1, characterized in that, Based on the initial detection results and the tire pressure change trend information, the detection results of the tire are determined, including: The initial detection results are periodically verified to obtain verification results, wherein the verification results are used to indicate whether there are periodic abnormal events in the tire within the target phase angle window; In response to the verification result indicating that the tire has the periodic abnormal event within the target phase angle window, the tire pressure change trend information of the tire is determined based on the tire pressure data of the tire within a first preset time window; Based on the degree to which the initial acoustic features deviate from the expected acoustic features as characterized by the initial detection results, an acoustic anomaly confidence level is determined, and based on the tire pressure change trend information, a tire pressure anomaly confidence level is determined, wherein the acoustic anomaly confidence level is used to characterize the degree of certainty that the tire has an acoustic anomaly, and the tire pressure anomaly confidence level is used to characterize the degree of certainty that the tire has a tire pressure anomaly. The detection result of the tire is determined based on the acoustic anomaly confidence level and the tire pressure anomaly confidence level.

6. The method according to claim 5, characterized in that, The initial detection results are periodically verified to obtain verification results, including: Obtain the Euclidean mean value corresponding to multiple phase angle windows of the tire, wherein the Euclidean mean value is used to characterize the average value of the Euclidean norm corresponding to the multiple phase angle windows within a second preset time window, the multiple phase angle windows are obtained by equally dividing the rotation period of the tire, and the multiple phase angle windows include the target phase angle window. The first Euclidean mean value corresponding to the target phase angle window is compared with the second Euclidean mean value corresponding to the phase angle windows other than the target phase angle window in the plurality of phase angle windows to obtain the comparison result. In response to the comparison result indicating that the first Euclidean mean is greater than the second Euclidean mean, a periodicity index corresponding to the target phase angle window is determined based on the first Euclidean mean corresponding to the target phase angle window and the rotation angle range corresponding to the target phase angle window, wherein the periodicity index is used to characterize the linear correlation between the first Euclidean mean and the rotation angle range. The verification result is determined based on the periodicity index and the periodicity index threshold.

7. The method according to claim 6, characterized in that, Based on the periodic indicator and the periodic indicator threshold, the verification result is determined, including: In response to the periodicity index being less than or equal to the periodicity index threshold, the verification result is determined to indicate that the tire has the periodic abnormal event within the target phase angle window; In response to the periodicity index being greater than the periodicity index threshold, the verification result is determined to indicate that the tire does not have the periodic abnormal event within the target phase angle window.

8. The method according to claim 5, characterized in that, Based on the acoustic anomaly confidence level and the tire pressure anomaly confidence level, the detection result of the tire is determined, including: The acoustic anomaly confidence level and the tire pressure anomaly confidence level are fused to obtain the overall confidence level of the tire, wherein the overall confidence level is used to characterize the degree of certainty that the tire has an anomaly. In response to the overall confidence level being greater than a confidence threshold, it is determined that the detection result of the tire indicates that the tire has the abnormality; The method further includes: In response to the abnormality in the tire, an alarm is triggered for the tire.

9. A vehicle, characterized in that, include: Memory, which stores executable programs; A processor for running the program, wherein the program, when running, performs the method according to any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored executable program, wherein, when the executable program is executed, it controls the device on which the storage medium is located to perform the method according to any one of claims 1 to 8.

Images