Vehicle anomaly detection method, vehicle, and storage medium

Abstract

The application discloses a vehicle anomaly detection method, a vehicle and a storage medium, and relates to the technical field of vehicle detection. The method comprises the following steps: acquiring original audio data, scene perception data and chassis state data of a target vehicle; performing data conversion processing on the original audio data, the scene perception data and the chassis state data to obtain a data conversion result; performing abnormal sound analysis processing on the data conversion result to obtain an abnormal sound analysis result, and performing noise evaluation processing on the data conversion result to obtain a noise evaluation result; performing verification processing on the abnormal sound analysis result and the noise evaluation result to obtain an anomaly detection result, wherein the anomaly detection result is used to represent abnormal state information of the target vehicle; and displaying the anomaly detection result in a graphical user interface of the target vehicle. The application solves the technical problem that, in the related art, vehicle noise and abnormal sound monitoring is insufficient, thereby affecting driving safety.

Description

Technical Field

[0001] This application relates to the field of vehicle inspection technology, and more specifically, to a vehicle anomaly detection method, a vehicle, and a storage medium. Background Technology

[0002] Effective monitoring of vehicle noise and abnormal sounds plays a crucial role in improving driving experience and road safety. Current technologies for vehicle noise and abnormal sound detection typically involve the vehicle maintenance or development phase, lacking the capability to monitor noise and abnormal sounds throughout the vehicle's entire lifecycle. This results in insufficient real-time assessment of the vehicle's health status, hindering timely warnings and maintenance for abnormal situations, thus impacting driving safety and the user's riding experience. Therefore, how to effectively monitor vehicle noise and abnormal sounds throughout its entire lifecycle is one of the important technical challenges in related fields.

[0003] There is currently no effective solution to the above problems. Summary of the Invention

[0004] This application provides a vehicle anomaly detection method, a vehicle, and a storage medium to at least solve the technical problem in the related art of insufficient monitoring of vehicle noise and abnormal sounds, which in turn affects driving safety.

[0005] According to one aspect of the embodiments of this application, a vehicle anomaly detection method is provided, comprising: acquiring raw audio data, context-aware data, and chassis state data of a target vehicle, wherein the raw audio data is used to represent audio data collected from multiple vehicle sound zones, the context-aware data is used to represent the fusion result of vehicle state data, vehicle network data, and driving environment data of the target vehicle, and the chassis state data is used to represent chassis perception data and suspension stiffness data of the target vehicle; performing data conversion processing on the raw audio data, context-aware data, and chassis state data to obtain a data conversion result, wherein the data conversion result is used to represent the raw audio data. The data conversion process involves: 1) Vectorizing the context-aware data and chassis status data; 2) Performing noise analysis on the converted data to obtain noise analysis results, and 3) Performing noise assessment on the converted data to obtain noise assessment results. The noise analysis results represent the noise correlation and noise region information of the target vehicle, while the noise assessment results represent the noise anomalies and sources of the target vehicle. The noise analysis and noise assessment results are then validated to obtain anomaly detection results, which represent the abnormal state information of the target vehicle. Finally, the anomaly detection results are displayed in the graphical user interface of the target vehicle.

[0006] Optionally, acquiring context-aware data includes: acquiring vehicle status data, vehicle-to-everything (V2X) data, and driving environment data. The vehicle status data includes at least the execution status data of multiple vehicle actuators within the target vehicle. The V2X data includes at least the meteorological data, geographic data, and preset database data associated with the target vehicle. The driving environment data includes at least the road condition data and passenger experience data of the target vehicle. The vehicle status data, V2X data, and driving environment data are then fused to obtain context-aware data.

[0007] Optionally, the data conversion results are subjected to abnormal noise analysis processing to obtain abnormal noise analysis results, including: filtering the data conversion results based on abnormal noise soundprint conditions to obtain abnormal noise filtering results, wherein the abnormal noise soundprint conditions are used to identify abnormal noise-related data in the data conversion results; and using a target abnormal noise localization model to predictively analyze the abnormal noise filtering results to obtain abnormal noise analysis results, wherein the target abnormal noise localization model is obtained by machine learning training an initial abnormal noise localization model using multiple sets of abnormal noise sample data, and the multiple sets of abnormal noise sample data include: abnormal noise input data and abnormal noise annotation data.

[0008] Optionally, noise assessment processing is performed on the data conversion results to obtain noise assessment results, including: filtering the data conversion results based on noise voiceprint conditions to obtain noise filtering results, wherein the noise voiceprint conditions are used to identify noise-related data in the data conversion results; and using a target noise assessment model to predict and analyze the noise filtering results to obtain noise assessment results, wherein the target noise assessment model is obtained by machine learning training of an initial noise assessment model using multiple sets of noise sample data, and the multiple sets of noise sample data include: noise input data and noise labeled data.

[0009] Optionally, the noise screening results are predicted and analyzed using the target noise assessment model to obtain the noise assessment results, including: evaluating the noise screening results based on the target noise assessment model to obtain the initial assessment results, wherein the initial assessment results are used to represent the noise distribution status information of the target vehicle; in response to the initial assessment results not meeting the preset noise range conditions, the initial assessment results are predicted and analyzed based on the target noise assessment model to obtain the noise assessment results.

[0010] Optionally, the abnormal noise analysis results and noise assessment results are validated to obtain anomaly detection results, including: cross-validating the abnormal noise analysis results and noise assessment results to obtain cross-validation results, wherein the cross-validation results are used to determine the abnormal location information and abnormal cause information of the target vehicle; and determining the anomaly detection results based on the cross-validation results.

[0011] Optionally, the vehicle anomaly detection method further includes: generating anomaly broadcast information based on the anomaly detection result, wherein the anomaly broadcast information is used to notify the target user of the abnormal status of the target vehicle; and playing the anomaly broadcast information using the target vehicle's audio component.

[0012] Optionally, the multiple vehicle sound zones include: the front cabin sound zone, the passenger cabin sound zone, and the rear cabin sound zone.

[0013] According to another aspect of the embodiments of this application, a vehicle anomaly detection device is also provided, comprising: an acquisition module, configured to acquire raw audio data, context perception data, and chassis state data of a target vehicle, wherein the raw audio data represents audio data collected from multiple vehicle sound zones, the context perception data represents the fusion result of vehicle state data, vehicle network data, and driving environment data of the target vehicle, and the chassis state data represents chassis perception data and suspension stiffness data of the target vehicle; and a conversion module, configured to perform data conversion processing on the raw audio data, context perception data, and chassis state data to obtain a data conversion result, wherein the data conversion result represents the raw audio data, context perception data, and chassis state data. The system comprises: a vectorized representation of the perceived data and chassis status data; a processing module for performing abnormal noise analysis on the data conversion results to obtain abnormal noise analysis results, and a noise assessment module for performing noise assessment on the data conversion results to obtain noise assessment results. The abnormal noise analysis results are used to represent the abnormal noise correlation and abnormal noise area information of the target vehicle, and the noise assessment results are used to represent the noise anomaly results and noise anomaly sources of the target vehicle; a verification module for verifying the abnormal noise analysis results and noise assessment results to obtain anomaly detection results. The anomaly detection results are used to represent the abnormal state information of the target vehicle; and a display module for displaying the anomaly detection results in the graphical user interface of the target vehicle.

[0014] Optionally, the acquisition module is also used to: acquire vehicle status data, vehicle network data, and driving environment data, wherein the vehicle status data includes at least: execution status data of multiple vehicle actuators in the target vehicle, the vehicle network data includes at least: meteorological data, geographical data, and preset database data associated with the target vehicle, and the driving environment data includes at least: road condition data and passenger experience data of the target vehicle; and to fuse the vehicle status data, vehicle network data, and driving environment data to obtain context perception data.

[0015] Optionally, the processing module is also used to: filter the data conversion results based on abnormal noise soundprint conditions to obtain abnormal noise filtering results, wherein the abnormal noise soundprint conditions are used to identify abnormal noise-related data in the data conversion results; and use the target abnormal noise location model to predict and analyze the abnormal noise filtering results to obtain abnormal noise analysis results, wherein the target abnormal noise location model is obtained by machine learning training of the initial abnormal noise location model using multiple sets of abnormal noise sample data, and the multiple sets of abnormal noise sample data include: abnormal noise input data and abnormal noise annotation data.

[0016] Optionally, the processing module is further configured to: filter the data conversion results based on noise voiceprint conditions to obtain noise filtering results, wherein the noise voiceprint conditions are used to identify noise-related data in the data conversion results; and perform predictive analysis on the noise filtering results using a target noise assessment model to obtain noise assessment results, wherein the target noise assessment model is obtained by machine learning training of an initial noise assessment model using multiple sets of noise sample data, and the multiple sets of noise sample data include: noise input data and noise labeled data.

[0017] Optionally, the processing module is further configured to: evaluate the noise screening results based on the target noise evaluation model to obtain an initial evaluation result, wherein the initial evaluation result is used to represent the noise distribution status information of the target vehicle; and, in response to the initial evaluation result not meeting the preset noise range condition, perform predictive analysis on the initial evaluation result based on the target noise evaluation model to obtain a noise evaluation result.

[0018] Optionally, the verification module is also used to: cross-validate the abnormal noise analysis results and the noise assessment results to obtain cross-validation results, wherein the cross-validation results are used to determine the abnormal location information and abnormal cause information of the target vehicle; and determine the abnormal detection results based on the cross-validation results.

[0019] Optionally, the vehicle anomaly detection device further includes: a generation module for generating anomaly broadcast information based on the anomaly detection result, wherein the anomaly broadcast information is used to notify the target user of the abnormal status of the target vehicle; and a playback module for playing the anomaly broadcast information using the target vehicle's audio components.

[0020] Optionally, the multiple vehicle sound zones include: the front cabin sound zone, the passenger cabin sound zone, and the rear cabin sound zone.

[0021] 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.

[0022] According to another aspect of the embodiments of this application, an electronic device 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.

[0023] 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.

[0024] 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.

[0025] 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.

[0026] 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.

[0027] In this embodiment, the original audio data, context-aware data, and chassis status data of the target vehicle are first acquired. The original audio data represents audio data collected from multiple vehicle sound zones. The context-aware data represents the fusion result of the target vehicle's vehicle status data, vehicle network data, and driving environment data. The chassis status data represents the target vehicle's chassis perception data and suspension stiffness data. Then, the original audio data, context-aware data, and chassis status data undergo data transformation processing to obtain a data transformation result, which represents the original audio data, context-aware data, and chassis status data. The corresponding vectorized representation results are then processed. Further, the data conversion results undergo abnormal noise analysis to obtain abnormal noise analysis results, and the data conversion results undergo noise assessment to obtain noise assessment results. The abnormal noise analysis results are used to represent the abnormal noise correlation and abnormal noise area information of the target vehicle, while the noise assessment results are used to represent the noise anomaly results and sources of the target vehicle. In addition, the abnormal noise analysis results and noise assessment results are verified to obtain anomaly detection results, which are used to represent the abnormal state information of the target vehicle. Finally, the anomaly detection results are displayed in the graphical user interface of the target vehicle. This application first achieves comprehensive monitoring of the vehicle's internal and external environment by acquiring the target vehicle's original audio data, context-aware data, and chassis status data. This ensures that multiple potential factors affecting vehicle noise and abnormal noises can be collected. Secondly, the original audio data, context-aware data, and chassis status data undergo data conversion processing, which improves the operability of the data and reduces the complexity of data analysis. Next, the data conversion results are processed for abnormal noise analysis to obtain abnormal noise analysis results, achieving precise location of abnormal noises. The data conversion results are also processed for noise assessment to obtain noise assessment results, achieving precise noise assessment. Furthermore, the abnormal noise analysis results and noise assessment results are verified to ensure the accuracy of abnormal noise location and noise assessment, avoiding false alarms or missed alarms and improving the reliability of the detection results. Finally, the graphical display of the abnormality detection results allows users to intuitively understand the vehicle's health status and obtain timely feedback on vehicle anomalies. This human-computer interaction method improves the user experience and promotes proactive management of the vehicle's status, helping to take measures before problems worsen and further ensuring driving safety. Therefore, this application can achieve the technical effect of effectively monitoring vehicle noise and abnormal noises, thereby solving the technical problem of insufficient monitoring of vehicle noise and abnormal noises in related technologies, which affects driving safety. Attached Figure Description

[0028] 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: Figure 1 This is a flowchart of a vehicle anomaly detection method according to an embodiment of this application; Figure 2 This is a system architecture diagram for acquiring context-aware data according to an embodiment of this application; Figure 3 This is a system architecture diagram for acquiring raw audio data according to an embodiment of this application; Figure 4 This is an example diagram illustrating the acquisition of raw audio data according to an embodiment of this application; Figure 5 This is a structural block diagram of a vehicle anomaly detection system according to an embodiment of this application; Figure 6 This is a structural block diagram of a vehicle anomaly detection device according to an embodiment of this application. Detailed Implementation

[0029] 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.

[0030] 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.

[0031] According to an embodiment of this application, a method embodiment for vehicle anomaly detection 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. Furthermore, 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.

[0032] This embodiment provides a method for detecting vehicle anomalies. Figure 1 This is a flowchart of a vehicle anomaly detection method according to an embodiment of this application, such as... Figure 1 As shown, the method includes the following steps: Step S11: Obtain the original audio data, context perception data, and chassis status data of the target vehicle. The original audio data represents audio data collected from multiple vehicle sound zones. The context perception data represents the fusion result of the target vehicle's vehicle status data, vehicle network data, and driving environment data. The chassis status data represents the target vehicle's chassis perception data and suspension stiffness data.

[0033] The aforementioned raw audio data refers to sound signals from multiple vehicle areas collected from multiple vehicle sound zones.

[0034] Optionally, the raw audio data includes audio data from vehicle parts such as the engine compartment, cabin, and trunk.

[0035] Optionally, the target vehicle is equipped with a multi-zone microphone array. The microphone array is arranged in a specific way to capture sound of different frequencies and directions, providing rich and detailed audio information for subsequent analysis.

[0036] The aforementioned context-aware data refers to data reflecting the vehicle's operating environment and status obtained by fusing vehicle status data, vehicle network data, and driving environment data of the target vehicle.

[0037] Optionally, vehicle status data includes, but is not limited to, vehicle speed, acceleration, engine speed, etc.; vehicle network data covers weather information, road conditions, traffic signals, etc.; and driving environment data involves natural environmental parameters such as temperature, humidity, and wind speed.

[0038] In one optional embodiment, vehicle status data is collected in real time using multiple onboard sensors, vehicle-to-everything (V2X) data is acquired via an onboard network, and driving environment data is collected using an onboard weather station. The vehicle status data, V2X data, and driving environment data are then integrated to obtain context-awareness data.

[0039] The aforementioned chassis status data includes chassis sensing data and suspension stiffness data. Chassis sensing data refers to real-time monitoring data of vibration, wear, temperature, etc., of various chassis components, while suspension stiffness data reflects the dynamic stiffness characteristics of the suspension system under different loads and driving conditions, and is an important factor affecting vehicle noise and abnormal sounds.

[0040] Optionally, the chassis's operating status can be monitored in real time using sensors installed at key locations on the vehicle chassis, such as accelerometers, vibration sensors, and temperature sensors. Simultaneously, suspension stiffness data can be calculated by measuring the vehicle's dynamic response under different road conditions.

[0041] Step S12: Perform data transformation processing on the original audio data, context-aware data, and chassis status data to obtain data transformation results. The data transformation results are used to represent the vectorized representation results corresponding to the original audio data, context-aware data, and chassis status data.

[0042] In one alternative embodiment, the original audio data is first converted into a digital audio signal. Then, the digital audio signal is divided into a series of short frames using a sliding window method. Each frame represents audio information within a certain time period. Furthermore, the audio signal within each frame is converted into a corresponding spectral representation (e.g., frequency domain features are extracted through Fourier transform). Finally, the spectral representation is converted into an audio feature vector.

[0043] In one optional embodiment, the vehicle status data, vehicle network data, and driving environment data in the context perception data are converted into corresponding numerical representations, and the multiple numerical representations are concatenated to obtain the context perception feature vector corresponding to the context perception data.

[0044] In one optional embodiment, the chassis status data is quantified to obtain the chassis status feature vector corresponding to the chassis status data.

[0045] Audio feature vectors, context-aware feature vectors, and chassis status feature vectors are used as the data transformation results.

[0046] Step S13: Perform abnormal noise analysis on the data conversion results to obtain abnormal noise analysis results, and perform noise assessment on the data conversion results to obtain noise assessment results. The abnormal noise analysis results are used to represent the abnormal noise correlation and abnormal noise area information of the target vehicle, and the noise assessment results are used to represent the noise anomaly results and noise anomaly sources of the target vehicle.

[0047] The above-mentioned abnormal noise correlation refers to the relationship between abnormal noise and vehicle dynamic behavior, vehicle status, and external environment.

[0048] The above information on abnormal noise areas refers to the location information of the area where the abnormal noise occurs.

[0049] The above noise anomaly results are used to characterize whether the noise level of the target vehicle exceeds the normal range.

[0050] The aforementioned sources of noise anomalies refer to the sources of noise anomalies that have been located.

[0051] In one optional embodiment, a pre-trained abnormal noise localization neural network model is used to analyze the abnormal noise in the data conversion results, identifying the abnormal noise features in the data conversion results. Simultaneously, the correlation between the abnormal noise features and vehicle dynamic behavior, vehicle state, and external environment is further analyzed, and the area where the abnormal noise occurs is determined, thus obtaining the abnormal noise analysis results.

[0052] In one optional embodiment, a pre-trained noise assessment neural network model is used to analyze the data transformation results to determine whether the noise is abnormal and the possible sources of the abnormal noise. Optionally, the noise typically refers to the continuous sound generated by the vehicle during operation, such as engine noise, tire friction noise, etc.

[0053] Step S14: Verify the abnormal noise analysis results and noise assessment results to obtain the anomaly detection results, which are used to represent the abnormal state information of the target vehicle.

[0054] The above abnormal status information refers to the descriptive information of any abnormal conditions existing in the target vehicle.

[0055] Optionally, the abnormal noise analysis results and noise assessment results can be cross-validated, using the abnormal noise analysis results to verify the noise assessment results, and vice versa, thereby ensuring the accuracy of the abnormal noise analysis results and noise assessment results.

[0056] After verifying the abnormal noise analysis results and noise assessment results, the verified abnormal noise analysis results and noise assessment results are integrated to obtain the anomaly detection results.

[0057] Step S15: Display the anomaly detection results in the graphical user interface of the target vehicle.

[0058] The graphical user interface (GUI) mentioned above refers to an interface through which users interact with the software using graphical elements (such as icons, menus, and windows). A GUI allows users to intuitively control and operate the software through actions such as clicking and dragging.

[0059] In one alternative embodiment, a data visualization tool or library is used to convert the anomaly detection results into an easily understandable chart format, and the converted chart is displayed on the target vehicle's graphical user interface. For example, a heatmap can be used to display noise and abnormal noise levels in different areas, with the intensity of the color representing the severity of the anomaly.

[0060] Steps S11 to S15 above, by acquiring the target vehicle's original audio data, context-aware data, and chassis status data, achieve comprehensive monitoring of the vehicle's internal and external environment. This ensures that multiple potential factors affecting vehicle noise and abnormal sounds can be collected. Secondly, the original audio data, context-aware data, and chassis status data undergo data transformation processing, which improves data operability and reduces the complexity of data analysis. Next, the data transformation results are processed for abnormal sound analysis, achieving precise location of abnormal sounds. The data transformation results are also processed for noise assessment, achieving accurate noise evaluation. Furthermore, the abnormal sound analysis results and noise assessment results are verified to ensure the accuracy of abnormal sound location and noise assessment, avoiding false alarms or missed alarms and improving the reliability of the detection results. Finally, the graphical display of anomaly detection results allows users to intuitively understand the vehicle's health status and obtain timely feedback on vehicle anomalies. This human-computer interaction method enhances the user experience and promotes proactive management of the vehicle's status, helping to take measures before problems worsen and further ensuring driving safety. Therefore, this application achieves the technical effect of effectively monitoring vehicle noise and abnormal sounds, thereby solving the technical problem of insufficient monitoring of vehicle noise and abnormal sounds in related technologies, which in turn affects driving safety.

[0061] The vehicle anomaly detection method in the embodiments of this application will be further described below.

[0062] Optionally, in step S11, acquiring context-aware data includes: acquiring vehicle status data, vehicle network data, and driving environment data, wherein the vehicle status data includes at least: execution status data of multiple vehicle actuators in the target vehicle, the vehicle network data includes at least: meteorological data, geographical data, and preset database data associated with the target vehicle, and the driving environment data includes at least: road condition data and passenger experience data of the target vehicle; the vehicle status data, vehicle network data, and driving environment data are fused to obtain context-aware data.

[0063] The aforementioned vehicle status data refers to the physical status data of the target vehicle when it is in motion or stationary, including but not limited to engine speed, vehicle speed, tire pressure, brake pressure, suspension status, air conditioning system operating status, door and window open / closed status, engine temperature, battery voltage, etc.

[0064] The aforementioned actuators refer to components that can perform physical operations based on electronic signals, such as engine control units, braking systems, power steering systems, and air conditioning compressors.

[0065] The execution status data of the aforementioned vehicle actuators includes the actuator's working mode, frequency, force, etc.

[0066] Optionally, vehicle status data can be collected in real time using a variety of onboard sensors.

[0067] The aforementioned vehicle-to-everything (V2X) data refers to the data that vehicles exchange with the outside world via the internet, including weather, road conditions, traffic information, and vehicle historical data, which can provide comprehensive information about the vehicle's environment.

[0068] The aforementioned preset database data refers to reference database data provided by car manufacturers or third parties, which covers common vehicle faults, maintenance history, optimization suggestions, etc., and helps to diagnose and predict vehicle status.

[0069] In one alternative embodiment, vehicle network data is acquired via an in-vehicle network.

[0070] The aforementioned road condition data includes road surface smoothness, slope, and slipperiness.

[0071] The above-mentioned riding experience data reflects passengers' perception of comfort during the vehicle's journey, and can be indirectly obtained by capturing passenger facial expressions and voice feedback through cameras.

[0072] In one alternative embodiment, road surface sensors or cameras located on the underside of the vehicle are used for image processing to identify the type and condition of the road, such as urban roads, highways, rural dirt roads, or slippery surfaces. Occupant reactions are collected via in-cabin cameras and a voice recognition system to assess ride comfort, such as whether occupants have voice feedback complaining about excessive noise.

[0073] Vehicle status data, vehicle network data, and driving environment data are merged to obtain context perception data.

[0074] Through the above steps, a comprehensive perception of the vehicle's internal and external environment is achieved, forming a comprehensive context perception database.

[0075] Figure 2 This is a system architecture diagram for acquiring context-aware data according to an embodiment of this application, such as... Figure 2 As shown, the system architecture for acquiring context-aware data includes a vision module, a connectivity module, a hardware control module, and a data fusion module. The vision module collects driving environment data, the hardware control module collects vehicle status data, and the connectivity module collects vehicle-to-everything (V2X) data. The data fusion module merges the vehicle status data, V2X data, and driving environment data to obtain the context-aware data.

[0076] Optionally, in step S13, the abnormal noise analysis process is performed on the data conversion result to obtain the abnormal noise analysis result, including the following steps: Step S131: The data conversion results are filtered based on the abnormal noise soundprint conditions to obtain abnormal noise filtering results. The abnormal noise soundprint conditions are used to identify abnormal noise-related data in the data conversion results. Step S132: Use the target abnormal noise location model to predict and analyze the abnormal noise screening results to obtain the abnormal noise analysis results. The target abnormal noise location model is obtained by machine learning training the initial abnormal noise location model with multiple sets of abnormal noise sample data. The multiple sets of abnormal noise sample data include: abnormal noise input data and abnormal noise annotation data.

[0077] The aforementioned abnormal noise soundprint conditions include the frequency distribution, temporal characteristics (such as suddenness and intermittency), and sound pressure level of the abnormal noise. These abnormal noise soundprint conditions are used to accurately locate and identify abnormal noises in vehicles from complex audio data.

[0078] Optionally, abnormal noise features can be extracted from the data conversion results to identify features helpful for abnormal noise identification, such as instantaneous energy and zero-crossing rate. Further, the extracted features are filtered according to predefined abnormal noise soundprint conditions (e.g., by threshold filtering). For example, an energy threshold is set; any sound segment with instantaneous energy exceeding this threshold is identified as a potential abnormal noise, thus filtering out abnormal noise-related data from the data conversion results. After the above filtering process, the data segments that meet the abnormal noise soundprint conditions are retained; these are considered data related to abnormal noise, i.e., abnormal noise-related data.

[0079] Optionally, the aforementioned target abnormal noise localization model is a machine learning model pre-trained based on multiple sets of abnormal noise sample data. The multiple sets of abnormal noise sample data include: abnormal noise input data and abnormal noise annotation data. Abnormal noise input data refers to the audio recording of the abnormal noise, while the abnormal noise annotation data includes at least the specific location and cause of the abnormal noise.

[0080] The abnormal noise screening results are input into the trained target abnormal noise localization model. The model uses its learned knowledge to analyze the input abnormal noise screening results and predict the specific location and cause of the abnormal noise.

[0081] Steps S131 to S132 above apply the abnormal noise soundprint conditions to the screening of data conversion results, and use the target abnormal noise localization model to predict and analyze the abnormal noise screening results, so as to achieve accurate identification and localization of vehicle abnormal noise in complex environments.

[0082] Optionally, in step S13, noise evaluation processing is performed on the data conversion result to obtain the noise evaluation result, including the following steps: Step S133: The data conversion results are filtered based on the noise voiceprint conditions to obtain noise filtering results, wherein the noise voiceprint conditions are used to identify noise-related data in the data conversion results. Step S134: The noise screening results are predicted and analyzed using the target noise assessment model to obtain the noise assessment results. The target noise assessment model is obtained by machine learning training of the initial noise assessment model using multiple sets of noise sample data. The multiple sets of noise sample data include noise input data and noise labeling data.

[0083] The aforementioned noise acoustic signature conditions refer to conditions set based on the frequency distribution, time series changes, and spectral characteristics of the noise.

[0084] In one alternative embodiment, the data conversion results are filtered based on noise acoustic characteristics, retaining data that matches the noise features, i.e., noise-related data.

[0085] Optionally, the aforementioned target noise assessment model is a machine learning model pre-trained based on multiple sets of noise sample data. The multiple sets of noise sample data include: noise input data and noise labeled data. Noise input data refers to the actual collected noise signal, and noise labeled data includes at least the noise distribution and its causes.

[0086] Optionally, the noise screening results can be input into the target noise assessment model, which can perform predictive analysis on the noise screening results to obtain the noise assessment results.

[0087] Steps S133 to S134 above apply noise soundprint conditions to the screening of data conversion results, and use the target noise assessment model to predict and analyze the noise screening results, thereby achieving accurate identification and assessment of vehicle internal and external noise.

[0088] Optionally, the noise screening results can be predicted and analyzed using a target noise assessment model to obtain the noise assessment results, including the following steps: Step S1341: The noise screening results are evaluated based on the target noise evaluation model to obtain the initial evaluation results, wherein the initial evaluation results are used to represent the noise distribution status information of the target vehicle. Step S1342: In response to the initial evaluation result not meeting the preset noise range condition, predictive analysis is performed on the initial evaluation result based on the target noise evaluation model to obtain the noise evaluation result.

[0089] In one optional embodiment, the noise screening results are input into the target noise evaluation model, which outputs an initial evaluation result. This result represents the noise distribution status information of the target vehicle in the current state in numerical or grade form, including but not limited to the noise level of each sound zone and the frequency characteristics of the noise.

[0090] The aforementioned preset noise range conditions are used to determine whether the noise of the target vehicle is within an acceptable range. Optionally, different noise standards can be set according to different vehicle types and usage scenarios. Under normal operating conditions, the noise levels in each sound zone inside the vehicle should be lower than a specific value.

[0091] If the initial assessment results do not meet the preset noise range conditions, meaning the noise level of the target vehicle exceeds the acceptable range, further analysis is performed on the initial assessment results based on the target noise assessment model. Optionally, the target noise assessment model conducts an in-depth analysis of the noise's spectral characteristics, temporal characteristics, and their correlation with vehicle dynamic data to identify the causes of excessive noise, whether it is related to vehicle malfunctions, component aging, or changes in the external environment. Finally, it outputs a detailed assessment and recommendations on the vehicle's noise status, i.e., the noise assessment result.

[0092] Steps S1341 to S1342 above utilize the target noise assessment model to predict and analyze the noise screening results, thereby obtaining the noise assessment results and achieving accurate assessment of vehicle noise. This helps to provide early warning of potential vehicle hazards and ensure driving safety.

[0093] Optionally, in step S14, the abnormal noise analysis results and noise assessment results are verified to obtain the anomaly detection results, including the following steps: Step S141: Cross-validate the abnormal noise analysis results and the noise assessment results to obtain cross-validation results. The cross-validation results are used to determine the abnormal location information and abnormal cause information of the target vehicle. Step S142: Determine the anomaly detection results based on the cross-validation results.

[0094] The aforementioned cross-validation refers to the mutual verification between the abnormal noise analysis results and the noise assessment results to confirm their consistency and reliability.

[0095] Optionally, cross-validation can be performed on the abnormal noise analysis results and the noise assessment results. For example, the timestamps and frequency characteristics in the abnormal noise analysis results are matched with the noise levels at the same time points in the noise assessment results. If, at a specific time point, the abnormal noise analysis results indicate the occurrence of abnormal noise, while the noise assessment results show a significant increase in the noise level in the area at the same time point, then there is a correlation between the two results. Cross-validation ensures the consistency between the abnormal noise analysis results and the noise assessment results, avoiding potential misjudgments that might occur with a single model.

[0096] The above anomaly detection results include information on the location and cause of the anomaly, which are used to guide subsequent vehicle maintenance and noise optimization.

[0097] Optionally, the abnormal noise analysis results and noise assessment results confirmed after cross-validation are fused to obtain the anomaly detection results.

[0098] Steps S141 to S142 above cross-validate the abnormal noise analysis results and noise assessment results to obtain cross-validation results, and determine the abnormal detection results based on the cross-validation results, thereby realizing a comprehensive analysis and accurate diagnosis of abnormal vehicle sounds.

[0099] Optionally, the vehicle anomaly detection method may also include the following steps: Step S161: Generate anomaly broadcast information based on the anomaly detection results, wherein the anomaly broadcast information is used to notify the target user of the abnormal status of the target vehicle. Step S162: Play an abnormal announcement message using the target vehicle's audio component.

[0100] The above-mentioned abnormality broadcast information is used to clearly and accurately convey the abnormal status of the vehicle and its possible impact to the vehicle's target users.

[0101] Optionally, anomaly reporting information is generated based on the anomaly detection results. The anomaly reporting information includes the specific location of the anomaly, possible causes, recommended operating procedures, and urgency level.

[0102] The aforementioned audio components refer to the audio system or other sound output devices equipped in the vehicle, such as speakers and buzzers. They are responsible for converting abnormal broadcast information into sound signals to notify the driver or passengers in the form of sound.

[0103] Once the anomaly alert is ready, select an appropriate audio component to broadcast the message. For example, when driving at high speed, the message can be loudly announced to the driver, while when the vehicle is stationary or traveling at low speed, a gentler voice output can be used to avoid startling the driver. Furthermore, to improve the effectiveness of the message, a mechanism can be set up to repeatedly play the anomaly alert until the user confirms receipt.

[0104] Steps S161 to S162 above generate abnormal broadcast information based on the abnormality detection results, and play the abnormal broadcast information using the audio component of the target vehicle, thereby realizing timely notification of abnormal vehicle status and ensuring driving safety.

[0105] Optionally, the multiple vehicle sound zones include: the front cabin sound zone, the passenger cabin sound zone, and the rear cabin sound zone.

[0106] Optionally, the front cabin noise zone refers to the noise area covered by the area under the hood and in front of the driver's cabin. The front cabin noise zone is where the vehicle's powertrain, transmission, and some electronic control systems are located, making it one of the main sources of abnormal noises and squeaks.

[0107] Optionally, the cabin sound zone covers the sound zone around the passenger seat, which is the area where passengers directly contact and perceive vehicle noise and abnormal sounds, and is also a key area for multi-zone intelligent language communication.

[0108] Optionally, the rear compartment noise zone covers the rear of the vehicle, including the luggage compartment and rear axle area, and is a key noise zone for monitoring abnormal noises from the rear mechanical structure and external noise.

[0109] Figure 3 This is a system structure diagram for acquiring raw audio data according to an embodiment of this application, such as... Figure 3 As shown, the target vehicle is divided into a front compartment, a passenger compartment (including seats), and a rear compartment. Each compartment is equipped with a microphone to collect sound signals from the corresponding compartment.

[0110] Figure 4 This is an example diagram illustrating the acquisition of raw audio data according to an embodiment of this application. The sound signals collected by the front cabin microphone, cockpit microphone, and rear cabin microphone are shown below. Figure 4 As shown.

[0111] By arranging microphone arrays in multiple vehicle sound zones, high-precision detection of abnormal noises and sounds throughout the entire vehicle can be achieved.

[0112] 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.

[0113] Figure 5 This is a structural block diagram of a vehicle anomaly detection system according to an embodiment of this application, such as... Figure 5 As shown, the vehicle anomaly detection system includes a multi-zone sound pickup system, an AI (Artificial Intelligence) cockpit system, a signal processing module, an AI chassis system, an abnormal noise location model, a noise assessment model, a comprehensive diagnostic model, an interactive system, and an enterprise database.

[0114] The aforementioned system utilizes the vehicle's built-in multi-zone microphones, vehicle computing unit, AI cockpit system, and AI chassis system to link the abnormal noises and sounds of the multi-zone intelligent vehicle with the vehicle's dynamic data and actuator data. By establishing the relationship between abnormal noises and sounds and the vehicle's status and external environment, a unified assessment result of the vehicle's dynamic and static abnormal noise and sound conditions is generated.

[0115] Specifically, the multi-zone microphone system mainly consists of in-cabin and out-of-cabin microphone systems. The microphones are arranged in a specific way and each can operate independently. The number of microphones in the vehicle can be increased or decreased depending on the vehicle's configuration level, and the number of microphones can be flexibly arranged according to the value of the components in each area.

[0116] To simplify multi-zone sound pickup systems, the entire vehicle is typically divided into three zones: the front compartment, the passenger compartment, and the rear compartment. These three zones work together to acquire noise and vibration data from the vehicle's surroundings and interior. The front compartment is primarily used to collect noise and vibration data from the front of the vehicle, and its layout can be based on the concentration of value or noise within the front compartment. Value concentration refers to the total value of vehicle components per unit area, while noise concentration refers to the area with the highest noise level under normal conditions.

[0117] The cockpit area primarily provides users with multi-zone intelligent voice interaction and collects noise and abnormal noise data from the passenger area. Since the cockpit area mainly serves the passengers, its layout is based on passenger voice functionality. Typically, the passenger cabin sound zone is further subdivided to provide more independent voice functions. Moreover, since passengers are the direct recipients of noise and abnormal noise, increasing the layout density of the cockpit area helps improve the accuracy of noise and abnormal noise detection. Optionally, the cockpit area can be subdivided into four sub-sound zones, which can be flexibly adjusted according to the vehicle configuration.

[0118] The rear compartment primarily collects noise and rattle data from the rear of the vehicle and the trunk area. Similar to the front compartment, its layout can be based on either noise concentration or value concentration principles. Sound collection from both the front and rear compartments plays a crucial role in determining the direction of noise and rattles. The multi-zone sound pickup system provides the entire system with raw noise and rattle data, which is then input to the signal processing module.

[0119] The AI ​​cockpit system primarily provides the fusion of vehicle status data, vehicle network data, and driving environment data, offering integrated data to correlate vehicle noise and vibration detection with vehicle component status, road conditions, and the natural environment. The AI ​​cockpit system comprises four functional modules: a vision module, a connectivity module, a hardware control module, and a data fusion module. The fused data is then fed into the signal processing module.

[0120] The vision module acquires data such as road type, speed bumps, and passenger comfort through the vehicle's built-in vision cameras. Road type and speed bumps can affect the types of abnormal noises and rattles, and are extremely helpful in locating the sources of these noises and rattles.

[0121] The connectivity module acquires data from the network through the vehicle's connectivity capabilities. This data is not limited to meteorological data, geographical data, and enterprise database data. By acquiring natural data about the vehicle's current location and combining it with the database data established for the entire vehicle, the detection accuracy of noise and abnormal sounds is further improved.

[0122] The hardware control module primarily provides vehicle actuator status data, which is used to analyze the relationship between vehicle actuators and noise or abnormal sounds. Vehicle actuators include domain controllers, actuators, sensors, etc. This module can sense the real-time status of the vehicle, including but not limited to vehicle speed, tire pressure, air conditioning operation status, door and window status, and wiper status.

[0123] Furthermore, to better detect the sources of vehicle noise and abnormal sounds, vehicle actuators are further divided into two categories: resident actuators and non-resident actuators. Resident actuators are those that will operate by default whether the vehicle is in static or dynamic operation; conversely, non-resident actuators need to feed back their execution status data to the data fusion module.

[0124] The data fusion module integrates vehicle status data, vehicle network data, and driving environment data, and then vectorizes them uniformly.

[0125] The signal processing module standardizes the raw audio data, context-aware data, and chassis status data. The current system contains three AI models: an abnormal noise localization model, a noise assessment model, and a comprehensive diagnostic model. The signal processing module needs to standardize the raw audio data, context-aware data, and chassis status data into input data for these three models.

[0126] The AI ​​chassis system is a crucial system for perceiving vehicle dynamic behavior. It dynamically senses vibration data from the road surface, as well as the vehicle's power and acceleration. The chassis's stiffness is also a significant factor influencing noise and abnormal sounds. Therefore, the AI ​​chassis system transmits the dynamic perception data and suspension stiffness data to the signal processing module.

[0127] Furthermore, based on the acoustic characteristics of abnormal noises and audible noises, the system processes them separately, employing two separate AI models to improve the accuracy of detection and assessment. Abnormal noises are characterized by their short duration, discontinuous acoustic patterns, and strong correlation with chassis vibration, abnormal component fit, or malfunctions; while noises are continuous, fluctuate relatively smoothly, last for a long time, and are highly correlated with the vehicle's external environment.

[0128] Based on the propagation characteristics of sound, the sounds received by all microphones should have time differences and amplitude variations, but the frequency differences should be small. This characteristic can be used to train anomaly localization models and noise assessment models.

[0129] Optionally, the abnormal noise localization model captures abnormal noise data from multi-zone sound data, determining the approximate area of ​​the abnormal noise based on its loudness (amplitude) and occurrence time. Furthermore, the model further explores the correlation between the abnormal noise's occurrence and chassis vibration and vehicle actuator status, narrowing down the noise's source range. Its output includes the correlation between actuators and the abnormal noise, as well as information on the noise's location; this data is then further imported into a comprehensive diagnostic model for further evaluation and judgment.

[0130] Optionally, the noise assessment model primarily evaluates long-term sound data, capturing the noise distribution within the vehicle under various vehicle conditions. Companies can use this model to establish the noise distribution of their vehicles under different conditions. Furthermore, the noise assessment model compares the data with the company's database to determine if the noise level is within acceptable limits. If the cabin noise level significantly exceeds the normal range, the model will further investigate the noise sources.

[0131] Optionally, the integrated diagnostic model comprehensively evaluates the results of the abnormal noise location model and the noise assessment model, and outputs the final checklist and noise assessment results.

[0132] Furthermore, the comprehensive diagnostic model's output data will be presented to the user through an interactive interface. The interaction methods are not limited to screen display or voice prompts.

[0133] According to another aspect of the embodiments of this application, a vehicle anomaly detection device is also provided. Figure 6 This is a structural block diagram of a vehicle anomaly detection device according to an embodiment of this application, such as... Figure 6 As shown, the vehicle anomaly detection device 600 includes: The acquisition module 601 is used to acquire the original audio data, context perception data and chassis status data of the target vehicle. The original audio data is used to represent audio data collected from multiple vehicle sound zones. The context perception data is used to represent the fusion result of the target vehicle's vehicle status data, vehicle network data and driving environment data. The chassis status data is used to represent the target vehicle's chassis perception data and suspension stiffness data. The conversion module 602 is used to perform data conversion processing on the original audio data, context-aware data and chassis status data to obtain data conversion results. The data conversion results are used to represent the vectorized representation results corresponding to the original audio data, context-aware data and chassis status data. The processing module 603 is used to perform abnormal noise analysis processing on the data conversion results to obtain abnormal noise analysis results, and to perform noise assessment processing on the data conversion results to obtain noise assessment results. The abnormal noise analysis results are used to represent the abnormal noise correlation and abnormal noise area information of the target vehicle, and the noise assessment results are used to represent the noise anomaly results and noise anomaly sources of the target vehicle. The verification module 604 is used to verify the abnormal noise analysis results and noise assessment results to obtain anomaly detection results, wherein the anomaly detection results are used to represent the abnormal state information of the target vehicle. Display module 605 is used to display the anomaly detection results in the graphical user interface of the target vehicle.

[0134] Optionally, the acquisition module 601 is further configured to: acquire vehicle status data, vehicle network data, and driving environment data, wherein the vehicle status data includes at least: execution status data of multiple vehicle actuators in the target vehicle, the vehicle network data includes at least: meteorological data, geographical data, and preset database data associated with the target vehicle, and the driving environment data includes at least: road condition data and passenger experience data of the target vehicle; and perform fusion processing on the vehicle status data, vehicle network data, and driving environment data to obtain context perception data.

[0135] Optionally, the processing module 603 is further configured to: filter the data conversion results based on the abnormal noise soundprint conditions to obtain abnormal noise filtering results, wherein the abnormal noise soundprint conditions are used to identify abnormal noise-related data in the data conversion results; and use the target abnormal noise location model to predict and analyze the abnormal noise filtering results to obtain abnormal noise analysis results, wherein the target abnormal noise location model is obtained by machine learning training of the initial abnormal noise location model using multiple sets of abnormal noise sample data, and the multiple sets of abnormal noise sample data include: abnormal noise input data and abnormal noise annotation data.

[0136] Optionally, the processing module 603 is further configured to: filter the data conversion results based on noise soundprint conditions to obtain noise filtering results, wherein the noise soundprint conditions are used to identify noise-related data in the data conversion results; and perform predictive analysis on the noise filtering results using a target noise assessment model to obtain noise assessment results, wherein the target noise assessment model is obtained by machine learning training of an initial noise assessment model using multiple sets of noise sample data, and the multiple sets of noise sample data include: noise input data and noise labeled data.

[0137] Optionally, the processing module 603 is further configured to: evaluate the noise screening results based on the target noise evaluation model to obtain an initial evaluation result, wherein the initial evaluation result is used to represent the noise distribution status information of the target vehicle; and in response to the initial evaluation result not meeting the preset noise range condition, perform predictive analysis on the initial evaluation result based on the target noise evaluation model to obtain a noise evaluation result.

[0138] Optionally, the verification module 604 is further configured to: cross-validate the abnormal noise analysis results and the noise assessment results to obtain cross-validation results, wherein the cross-validation results are used to determine the abnormal location information and abnormal cause information of the target vehicle; and determine the abnormal detection results based on the cross-validation results.

[0139] Optionally, the vehicle anomaly detection device further includes: a generation module for generating anomaly broadcast information based on the anomaly detection result, wherein the anomaly broadcast information is used to notify the target user of the abnormal status of the target vehicle; and a playback module for playing the anomaly broadcast information using the target vehicle's audio components.

[0140] Optionally, the multiple vehicle sound zones include: the front cabin sound zone, the passenger cabin sound zone, and the rear cabin sound zone.

[0141] 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.

[0142] 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.

[0143] According to another aspect of the embodiments of this application, a computer program product is provided, including a computer program that, when executed by a processor, implements the methods of various embodiments of this application.

[0144] 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 for storing a computer program, which, when executed by a processor, implements the methods in various embodiments of this application.

[0145] 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 described in the various embodiments of this application.

[0146] 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.

[0147] 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.

[0148] 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.

[0149] 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.

[0150] 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.

[0151] 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 detecting vehicle anomalies, characterized in that, include: The system acquires raw audio data, context-aware data, and chassis status data of the target vehicle. The raw audio data represents audio data collected from multiple vehicle sound zones. The context-aware data represents the fusion result of the target vehicle's vehicle status data, vehicle network data, and driving environment data. The chassis status data represents the target vehicle's chassis perception data and suspension stiffness data. The original audio data, the context-aware data, and the chassis status data are subjected to data transformation processing to obtain data transformation results, wherein the data transformation results are used to represent the vectorized representation results corresponding to the original audio data, the context-aware data, and the chassis status data; The data conversion results are filtered based on abnormal noise soundprint conditions to obtain abnormal noise filtering results, wherein the abnormal noise soundprint conditions are used to identify abnormal noise related data in the data conversion results; the abnormal noise filtering results are predicted and analyzed using a target abnormal noise localization model to obtain abnormal noise analysis results, wherein the abnormal noise analysis results are used to represent the abnormal noise correlation and abnormal noise area information of the target vehicle. The data conversion results are filtered based on noise acoustic print conditions to obtain noise filtering results, wherein the noise acoustic print conditions are used to identify noise-related data in the data conversion results; the noise filtering results are then predicted and analyzed using a target noise assessment model to obtain noise assessment results, wherein the noise assessment results are used to represent the noise anomaly results and noise anomaly sources of the target vehicle; The abnormal noise analysis results and the noise assessment results were cross-validated to obtain cross-validation results. Anomaly detection results are determined based on the cross-validation results, wherein the anomaly detection results are used to represent the abnormal state information of the target vehicle; The anomaly detection results are displayed in the graphical user interface of the target vehicle.

2. The vehicle anomaly detection method according to claim 1, characterized in that, Acquiring the context-aware data includes: The vehicle status data, the vehicle network data, and the driving environment data are acquired. The vehicle status data includes at least the execution status data of multiple vehicle actuators in the target vehicle. The vehicle network data includes at least the meteorological data, geographical data, and preset database data associated with the target vehicle. The driving environment data includes at least the road condition data and passenger experience data of the target vehicle. The vehicle status data, the vehicle network data, and the driving environment data are fused together to obtain the scenario perception data.

3. The vehicle anomaly detection method according to claim 1, characterized in that, The target abnormal noise localization model is obtained by machine learning training an initial abnormal noise localization model using multiple sets of abnormal noise sample data. The multiple sets of abnormal noise sample data include: abnormal noise input data and abnormal noise annotation data.

4. The vehicle anomaly detection method according to claim 1, characterized in that, The target noise assessment model is obtained by machine learning training of the initial noise assessment model using multiple sets of noise sample data, including noise input data and noise labeling data.

5. The vehicle anomaly detection method according to claim 4, characterized in that, The noise screening results are predicted and analyzed using the target noise assessment model to obtain the following noise assessment results: The noise screening results are evaluated based on the target noise assessment model to obtain an initial assessment result, wherein the initial assessment result is used to represent the noise distribution status information of the target vehicle; In response to the initial evaluation result not meeting the preset noise range condition, the initial evaluation result is predicted and analyzed based on the target noise evaluation model to obtain the noise evaluation result.

6. The vehicle anomaly detection method according to claim 1, characterized in that, The cross-validation results are used to determine the abnormal location information and abnormal cause information of the target vehicle.

7. The vehicle anomaly detection method according to claim 1, characterized in that, The method further includes: Anomaly reporting information is generated based on the anomaly detection results, wherein the anomaly reporting information is used to notify the target user of the abnormal status of the target vehicle; The abnormality announcement information is played using the audio component of the target vehicle.

8. The vehicle anomaly detection method according to claim 1, characterized in that, The multiple vehicle sound zones include: the front cabin sound zone, the passenger cabin sound zone, and the rear cabin sound zone.

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