Vehicle abnormity identification method and device, electronic equipment and vehicle
By converting the vehicle abnormal sound audio file into a unified format and using Euclidean distance comparison, the problem of inconsistent vehicle abnormal sound audio format is solved, and the cause of vehicle abnormal sound is quickly and accurately positioned, improving the troubleshooting efficiency.
Patent Information
- Application Number
- CN202411682900.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-22
- Publication Date
- 2025-03-28
AI Technical Summary
The prior art cannot quickly and accurately determine the cause of vehicle abnormality based on vehicle abnormal noise, mainly because the audio format of abnormal noise is not uniform, which makes it impossible to quickly compare and match.
By converting the pending audio file into the pending audio data in a preset format, it is consistent with the format of the pre-stored multiple aberrant audio data, the similarity between the two is determined using the Euclidean distance, thereby quickly and accurately determining the cause of the vehicle abnormality.
It realizes the rapid and accurate positioning of vehicle abnormal causes based on the pending audio files, reduces labor costs and time costs, and improves the efficiency of troubleshooting and repairs.
Smart Images
Figure CN119719935A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the field of vehicle control technology, and in particular to a method, device, electronic equipment and vehicle for identifying vehicle abnormality. Background Art
[0002] With the development of the vehicle industry, the cause of the abnormality can be diagnosed based on the abnormal sound of the vehicle. However, due to the non-uniform audio format of the abnormal sound of the vehicle, the cause of the abnormality of the vehicle cannot be determined quickly and accurately based on the abnormal sound of the vehicle.
[0003] In view of this, how to quickly and accurately determine the cause of vehicle abnormality based on vehicle abnormal noise has become a technical problem that needs to be solved urgently. Summary of the invention
[0004] In view of this, the purpose of the present disclosure is to provide a vehicle abnormality identification method, device, electronic equipment and vehicle to solve the problem in the prior art that the cause of the vehicle abnormality cannot be determined quickly and accurately based on the vehicle abnormal sound.
[0005] Based on the above objectives, the first aspect of the present disclosure provides a method for identifying vehicle abnormalities, the method comprising:
[0006] Determine an audio file to be processed, and process the audio file to be processed to obtain audio data to be processed in a preset format; wherein the audio data to be processed in the preset format is consistent with the format of a plurality of pre-stored abnormal sound audio data;
[0007] The similarity between the audio data to be processed in the preset format and each abnormal sound audio data is determined by using the Euclidean distance, target abnormal sound audio data with high similarity is determined from the multiple abnormal sound audio data, and the cause of the vehicle abnormality is determined based on the target abnormal sound audio data.
[0008] Based on the same inventive concept, the second aspect of the present disclosure proposes a vehicle abnormality identification device, comprising:
[0009] A processing module is configured to determine an audio file to be processed, and process the audio file to be processed to obtain audio data to be processed in a preset format; wherein the audio data to be processed in the preset format is consistent with the format of a plurality of pre-stored abnormal sound audio data;
[0010] The determination module is configured to determine the similarity between the audio data to be processed in the preset format and each abnormal sound audio data by using the Euclidean distance, determine the target abnormal sound audio data with high similarity from the multiple abnormal sound audio data, and determine the cause of the vehicle abnormality based on the target abnormal sound audio data.
[0011] Based on the same inventive concept, the third aspect of the present disclosure proposes an electronic device, including a memory, a processor, and a computer program stored in the memory and executable by the processor, wherein the processor implements the method as described above when executing the computer program.
[0012] Based on the same inventive concept, the fourth aspect of the present disclosure proposes a vehicle, which includes the vehicle abnormality identification device described in the second aspect or the electronic device described in the third aspect.
[0013] As can be seen from the above, the present disclosure provides a method, device, electronic device and vehicle for identifying vehicle abnormalities. An audio file to be processed is determined, and the audio file to be processed is processed to obtain audio data to be processed in a preset format. By converting the audio file to be processed into audio data to be processed in a preset format, the format of the audio data to be processed in the preset format is made consistent with the format of multiple abnormal sound audio data stored in advance, so as to facilitate the comparison of the audio data to be processed in the preset format with the multiple abnormal sound audio data. Each abnormal sound audio data corresponds to a cause of vehicle abnormality. The Euclidean distance is used to determine the similarity between the audio data to be processed in the preset format and each abnormal sound audio data, and the target abnormal sound audio data with high similarity is determined from the multiple abnormal sound audio data. Based on the target abnormal sound audio data, the cause of the vehicle abnormality can be quickly and accurately determined. In this way, the cause of the vehicle abnormality can be quickly and accurately located according to the audio file to be processed, which is convenient for troubleshooting and maintenance, and there is no need to manually determine the cause of the vehicle abnormality, reducing labor costs and time costs. BRIEF DESCRIPTION OF THE DRAWINGS
[0014] In order to more clearly illustrate the technical solutions in the present disclosure or related technologies, the drawings required for use in the embodiments or related technical descriptions are briefly introduced below. Obviously, the drawings described below are only embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0015] Figure 1 A flowchart of a method for identifying vehicle abnormalities according to an embodiment of the present disclosure;
[0016] Figure 2 It is a flow chart of the vehicle abnormal sound audio recognition method according to an embodiment of the present disclosure;
[0017] Figure 3 A schematic diagram of the structure of a vehicle abnormality recognition device according to an embodiment of the present disclosure;
[0018] Figure 4 It is a schematic diagram of the structure of an electronic device according to an embodiment of the present disclosure. DETAILED DESCRIPTION
[0019] In order to make the objectives, technical solutions and advantages of the present disclosure more clearly understood, the present disclosure is further described in detail below in combination with specific embodiments and with reference to the accompanying drawings.
[0020] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the embodiments of the present disclosure should be understood by people with ordinary skills in the field to which the present disclosure belongs. The "first", "second" and similar words used in the embodiments of the present disclosure do not indicate any order, quantity or importance, but are only used to distinguish different components. "Including" or "comprising" and similar words mean that the elements or objects appearing before the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.
[0021] Based on the description of the background technology, with the rapid development of the vehicle industry and the continuous improvement of people's requirements for driving experience, vehicle maintenance and fault diagnosis have become particularly important. Abnormal noise in vehicles is a common manifestation of vehicle failure, and important mechanical health information is often hidden behind it. Currently, visual inspection and experience-based fault judgment is time-consuming and inefficient, and often cannot meet the needs of fast and accurate diagnosis.
[0022] In this context, vehicle abnormal sound audio recognition came into being, providing a new perspective for vehicle abnormal sound diagnosis. It can identify samples similar to the abnormal sound audio to be checked from a massive abnormal sound sample library in a short time, and quickly locate the problem based on comparative analysis of known fault modes. This not only improves the accuracy of diagnosis, but also greatly shortens the troubleshooting cycle and reduces the diagnosis cost, which is of great significance to improving the overall technical level of vehicle abnormal sound diagnosis.
[0023] During vehicle R&D, manufacturing, road test inspection, and after-sales market, a large number of abnormal noise problems will be encountered. In the current information-developed era, abnormal noise problems are often recorded by mobile phones in the form of video or audio for collection and transmission analysis. The R&D departments of various OEMs generally have such abnormal noise audio or video libraries for case collection and accumulation, providing directional guidance for abnormal noise problem troubleshooting. However, there are the following problems in the actual application of these sample libraries:
[0024] 1. The sample library often contains a lot of content. When troubleshooting the current abnormal sound problem, it takes a long time to manually compare the abnormal sound case samples, and most of them are subjective judgments with low accuracy;
[0025] 2. The sample library includes video files and audio files, and there is no unified format for automatic retrieval;
[0026] 3. The video files and audio files in the sample library have different durations, and different recording devices lead to different sampling frequencies and different numbers of sampling channels, which makes it impossible to directly perform corresponding comparison and matching.
[0027] As mentioned above, how to quickly and accurately determine the cause of vehicle abnormality based on vehicle abnormal noise has become an important research issue.
[0028] Based on the above description, if Figure 1 As shown, the vehicle abnormality identification method proposed in this embodiment includes:
[0029] Step 101, determining an audio file to be processed, and processing the audio file to be processed to obtain audio data to be processed in a preset format; wherein the audio data to be processed in the preset format is consistent with the format of a plurality of pre-stored abnormal sound audio data.
[0030] In a specific implementation, the audio and / or video to be processed are received, and the audio and / or video to be processed are processed to obtain an audio file to be processed. The audio file to be processed is an audio file when an abnormality occurs in the vehicle.
[0031] In addition, multiple abnormal sound audio files are pre-stored, and multiple abnormal sound audio data in a preset format are obtained by processing the multiple abnormal sound audio files. In this way, the audio data to be processed and the multiple abnormal sound audio data are both in a preset format, which can achieve the unification of the formats of the audio data to be processed and the multiple abnormal sound audio data, so that the audio data to be processed in the preset format can be quickly compared with the multiple abnormal sound audio data.
[0032] Step 102, using Euclidean distance, determine the similarity between the audio data to be processed in the preset format and each abnormal sound audio data, determine target abnormal sound audio data with high similarity from the multiple abnormal sound audio data, and determine the cause of the vehicle abnormality based on the target abnormal sound audio data.
[0033] In specific implementation, Euclidean distance is also called Euclidean distance, which is the real distance between two points in m-dimensional space. In the field of pattern recognition, Euclidean distance is often used to measure the similarity between two patterns or objects.
[0034] A plurality of abnormal sound audio data are pre-stored, each of which corresponds to an abnormal cause. The similarity between the audio data to be processed and each abnormal sound audio data is determined; and target abnormal sound audio data having a high similarity to the audio data to be processed in a preset format is determined from the plurality of abnormal sound audio data.
[0035] For example, the pre-stored multiple abnormal sound audio data include: a first abnormal sound audio, a second abnormal sound audio, and a third abnormal sound audio. The first abnormal sound audio corresponds to a first abnormal cause, the second abnormal sound audio corresponds to a second abnormal cause, and the third abnormal sound audio corresponds to a third abnormal cause. From the multiple abnormal sound audio data, a second abnormal sound audio having a high similarity to the audio data to be processed in a preset format is determined, and the abnormal cause of the vehicle is determined to be the second abnormal cause.
[0036] In addition, when there is no target abnormal sound audio data with high similarity to the audio data to be processed among the multiple abnormal sound audio data, after determining the cause of the vehicle abnormality, the audio data to be processed is associated with the cause of the vehicle abnormality and stored. In this way, the pre-stored abnormal sound audio library can be further enriched, so that the cause of the vehicle abnormality can be quickly and accurately identified based on the audio data to be processed later.
[0037] The audio file to be processed can be collected by a mobile terminal device or a vehicle end. For example, during the vehicle R&D, manufacturing, testing and after-sales process, the vehicle sends out audio to be processed, and the audio to be processed is collected by a mobile terminal or a vehicle end, and the audio to be processed is converted into audio data to be processed in a preset format. The audio data to be processed in the preset format is compared with a plurality of pre-stored abnormal sound audio data to determine the cause of the vehicle abnormality.
[0038] Through the above embodiment, the audio file to be processed is determined, and the audio file to be processed is processed to obtain the audio data to be processed in a preset format. By converting the audio file to be processed into the audio data to be processed in a preset format, the audio data to be processed in the preset format is made consistent with the format of the multiple abnormal sound audio data stored in advance, so as to facilitate the comparison of the audio data to be processed in the preset format with the multiple abnormal sound audio data. Each abnormal sound audio data corresponds to a cause of vehicle abnormality. The similarity between the audio data to be processed in the preset format and each abnormal sound audio data is determined by using the Euclidean distance, and the target abnormal sound audio data with high similarity is determined from the multiple abnormal sound audio data. Based on the target abnormal sound audio data, the cause of the vehicle abnormality can be quickly and accurately determined. In this way, the cause of the vehicle abnormality can be quickly and accurately located according to the audio file to be processed, which is convenient for troubleshooting and maintenance, and there is no need to manually determine the cause of the vehicle abnormality, thereby reducing labor costs and time costs.
[0039] In some embodiments, step 101 includes:
[0040] Step 1011, read and process the audio file to be processed to obtain the read audio data to be processed.
[0041] Step 1012, normalizing the read audio data to be processed to obtain audio data to be processed in a preset format.
[0042] In specific implementation, the reading process includes: decoding process and conversion process. The audio file to be processed is decoded to obtain decoded audio data to be processed. The decoded audio data to be processed is converted to obtain a double-precision floating-point number, and the double-precision floating-point number is used as the read audio data to be processed.
[0043] Normalization processing includes: single-channel conversion processing, sampling rate conversion processing and duration conversion processing. Convert the read audio data to be processed into single-channel audio data to be processed. Convert the single-channel audio data to be processed into audio data to be processed at a preset sampling rate. Convert the audio data to be processed at a preset sampling rate into audio data to be processed at a preset duration. Use the audio data to be processed at a preset duration as audio data to be processed in a preset format.
[0044] In addition, multiple abnormal sound audio files are pre-stored. The multiple abnormal sound audio files are read and processed to obtain multiple abnormal sound audio data after reading. The multiple abnormal sound audio data after reading are normalized to obtain multiple abnormal sound audio data in a preset format.
[0045] Through the above scheme, by reading and processing the audio file to be processed, the read audio data to be processed can be easily processed later with higher accuracy. By normalizing the read audio data to be processed, the formats of the preset audio data to be processed and the multiple abnormal sound audio data can be unified, so that the preset audio data to be processed and the multiple abnormal sound audio data can be quickly compared and processed.
[0046] In some embodiments, step 1011 includes:
[0047] Step 1011A: determine a target decoder according to the file header information of the audio file to be processed, and use the target decoder to decode the audio file to be processed to obtain decoded audio data to be processed.
[0048] In specific implementation, the file header information includes: sampling rate, bit depth and number of channels.
[0049] A plurality of decoders are pre-stored; a target decoder is determined from the plurality of decoders according to the file header information of the audio file to be processed; and the audio file to be processed is decoded by the target decoder to obtain decoded audio data to be processed.
[0050] Step 1011B: convert the decoded audio data to be processed to obtain a double-precision floating-point number, and use the double-precision floating-point number as the read audio data to be processed.
[0051] In specific implementation, double precision floating point number (double) is a data type used by computers, using 64 bits (8 bytes) to store a floating point number. It can represent 15 or 16 significant digits in decimal, and the absolute value range of the number it can represent is approximately: -1.79E+308 to +1.79E+308.
[0052] The decoded audio data to be processed is converted to obtain a double-precision floating point number, and the double-precision floating point number is used as the audio data to be processed after reading. In this way, due to the large number of digits of the double-precision floating point number, the audio data to be processed after reading has higher precision and is more accurate.
[0053] Through the above scheme, the audio file to be processed is decoded and converted into the audio data to be processed, which is convenient for subsequent processing of the audio data to be processed. By converting the decoded audio data to be processed into double-precision floating-point numbers as the audio data to be processed after reading, the precision of the audio data to be processed after reading is higher and more accurate.
[0054] In some embodiments, step 1012 includes:
[0055] Step 1012A: determine the number of channels of the read audio data to be processed, and convert the read audio data to be processed into single-channel audio data to be processed according to the number of channels.
[0056] During specific implementation, the number of channels of the read data to be processed is determined, and the read data to be processed is averaged to obtain single-channel audio data to be processed.
[0057]
[0058] Among them, y(t) is the single-channel audio data to be processed, t is the time point, x 1 (t) is the sample value of the audio data to be processed after reading the first channel, x 2 (t) is the sample value of the audio data to be processed after reading the second channel, x n (t) is the sample value of the audio data to be processed after being read out of the nth channel, and n is the number of channels of the audio data to be processed after being read out.
[0059] Step 1012B: convert the single-channel audio data to be processed into audio data to be processed at a preset sampling rate.
[0060] In a specific implementation, the audio sampling rate of the audio data to be processed of a single channel is obtained; the audio sampling rate is compared with the preset sampling rate; in response to determining that the audio sampling rate is greater than the preset sampling rate, the audio sampling rate of the audio data to be processed of the single channel is converted to the preset sampling rate by a down-sampling method; in response to determining that the audio sampling rate is less than the preset sampling rate, the audio sampling rate of the audio data to be processed of the single channel is converted to the preset sampling rate by an up-sampling method.
[0061] Among them, undersampling and oversampling are commonly used methods in data preprocessing, mainly used to process unbalanced data sets. Undersampling is to balance the data set by reducing the number of majority class samples, and the undersampling methods include random undersampling and cluster undersampling. Oversampling is to balance the data set by increasing the number of minority class samples, and the oversampling methods include random upsampling.
[0062] For example, the preset sampling rate is 44100Hz. Determine whether the audio sampling rate of the single-channel audio data to be processed is 44100Hz, that is, determine whether the audio sampling rate Fs=44100Hz; when the audio sampling rate Fs>44100Hz, use downsampling to convert the single-channel audio data to be processed into audio data to be processed with a sampling rate of 44100Hz; when the audio sampling rate Fs<44100Hz, use upsampling to convert the single-channel audio data to be processed into audio data to be processed with a sampling rate of 44100Hz.
[0063] Step 1012C: convert the to-be-processed audio data of the preset sampling rate into to-be-processed audio data of the preset duration, and use the to-be-processed audio data of the preset duration as to-be-processed audio data of the preset format.
[0064] In a specific implementation, the to-be-processed audio data in a preset format is to-be-processed audio data of a single channel, a preset sampling rate and a preset duration.
[0065] Obtain the audio duration of the audio data to be processed at a preset sampling rate; compare the audio duration with the preset duration; in response to determining that the audio duration is greater than the preset duration, intercept the audio data to be processed of the preset duration at the front end of the audio data to be processed at the preset sampling rate; in response to determining that the audio duration is less than the preset duration, add a zero value to the end of the audio data to be processed at the preset sampling rate to obtain the audio data to be processed of the preset duration.
[0066] Specifically, the number of sampling points (i.e., length) of the audio data to be processed at a preset sampling rate is obtained; the ratio of the number of sampling points to the preset sampling rate is used as the audio duration. The relationship between the audio duration and the number of sampling points is expressed as: audio duration = number of sampling points / preset sampling rate.
[0067] For example, the preset duration is 10 seconds, and the preset sampling rate is expressed as Fs. Determine whether the audio duration of the to-be-processed audio data of the preset sampling rate is 10 seconds; when the audio duration is greater than 10 seconds, intercept the first 10 seconds (i.e., from the 1st sampling point to the 10×Fs sampling point) of the to-be-processed audio data of the preset sampling rate; when the audio duration is less than 10 seconds, calculate the sampling point m that differs from 10 seconds, establish a signal data composed of m zero values, and add the signal data composed of m zero values to the end of the to-be-processed audio data of the preset sampling rate to obtain 10 seconds of to-be-processed audio data.
[0068] Through the above scheme, by converting the read audio data to be processed into single-channel audio data to be processed, converting the single-channel audio data to be processed into audio data to be processed at a preset sampling rate, and converting the audio data to be processed at a preset sampling rate into audio data to be processed with a preset duration as audio data to be processed in a preset format, the audio data to be processed in a preset format and a plurality of abnormal sound audio data can be unified in format in terms of the number of channels, sampling rate and duration, so that the audio data to be processed in a preset format can be quickly compared and processed with the plurality of abnormal sound audio data.
[0069] In some embodiments, step 102 includes:
[0070] Step 1021: vectorize the to-be-processed audio data in the preset format to obtain a to-be-processed feature vector, and vectorize the multiple abnormal sound audio data to obtain multiple abnormal sound feature vectors.
[0071] Step 1022: Determine the similarity between the audio data to be processed in the preset format and each abnormal sound audio data according to the feature vector to be processed and the plurality of abnormal sound feature vectors by using the Euclidean distance.
[0072] Step 1023: determine target abnormal sound audio data having a high similarity to the audio data to be processed in the preset format from the multiple abnormal sound audio data.
[0073] In specific implementation, the number of decomposition layers of the audio data to be processed in a preset format is determined, the audio data to be processed in a preset format is decomposed according to the number of decomposition layers using wavelet basis functions to obtain decomposition coefficients to be processed, and the decomposition coefficients to be processed are extracted to obtain feature vectors to be processed.
[0074] Determine the decomposition level of the abnormal sound audio data. Use the wavelet basis function to decompose the abnormal sound audio data according to the decomposition level to obtain abnormal sound decomposition coefficients, and extract the abnormal sound decomposition coefficients to obtain abnormal sound feature vectors.
[0075] According to the feature vector to be processed and the plurality of abnormal sound feature vectors, the similarity between the audio data to be processed in a preset format and each abnormal sound audio data is determined. From the plurality of abnormal sound audio data, target abnormal sound audio data having a high similarity to the audio data to be processed in the preset format is determined.
[0076] Through the above scheme, by decomposing and processing the audio data to be processed in a preset format, the feature vector to be processed can be accurately obtained. By decomposing and processing the abnormal sound audio data, the abnormal sound feature vector can be accurately obtained. According to the feature vector to be processed and multiple abnormal sound feature vectors, the similarity between the audio data to be processed in a preset format and each abnormal sound audio data can be accurately determined, so that the target abnormal sound audio data with high similarity to the audio data to be processed in the preset format can be accurately determined from the multiple abnormal sound audio data according to the similarity, so as to determine the cause of the vehicle abnormality based on the target abnormal sound audio data.
[0077] In some embodiments, step 1021 includes:
[0078] Step 1021A, determining the number of decomposition layers of the to-be-processed audio data in the preset format, and determining the number of decomposition times based on the number of decomposition layers.
[0079] In specific implementation, the decomposition layer number L of the audio data to be processed in a preset format is determined, and the number of decompositions can be determined based on the decomposition layer number L. The audio data to be processed in the preset format is decomposed using the wavelet basis function and the decomposition layer number L to obtain a feature vector to be processed.
[0080] Step 1021B, in response to determining that the decomposition times are twice, performing a first decomposition process on the audio data to be processed in the preset format using a wavelet basis function to obtain a first decomposition coefficient, performing a second decomposition process on the first decomposition coefficient using a wavelet basis function to obtain a second decomposition coefficient, and using the second decomposition coefficient as a feature vector to be processed.
[0081] In specific implementation, the number of decompositions of the audio to be processed in the preset format is determined according to the number of decomposition layers. 3 When the number of decomposition layers L=2, there are obvious advantages in computational efficiency and accuracy. When the number of decomposition layers L=2, the audio data to be processed in the preset format is decomposed twice to obtain the feature vector to be processed.
[0082] The decomposition coefficients include: approximate coefficients and detail coefficients. The first decomposition process is performed on the audio data to be processed in the preset format to obtain the first approximate coefficient a 1 (k) and the first detail coefficient d 1 (k). The specific formula is as follows:
[0083]
[0084] Among them, a 1 (k) is the first approximation coefficient, d 1 (k) is the first detail coefficient, x(n) is the audio data to be processed in a preset format, is a scaling function about 1,k(n), ψ 1,k(n) is the wavelet function about 1,k(n), ψ 1,k(n) =ψ(2 -1 nk).
[0085] The first approximate coefficient a obtained by the first decomposition process 1 (k) and the first detail coefficient d 1 (k) Perform a second decomposition process to obtain higher-level wavelet packet decomposition coefficients. The specific formula is as follows:
[0086]
[0087] Among them, a j+1 (k) is the second approximate coefficient, d j+1 (k) is the second detail coefficient, is the third approximate coefficient, is the third detail coefficient, a j (n) is the previous approximate coefficient, d j (n) is the previous detail coefficient, is a scaling function about 1,k(n), ψ 1,k(n) is the wavelet function about 1,k(n), ψ 1,k(n) =ψ(2 -1 nk).
[0088] When the decomposition coefficient L=2, the second decomposition coefficient is extracted to obtain a feature vector to be processed. The feature vector to be processed is a feature vector representing the audio data to be processed in a preset format.
[0089] When the decomposition coefficient L>2, the above decomposition process is repeated until the predetermined decomposition level L is reached to obtain all wavelet packet coefficients. The wavelet packet coefficients are extracted to obtain the feature vector to be processed.
[0090] Through the above scheme, according to the number of decomposition layers, the audio data to be processed in the preset format is decomposed twice to obtain the feature vector to be processed. When the number of decomposition layers is 2, the audio data to be processed in the preset format is decomposed twice, so that the second decomposition coefficient obtained is a wavelet packet decomposition coefficient of a higher level, with high calculation efficiency and high accuracy, and the feature vector to be processed can be accurately obtained by extracting the second decomposition coefficient.
[0091] In some embodiments, step 1022 includes:
[0092] Step 1022A: determine the Euclidean distance between the feature vector to be processed and each abnormal sound feature vector.
[0093] In a specific implementation, the Euclidean distance between the feature vector to be processed and the abnormal sound feature vector is obtained by calculating the square root of the sum of squares of differences between the feature vector to be processed and the abnormal sound feature vector in each dimension.
[0094] For example, the feature vector to be processed is p=(p 1 ,p 2 ,…,p n ), the abnormal noise characteristic vector is q=(q 1 ,q 2 ,…,q n ), then the Euclidean distance between the feature vector to be processed and the abnormal sound feature vector is:
[0095]
[0096] Among them, d(p,q) is the Euclidean distance between the feature vector to be processed and the abnormal sound feature vector, and the feature vector to be processed p = (p 1 ,p 2 ,…,p n ) and the abnormal sound characteristic vector q=(q 1 ,q 2 ,…,q n ) are all n-dimensional vectors.
[0097] Step 1022B: Determine the similarity between the to-be-processed audio data in the preset format and each abnormal sound audio data based on the Euclidean distance.
[0098] In specific implementation, the similarity between the audio data to be processed in a preset format and each abnormal sound audio data is obtained by subtracting the Euclidean distance from 1 and performing the percentage calculation. The similarity is expressed as:
[0099] Sim=(1-d)×100%
[0100] Wherein, Sim is the similarity between the audio data to be processed in the preset format and each abnormal sound audio data, and d is the Euclidean distance between the feature vector to be processed and the abnormal sound feature vector. When Sim≤0, it is considered that the audio data to be processed in the preset format is not similar to the abnormal sound audio data.
[0101] The smaller the Euclidean distance between the feature vector to be processed and the abnormal sound feature vector, the closer the feature vector to be processed and the abnormal sound feature vector are, and the higher the similarity between the audio data to be processed in the preset format and the abnormal sound audio data.
[0102] The larger the Euclidean distance between the feature vector to be processed and the abnormal sound feature vector, the farther the feature vector to be processed is from the abnormal sound feature vector, and the lower the similarity between the audio data to be processed in the preset format and the abnormal sound audio data.
[0103] Through the above solution, by determining the Euclidean distance between the feature vector to be processed and each abnormal sound feature vector, based on the Euclidean distance, the similarity between the audio data to be processed in a preset format and each abnormal sound audio data can be accurately determined.
[0104] Through the above embodiment, the audio file to be processed is determined, and the audio file to be processed is processed to obtain the audio data to be processed in a preset format. By converting the audio file to be processed into the audio data to be processed in a preset format, the audio data to be processed in the preset format is made consistent with the format of the multiple abnormal sound audio data stored in advance, so as to facilitate the comparison of the audio data to be processed in the preset format with the multiple abnormal sound audio data. Each abnormal sound audio data corresponds to a cause of vehicle abnormality. The similarity between the audio data to be processed in the preset format and each abnormal sound audio data is determined by using the Euclidean distance, and the target abnormal sound audio data with high similarity is determined from the multiple abnormal sound audio data. Based on the target abnormal sound audio data, the cause of the vehicle abnormality can be quickly and accurately determined. In this way, the cause of the vehicle abnormality can be quickly and accurately located according to the audio file to be processed, which is convenient for troubleshooting and maintenance, and there is no need to manually determine the cause of the vehicle abnormality, thereby reducing labor costs and time costs.
[0105] It should be noted that the embodiments of the present disclosure may be further described in the following manner:
[0106] Figure 2 FIG. 1 is a flow chart of a method for identifying abnormal vehicle sound according to an embodiment of the present disclosure. Figure 2 As shown, the vehicle abnormal sound audio recognition method includes:
[0107] Step 1: read the abnormal sound audio to be identified and obtain audio signal data, sampling rate, and sampling channel parameters.
[0108] Convert the read video and audio files of different formats into audio data of a unified format. Specifically, it includes: automatically selecting the appropriate decoder to process the file according to the file header of the video and audio file (including sampling rate, bit depth, number of channels, etc.), converting the decoded data into double-precision floating point numbers, and further returning the decoded and converted audio data, sampling rate, and number of channels to the calling link for use.
[0109] Step 2: Convert and unify the sampling frequency, number of channels, and duration of audio data.
[0110] Call the different file audio data read and converted above and perform unified processing so that they can be compared later under a unified standard. Specifically include:
[0111] (1) The number of channels in the read file is converted to a single channel, and the data of each channel is superimposed in parallel and averaged to generate a new single-channel data. The principle is: for a given time point t, if the original audio signal has n channels, the sample values of each channel are x 1 (t), x 2 (t), …, x n (t), then the mono signal y(t) can be calculated by the following formula:
[0112]
[0113] (2) Call the sampling rates of the audio data of different files after reading and converting to determine whether the sampling rates are consistent. Because the sampling frequencies of sound signals of video and audio files of different formats are different, in order to unify the processing, the audio data sampling rate is unified to 44100 Hz by combining upsampling and downsampling. The principle is to first determine whether the sampling rate of the read audio signal is equal to 44100 Hz, that is, whether Fs=44100 Hz. If not, 44100 Hz is used as the benchmark. For data with Fs>44100 Hz, the downsampling method is used, and for data with Fs<44100 Hz, the upsampling method is used. The purpose is to unify the sampling frequency to 44100 Hz.
[0114] (3) The audio data converted to a uniform sampling rate is uniformly adjusted for audio duration. For audio duration greater than 10 seconds, the first 10 seconds of audio data are intercepted; for audio data less than 10 seconds, zero values are added to the end to make up for the 10-second duration. The specific method is as follows: read the length of the read and converted audio data, i.e., the number of sampling points, divide it by the sampling frequency 44100Hz to get the duration, and compare it with 10 seconds. For data greater than or equal to 10 seconds, read the first 10 seconds of data, i.e., read the data from the 1st sampling point to the 10×Fs point; for data less than 10 seconds, first calculate the number of sampling points m that differ from the 10-second duration, and then create a signal data consisting of m zero values, which are added to the original signal data to make its duration meet 10 seconds. Filling zero values will not affect the characteristics of the original signal, because zero values will not introduce additional frequency components. For time domain signals, filling zero values is equivalent to extending the silent part of the signal.
[0115] Step 3: extract the wavelet packet decomposition coefficients as the audio feature vector.
[0116] (1) Using the audio data after reading and normalization, the wavelet packet classification coefficient is extracted as the feature vector. The specific method is as follows: call the processed data signal x(t), select the appropriate wavelet basis function and decomposition layer number L, perform the first wavelet decomposition on the signal x(t), and obtain the approximate coefficient a 1 and detail factor d 1 :
[0117]
[0118] in, and ψ 1,k(n) =ψ(2 -1 nk).
[0119] Then the approximate coefficient a obtained by the first decomposition 1 and detail factor d 1 Further decomposition is performed to obtain higher-level wavelet packet decomposition coefficients:
[0120]
[0121] in, and ψ 1,k(n) =ψ(2 -1 nk).
[0122] Repeat the above process until the predetermined decomposition level L is reached, all wavelet packet coefficients are obtained, and then the wavelet packet coefficients are extracted as a feature vector representing the audio signal data (based on the actual debugging selection db 3 Wavelet basis functions and L=2 decomposition levels have obvious advantages in computational efficiency and accuracy).
[0123] (2) Further, the Euclidean distance between the feature vectors of the two extracted audio signals is calculated, and then the Euclidean distance is subtracted from 1, and then the percentage is calculated to obtain a percentage, which is used as the similarity. The closer the distance, the higher the similarity. The calculation principle is as follows: the Euclidean distance between the two n-dimensional vectors is obtained by calculating the square root of the sum of the squares of the differences in each dimension. Specifically, assuming that there are two n-dimensional feature vectors p = (p 1 ,p 2 ,…,p n ) and q=(q 1 ,q 2 ,…,q n ), the Euclidean distance in Euclidean space is defined as:
[0124]
[0125] Similarity principle: The smaller the Euclidean distance, the closer the two vectors are, and the more similar the corresponding samples are.
[0126] Dissimilarity principle: The larger the Euclidean distance, the farther away the two vectors are, and the more dissimilar the corresponding samples are.
[0127] Then, assuming that the similarity is Sim, then Sim=(1-d)×100%. If it is less than or equal to 0, it is considered to be dissimilar.
[0128] After step 3 and before step 4, it also includes: establishing an initialized similarity structure array for subsequent storage of similarity and file names. Specifically, obtain the storage path of the abnormal sound audio and video database folder; traverse all abnormal sound audio and video files in the database under the above storage path, read and convert these files in turn to a unified sampling frequency, single channel, and 10-second data duration, and extract the wavelet packet decomposition coefficients as audio feature vectors in turn, perform Euclidean distance calculations with the feature vectors extracted from the audio to be identified in turn, and further obtain the similarity, and store the similarity and the corresponding abnormal sound file name in the initialized structure array.
[0129] Step 4, sort the similarities in the structure array in descending order and obtain the sorted index; output the result of the first 5 corresponding file names and similarities or a "no similar audio" string.
[0130] The feature vector of the abnormal sound audio file to be identified is obtained according to the above series of methods, and then the abnormal sound case files in the abnormal sound audio and video database are traversed in turn, and the feature vector of each abnormal sound file is calculated according to the above series of methods. The Euclidean distance with the feature vector of the abnormal sound file to be identified is calculated in turn, and a structure array is created. The similarity and file name corresponding to each audio file are stored in this structure array, and the similarity field is extracted from this structure array. The similarity is sorted in descending order, and the sorted index is obtained. The first 5 corresponding file names and similarities are output and displayed. If there is no matching similarity calculation result, the "No similar audio" string is displayed to complete the matching and identification of the abnormal sound audio file to be identified.
[0131] Through the above embodiments, the problem of diverse and non-uniform formats of abnormal sound database files can be solved. The problems of inconsistent sampling channels, inconsistent file sampling frequencies, and inconsistent file durations of abnormal sound database files can be solved. The database files corresponding to the abnormal sound audio to be identified can be quickly and efficiently retrieved and matched to complete the abnormal sound identification process. It can be applied to scenarios such as vehicle abnormal sound performance loading, inspection, and after-sales market, and can transmit and server-side retrieve abnormal sound audio and video files recorded on site, improve the efficiency of abnormal sound problem troubleshooting and identification, and save labor costs and time costs for problem handling and troubleshooting.
[0132] It should be noted that the method of the embodiment of the present disclosure can be performed by a single device, such as a computer or a server. The method of the present embodiment can also be applied in a distributed scenario and completed by multiple devices cooperating with each other. In the case of such a distributed scenario, one of the multiple devices can only perform one or more steps in the method of the embodiment of the present disclosure, and the multiple devices will interact with each other to complete the described method.
[0133] It should be noted that the above describes some embodiments of the present disclosure. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the above embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0134] Based on the same inventive concept, corresponding to any of the above-mentioned embodiment methods, the present disclosure also provides a vehicle abnormality identification device.
[0135] refer to Figure 3 The vehicle abnormality recognition device comprises:
[0136] The processing module 301 is configured to determine an audio file to be processed, and process the audio file to be processed to obtain audio data to be processed in a preset format; wherein the audio data to be processed in the preset format is consistent with the format of a plurality of pre-stored abnormal sound audio data;
[0137] The determination module 302 is configured to determine the similarity between the audio data to be processed in the preset format and each abnormal sound audio data by using the Euclidean distance, determine the target abnormal sound audio data with high similarity from the multiple abnormal sound audio data, and determine the cause of the vehicle abnormality based on the target abnormal sound audio data.
[0138] In some embodiments, the processing module 301 includes:
[0139] A reading and processing unit is configured to read and process the audio file to be processed to obtain the read audio data to be processed;
[0140] The normalization processing unit is configured to perform normalization processing on the read audio data to be processed to obtain the audio data to be processed in a preset format.
[0141] In some embodiments, the reading processing unit includes:
[0142] A decoding processing subunit is configured to determine a target decoder according to the file header information of the audio file to be processed, and use the target decoder to decode the audio file to be processed to obtain decoded audio data to be processed;
[0143] The conversion processing subunit is configured to perform conversion processing on the decoded audio data to be processed to obtain a double-precision floating point number, and use the double-precision floating point number as the read audio data to be processed.
[0144] In some embodiments, the normalization processing unit includes:
[0145] a channel conversion subunit, configured to determine the number of channels of the read audio data to be processed, and convert the read audio data to be processed into single-channel audio data to be processed according to the number of channels;
[0146] A sampling rate conversion subunit, configured to convert the single-channel audio data to be processed into audio data to be processed at a preset sampling rate;
[0147] The duration conversion subunit is configured to convert the to-be-processed audio data of the preset sampling rate into the to-be-processed audio data of the preset duration, and use the to-be-processed audio data of the preset duration as the to-be-processed audio data of the preset format.
[0148] In some embodiments, the determining module 302 includes:
[0149] A vector processing unit is configured to perform vector processing on the to-be-processed audio data in the preset format to obtain a to-be-processed feature vector, and to perform vector processing on the plurality of abnormal sound audio data to obtain a plurality of abnormal sound feature vectors;
[0150] A similarity determination unit is configured to determine the similarity between the to-be-processed audio data in the preset format and each abnormal sound audio data according to the to-be-processed feature vector and the plurality of abnormal sound feature vectors by using Euclidean distance;
[0151] The target abnormal sound audio data determining unit is configured to determine, from the plurality of abnormal sound audio data, target abnormal sound audio data having a high similarity to the audio data to be processed in the preset format.
[0152] In some embodiments, the vectorized processing unit includes:
[0153] A decomposition number determination subunit is configured to determine the number of decomposition layers of the to-be-processed audio data in the preset format, and determine the decomposition number based on the number of decomposition layers;
[0154] The decomposition processing subunit is configured to, in response to determining that the decomposition times are twice, perform a first decomposition processing on the audio data to be processed in the preset format using a wavelet basis function to obtain a first decomposition coefficient, perform a second decomposition processing on the first decomposition coefficient using a wavelet basis function to obtain a second decomposition coefficient, and use the second decomposition coefficient as a feature vector to be processed.
[0155] In some embodiments, the similarity determination unit includes:
[0156] A Euclidean distance determination subunit, configured to determine the Euclidean distance between the feature vector to be processed and each abnormal sound feature vector;
[0157] The similarity determination subunit is configured to determine the similarity between the to-be-processed audio data in the preset format and each abnormal sound audio data based on the Euclidean distance.
[0158] For the convenience of description, the above device is described by dividing it into various modules according to its functions. Of course, when implementing the present disclosure, the functions of each module can be implemented in the same or multiple software and / or hardware.
[0159] The device of the above embodiment is used to implement the corresponding vehicle abnormality identification method in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be repeated here.
[0160] Based on the same inventive concept, corresponding to any of the above-mentioned embodiments, the present disclosure also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method for identifying vehicle abnormalities described in any of the above embodiments is implemented.
[0161] Figure 4 A more specific schematic diagram of the hardware structure of an electronic device provided in this embodiment is shown, and the device may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040 are connected to each other through the bus 1050 in the device.
[0162] The processor 1010 can be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.
[0163] The memory 1020 may be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 may store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program codes are stored in the memory 1020 and are called and executed by the processor 1010.
[0164] The input / output interface 1030 is used to connect the input / output module to realize information input and output. The input / output module can be configured in the device as a component (not shown in the figure), or it can be externally connected to the device to provide corresponding functions. The input device may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output device may include a display, a speaker, a vibrator, an indicator light, etc.
[0165] The communication interface 1040 is used to connect a communication module (not shown in the figure) to realize communication interaction between the device and other devices. The communication module can realize communication through a wired mode (such as USB (Universal Serial Bus), network cable, etc.), or through a wireless mode (such as mobile network, WIFI (Wireless Fidelity), Bluetooth, etc.).
[0166] The bus 1050 includes a path that transmits information between the various components of the device (eg, the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040).
[0167] It should be noted that, although the above device only shows the processor 1010, the memory 1020, the input / output interface 1030, the communication interface 1040 and the bus 1050, in the specific implementation process, the device may also include other components necessary for normal operation. In addition, it can be understood by those skilled in the art that the above device may also only include the components necessary for implementing the embodiments of the present specification, and does not necessarily include all the components shown in the figure.
[0168] The electronic device of the above embodiment is used to implement the corresponding vehicle abnormality identification method in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be repeated here.
[0169] Based on the same inventive concept, corresponding to any of the above-mentioned embodiment methods, the present disclosure also provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the vehicle abnormality identification method as described in any of the above embodiments.
[0170] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device.
[0171] The computer instructions stored in the storage medium of the above embodiment are used to enable the computer to execute the vehicle abnormality identification method described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.
[0172] Based on the same inventive concept, corresponding to any of the above-mentioned embodiments and methods, the present application also provides a vehicle, including a vehicle abnormality identification device, or an electronic device, or a storage medium in the above-mentioned embodiments, and the vehicle equipment implements the vehicle abnormality identification method described in any of the above embodiments.
[0173] The vehicle of the above-mentioned embodiment is used to implement the vehicle abnormality identification method described in any of the above-mentioned embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.
[0174] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present disclosure is limited to these examples. Based on the concept of the present disclosure, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the different aspects of the embodiments of the present disclosure as described above, which are not provided in detail for the sake of simplicity.
[0175] In addition, to simplify the description and discussion, and in order not to make the embodiments of the present disclosure difficult to understand, the known power / ground connections to the integrated circuit (IC) chips and other components may or may not be shown in the provided figures. In addition, the device can be shown in the form of a block diagram to avoid making the embodiments of the present disclosure difficult to understand, and this also takes into account the fact that the details of the implementation of these block diagram devices are highly dependent on the platform on which the embodiments of the present disclosure will be implemented (that is, these details should be fully within the scope of understanding of those skilled in the art). Where specific details (e.g., circuits) are set forth to describe exemplary embodiments of the present disclosure, it is apparent to those skilled in the art that the embodiments of the present disclosure can be implemented without these specific details or with changes in these specific details. Therefore, these descriptions should be considered illustrative rather than restrictive.
[0176] Although the present disclosure has been described in conjunction with specific embodiments of the present disclosure, many replacements, modifications and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may use the embodiments discussed.
[0177] The embodiments of the present disclosure are intended to cover all such substitutions, modifications and variations that fall within the broad scope of the present disclosure. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present disclosure should be included in the scope of protection of the present disclosure.
Claims
1. A method for identifying vehicle abnormality, characterized in that: The method comprises: Determine an audio file to be processed, and process the audio file to be processed to obtain audio data to be processed in a preset format; wherein the audio data to be processed in the preset format is consistent with the format of a plurality of pre-stored abnormal sound audio data; The similarity between the audio data to be processed in the preset format and each abnormal sound audio data is determined by using the Euclidean distance, target abnormal sound audio data with high similarity is determined from the multiple abnormal sound audio data, and the cause of the vehicle abnormality is determined based on the target abnormal sound audio data.
2. The method according to claim 1, characterized in that The step of processing the audio file to be processed to obtain audio data to be processed in a preset format includes: Reading and processing the audio file to be processed to obtain the read audio data to be processed; The read audio data to be processed is normalized to obtain audio data to be processed in a preset format.
3. The method according to claim 2, characterized in that The step of reading and processing the audio file to be processed to obtain the read audio data to be processed includes: Determine a target decoder according to the file header information of the audio file to be processed, and use the target decoder to decode the audio file to be processed to obtain decoded audio data to be processed; The decoded audio data to be processed is converted to obtain a double-precision floating-point number, and the double-precision floating-point number is used as the read audio data to be processed.
4. The method according to claim 2, characterized in that: The step of normalizing the read audio data to be processed to obtain the audio data to be processed in a preset format includes: Determine the number of channels of the read audio data to be processed, and convert the read audio data to be processed into single-channel audio data to be processed according to the number of channels; Converting the single-channel audio data to be processed into audio data to be processed at a preset sampling rate; The to-be-processed audio data of the preset sampling rate is converted into to-be-processed audio data of a preset duration, and the to-be-processed audio data of the preset duration is used as the to-be-processed audio data of a preset format.
5. The method according to claim 1, characterized in that The method of using the Euclidean distance to determine the similarity between the audio data to be processed in the preset format and each abnormal sound audio data, and determining target abnormal sound audio data with high similarity from the multiple abnormal sound audio data, includes: Performing vector processing on the to-be-processed audio data in the preset format to obtain a to-be-processed feature vector, and performing vector processing on the multiple abnormal sound audio data to obtain multiple abnormal sound feature vectors; Determine the similarity between the audio data to be processed in the preset format and each abnormal sound audio data according to the feature vector to be processed and the multiple abnormal sound feature vectors by using the Euclidean distance; Target abnormal sound audio data having a high similarity to the audio data to be processed in the preset format is determined from the multiple abnormal sound audio data.
6. The method according to claim 5, characterized in that The vectorization processing of the to-be-processed audio data in the preset format to obtain the to-be-processed feature vector comprises: Determine the number of decomposition layers of the to-be-processed audio data in the preset format, and determine the number of decomposition times based on the number of decomposition layers; In response to determining that the decomposition times are twice, the audio data to be processed in the preset format is decomposed for the first time using a wavelet basis function to obtain a first decomposition coefficient, the first decomposition coefficient is decomposed for the second time using a wavelet basis function to obtain a second decomposition coefficient, and the second decomposition coefficient is used as a feature vector to be processed.
7. The method according to claim 5, characterized in that The using of the Euclidean distance to determine the similarity between the to-be-processed audio data in the preset format and each abnormal sound audio data according to the to-be-processed feature vector and the plurality of abnormal sound feature vectors includes: Determine the Euclidean distance between the feature vector to be processed and each abnormal sound feature vector; Based on the Euclidean distance, the similarity between the to-be-processed audio data in the preset format and each abnormal sound audio data is determined.
8. A vehicle abnormality recognition device, characterized in that: include: A processing module is configured to determine an audio file to be processed, and process the audio file to be processed to obtain audio data to be processed in a preset format; wherein the audio data to be processed in the preset format is consistent with the format of a plurality of pre-stored abnormal sound audio data; The determination module is configured to determine the similarity between the audio data to be processed in the preset format and each abnormal sound audio data, determine target abnormal sound audio data with high similarity from the multiple abnormal sound audio data, and determine the cause of the vehicle abnormality based on the target abnormal sound audio data.
9. An electronic device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and running on the processor, wherein when the processor executes the program, the method according to any one of claims 1 to 7 is implemented.
10. A vehicle, characterized in that: Includes the vehicle abnormality identification device according to claim 8 or the electronic device according to claim 9.