An accident report generation method and related device for an automobile event recording system

Abstract

The present invention discloses a method and related device for generating an accident report for an automobile event recording system, relating to the field of data processing technology. The method comprises: collecting target accident data from various automobile event recording systems; performing data format parsing on the target accident data based on field pattern recognition to obtain the data field distribution of each data type; performing synchronous and parallel data decoding based on the data field distribution of each data type to obtain a target readable physical quantity; comparing the target readable physical quantity with the accident vehicle status information; generating vehicle state change information based on the comparison results, and performing accident scenario and accident responsibility analysis in combination with a cause-and-effect graph; generating an accident report based on the vehicle state change information, accident scenario analysis data, and accident responsibility analysis data, and visualizing the vehicle state change information. The present invention can parse and decode data formats from different automobile event recording systems, improving the accuracy and comprehensiveness of accident report generation.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to an accident report generating method and related device for an automobile event recording system. Background Art

[0002] Vehicle event recording systems are widely used in modern vehicles. They record key information including vehicle speed, acceleration, braking status, and collision force. These systems are typically activated when an accident occurs and store data before and after the accident. However, the data formats used in these systems by different automakers vary, making it impossible for third parties to directly read and analyze this data. Dedicated software is required to parse and decode the data format, making responsibility and scenario analysis of vehicle accident data cumbersome and complex. Furthermore, scenario analysis and responsibility analysis of vehicle accident data are currently typically performed through data statistics conducted by relevant personnel. However, this method relies too much on the professional expertise of the relevant personnel and cannot ensure the accuracy and comprehensiveness of the accident scenario and responsibility analysis. As a result, the resulting accident report fails to accurately reflect the specific accident circumstances, making it difficult for users to clearly and intuitively understand the accident report. Summary of the Invention

[0003] The purpose of the present invention is to overcome the shortcomings of the existing technology. The present invention provides an accident report generation method and related devices for an automobile event recording system, which can realize data format parsing and decoding of different automobile event recording systems, and improve the accuracy and comprehensiveness of accident report generation.

[0004] In order to solve the above technical problems, the present invention provides a method for generating an accident report of an automobile event recording system, the method comprising:

[0005] Collect target accident data from corresponding automobile event recording systems based on the integrity verification method;

[0006] Performing data format analysis on the target accident data based on field pattern recognition to obtain data field distribution of each data type;

[0007] Perform data synchronous and parallel decoding based on the data field distribution of each data type to obtain the target readable physical quantity;

[0008] Comparing the target readable physical quantity with the accident vehicle status information generated by analyzing the accident scene evidence information to obtain a comparison result;

[0009] Generate vehicle state change information based on the comparison result, and construct a cause-and-effect graph, and perform accident scenario analysis and accident responsibility analysis based on the vehicle state change information using the cause-and-effect graph to obtain accident scenario analysis data and accident responsibility analysis data;

[0010] An accident report is generated based on the vehicle state change information, the accident scenario analysis data, and the accident responsibility analysis data, and the accident report and the vehicle state change information are visualized.

[0011] Optionally, the collecting of target accident data from corresponding automobile event recording systems based on the integrity verification method includes:

[0012] Identifying the bus protocol of each vehicle based on a controller area network bus analyzer and an on-board diagnostic system second generation reader, and connecting to the electronic control unit of each vehicle based on the bus protocol;

[0013] After connecting to the electronic control unit of each vehicle, the accident data of the vehicle event recording system in the electronic control unit of each vehicle is collected in real time;

[0014] The accident data is integrity checked based on a preset identification data block and a preset data collection constraint to obtain an integrity check result, and target accident data in the automobile event recording system corresponding to each vehicle is determined based on the integrity check result.

[0015] Optionally, performing data format parsing on the target accident data based on field pattern recognition to obtain data field distribution of each data type includes:

[0016] Performing header information analysis on the target accident data to obtain target header information;

[0017] Acquire vehicle data patterns of each automobile event recording system, and identify several target types to which target accident data belongs using a recognition model based on the vehicle data patterns;

[0018] Acquire target fields of target accident data based on a plurality of target types and target header information, and determine corresponding field patterns based on the target fields using field value arrangement;

[0019] Based on the field pattern, the target accident data is parsed for data format using association analysis and a preset format rule library to obtain data field distribution of each data type.

[0020] Optionally, performing data synchronous parallel decoding based on the data field distribution of each data type to obtain a target readable physical quantity includes:

[0021] Acquire historical data types, historical data field distributions, and historical decoding algorithms during a historical decoding process, construct an undirected knowledge representation graph based on the historical data types, historical data field distributions, and historical decoding algorithms, and construct a matching calculation model based on the undirected knowledge representation graph;

[0022] Calculate the matching degree between the data field distribution of each data type and each decoding algorithm based on the matching calculation model, and use the decoding algorithm with the highest matching degree as the target decoding algorithm for the data field distribution of the corresponding data type;

[0023] Based on the data field distribution of each data type and the corresponding target decoding algorithm, the corresponding decoder core is determined. Based on the decoder core, the corresponding target decoding algorithm is used in combination with timestamp control to perform data synchronous and parallel decoding on the data field distribution of each data type to obtain the target readable physical quantity.

[0024] Optionally, comparing the target readable physical quantity with accident vehicle status information generated by analyzing accident scene evidence information to obtain a comparison result includes:

[0025] Extracting accident scene evidence information, performing vehicle speed analysis based on the accident scene evidence information using uncertainty analysis to obtain target vehicle speed information;

[0026] Calculating the vehicle's loss of control displacement distance using a correction coefficient and an adhesion coefficient based on the accident scene evidence information;

[0027] Performing accident simulation analysis using an object finite element model based on the accident scene evidence information to obtain accident simulation analysis data;

[0028] The target readable physical quantity is compared with the accident vehicle state information generated by the target vehicle speed information, the vehicle out-of-control displacement distance and the accident simulation analysis data to obtain a comparison result.

[0029] Optionally, generating vehicle state change information based on the comparison result and constructing a cause-and-effect graph, performing accident scenario analysis and accident responsibility analysis based on the vehicle state change information using the cause-and-effect graph to obtain accident scenario analysis data and accident responsibility analysis data, includes:

[0030] optimizing the target readable physical quantity based on the comparison result to obtain an optimized readable physical quantity, and generating vehicle state change information based on the optimized readable physical quantity;

[0031] Acquire investigation and analysis text data of historical vehicle accident events, and generate accident causal relationship event pairs based on the investigation and analysis text data;

[0032] Extracting event entities based on the investigation and analysis text data to obtain target event entities, and constructing a cause-and-effect graph based on the accident causal relationship event pairs and the target event entities using tuple mapping;

[0033] Based on the event graph, the vehicle state change information is used to perform scenario deduction to obtain scenario deduction information, and based on the scenario deduction information, the accident scenario analysis is performed using a scenario association mechanism to obtain accident scenario analysis data;

[0034] Based on the event graph, the automobile state change information is used to perform an accident cause analysis to obtain accident cause information, and based on the accident cause information, an accident responsibility analysis is performed to obtain accident responsibility analysis data.

[0035] Optionally, generating an accident report based on the vehicle state change information, the accident scenario analysis data, and the accident responsibility analysis data, and visualizing the accident report and the vehicle state change information, includes:

[0036] Inputting the vehicle state change information, accident scenario analysis data, and accident responsibility analysis data into a preset report template to obtain an accident report;

[0037] Obtaining a chart type corresponding to the accident report and the vehicle status change information, and determining a data format type corresponding to the chart type;

[0038] Perform rendering parameter analysis based on the chart type and data format type to obtain target rendering parameters;

[0039] compressing the accident report and the vehicle state change information based on hierarchical compression to obtain compressed accident report and vehicle state change information;

[0040] The compressed accident report and vehicle status change information are transmitted to a visualization tool, and the visualization tool visualizes the compressed accident report and vehicle status change information based on a chart type, a data format type, and rendering parameters.

[0041] In addition, the present invention also provides an accident report generating device for an automobile event recording system, the device comprising:

[0042] Data acquisition module: used to collect target accident data from corresponding automobile event recording systems based on integrity verification method;

[0043] Data format parsing module: used to perform data format parsing on the target accident data based on field pattern recognition to obtain data field distribution of each data type;

[0044] Data synchronization and parallel decoding module: used to perform data synchronization and parallel decoding based on the data field distribution of each data type to obtain the target readable physical quantity;

[0045] Data comparison module: used to compare the target readable physical quantity with the accident vehicle status information generated by analyzing the accident scene evidence information to obtain a comparison result;

[0046] Accident scenario analysis and accident responsibility analysis module: used to generate vehicle state change information based on the comparison results, and construct a cause-and-effect graph, and use the cause-and-effect graph to perform accident scenario analysis and accident responsibility analysis based on the vehicle state change information, thereby obtaining accident scenario analysis data and accident responsibility analysis data;

[0047] Data visualization module: used to generate an accident report based on the vehicle state change information, accident scenario analysis data and accident responsibility analysis data, and visualize the accident report and vehicle state change information.

[0048] In addition, the present invention also provides an electronic device, which includes a processor and a memory, wherein the memory is used to store instructions, and the processor is used to call the instructions in the memory so that the electronic device executes the above-mentioned accident report generation method of the automobile event recording system.

[0049] In addition, the present invention also provides a computer-readable storage medium, which stores computer instructions. When the computer instructions are executed on an electronic device, the electronic device executes the accident report generation method of the automobile event recording system.

[0050] In an embodiment of the present invention, target accident data from corresponding vehicle event recording systems is collected based on an integrity verification method, ensuring the accuracy and integrity of the collected data. Data format analysis of the target accident data is performed based on field pattern recognition to obtain the data field distribution of each data type, improving the accuracy of data format analysis and adapting it to data format analysis in different vehicle event recording systems. Data is synchronously and parallelly decoded based on the data field distribution of each data type to obtain target readable physical quantities, ensuring data consistency and accuracy during the decoding process. Data from different vehicle event recording systems can be decoded into readable physical quantities without the need for specialized software. Vehicle state change information is generated based on the comparison results, and a causal graph is constructed. Accident scenario analysis and accident responsibility analysis are performed based on the vehicle state change information using the causal graph, making the obtained accident scenario analysis and accident responsibility analysis data more accurate. An accident report is generated based on the vehicle state change information, accident scenario analysis data, and accident responsibility analysis data, ensuring that the generated accident report more accurately reflects the specific accident situation. The accident report and vehicle state change information are visualized based on the chart type, data format type, and rendering parameters, improving the visualization effect and allowing users to clearly and intuitively understand the accident situation. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0052] Figure 1 1 is a flow chart of a method for generating an accident report of an automobile event recording system in an embodiment of the present invention;

[0053] Figure 2 is a flow chart of an accident report generation method of an automobile event recording system in another embodiment of the present invention;

[0054] Figure 3 2 is a schematic diagram of the structure of an accident report generating device of an automobile event recording system in an embodiment of the present invention;

[0055] Figure 4 It is a schematic diagram of the structure of an electronic device in an embodiment of the present invention. DETAILED DESCRIPTION

[0056] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0057] Example 1

[0058] See also Figure 1 , Figure 1 FIG. 1 is a flow chart of a method for generating an accident report of an automobile event recording system according to an embodiment of the present invention. The method includes:

[0059] S11: collecting target accident data from corresponding automobile event recording systems based on an integrity verification method;

[0060] In the specific implementation process of the present invention, the target accident data in the corresponding automobile event recording system is collected based on the integrity verification method, including: identifying the bus protocol of each vehicle based on a controller area network bus analyzer and a second-generation reader of the on-board diagnostic system, and connecting to the electronic control unit of each vehicle based on the bus protocol; after connecting to the electronic control unit of each vehicle, collecting the accident data of the automobile event recording system in the electronic control unit of each vehicle in real time; performing integrity verification on the accident data based on a preset identification data block and preset data collection constraints to obtain an integrity verification result, and determining the target accident data in the automobile event recording system corresponding to each vehicle based on the integrity verification result.

[0061] Specifically, a controller area network bus analyzer and a second-generation on-board diagnostic system reader are used to identify the bus protocol of each vehicle and connect to each vehicle's electronic control unit based on the bus protocol. After connecting to each vehicle's electronic control unit, accident data from the vehicle event recording system in each vehicle's electronic control unit is collected in real time. Specifically, data before and after the accident is collected from the vehicle event recording system. The accident data may include information such as vehicle speed, acceleration, brake status, accelerator pedal position, airbag status, and speed limit. The accident data is integrity-checked based on a preset identification data block and preset data collection constraints. The preset identification data block may include a timestamp and a controller area network number, and the preset data collection constraints may include the range of various values ​​in the accident data. An integrity check result is obtained, and target accident data corresponding to each vehicle in the vehicle event recording system is determined based on the integrity check result. If the accident data fails the integrity check, the currently collected data is not processed and new accident data is collected. If the accident data passes the integrity check, the collected accident data is used as the target accident data for subsequent processing.

[0062] S12: parsing the target accident data based on field pattern recognition to obtain data field distribution of each data type;

[0063] In the specific implementation process of the present invention, the data format of the target accident data is parsed based on field pattern recognition to obtain the data field distribution of each data type, including: performing header information analysis on the target accident data to obtain target header information; obtaining the vehicle data pattern of each automobile event recording system, and using the recognition model to identify several target types to which the target accident data belongs based on the vehicle data pattern; obtaining the target field of the target accident data based on several target types and target header information, and determining the corresponding field pattern based on the target field using field assignment arrangement; performing data format parsing on the target accident data based on the field pattern using association relationship analysis and a preset format rule library to obtain the data field distribution of each data type.

[0064] Specifically, the target accident data is subjected to header information analysis to obtain target header information, which includes a magic number, timestamp, data file version number, and checksum field. The vehicle data schema of each vehicle event recording system is obtained. The vehicle data schema varies from vehicle to vehicle. The vehicle's technical manual and standards are searched based on the vehicle's make and model. The vehicle data schema, such as steering wheel status and speed status, is determined based on the technical manual and standards. Based on the vehicle data schema, a recognition model is used to identify several target types to which the target accident data belongs. The recognition model is a convergent model obtained by training a sample dataset through a deep neural network. Each field type in the accident data has its corresponding recognition rule. For example, if a field contains speed units, its type is speed. Based on several target types and target header information, the target fields of the target accident data are obtained. Byte alignment and field splitting are performed according to the several target types and target header information to obtain the target fields. Based on the target fields, the corresponding field patterns are determined by field assignment arrangement. For each target type, the target fields are assigned values ​​to obtain the corresponding values ​​of the fields. The corresponding values ​​of the fields are arranged according to a preset field order to obtain the corresponding field patterns. For example, if the preset field order is brake status, airbag status, and speed, when the target type is speed, the field containing the speed is assigned a value of one and the remaining fields are assigned a value of zero to obtain the corresponding field pattern. Based on the field pattern, the target accident data is parsed into data formats using association analysis and a preset format rule library. The field pattern is matched with the historical field patterns in the preset format rule library, that is, the field pattern is analyzed for association with each historical field pattern. The association analysis can be performed by calculating similarity. The field relationship stored in the historical field pattern with the highest similarity is used as the target field relationship. The specific value of the field is parsed according to the target field relationship and the value range, that is, the data field distribution of each data type is obtained.

[0065] S13: Perform data synchronous parallel decoding based on the data field distribution of each data type to obtain the target readable physical quantity;

[0066] In the specific implementation process of the present invention, the data field distribution based on each data type is subjected to synchronous and parallel decoding to obtain the target readable physical quantity, including: obtaining the historical data type, historical data field distribution and historical decoding algorithm in the historical decoding process, and constructing an undirected knowledge representation graph based on the historical data type, historical data field distribution and historical decoding algorithm, and constructing a matching calculation model based on the undirected knowledge representation graph; calculating the matching degree of the data field distribution of each data type and each decoding algorithm based on the matching calculation model, and taking the decoding algorithm with the highest matching degree as the target decoding algorithm for the data field distribution of the corresponding data type; determining the corresponding decoder core based on the data field distribution of each data type and the corresponding target decoding algorithm, and performing data synchronous and parallel decoding on the data field distribution of each data type based on the decoder core using the corresponding target decoding algorithm combined with timestamp control to obtain the target readable physical quantity.

[0067] Specifically, the historical data types, historical data field distributions, and historical decoding algorithms in the historical decoding process are obtained, and the decoding algorithms include Base64 decoding, Huffman decoding, etc., and an undirected knowledge representation graph is constructed based on the historical data types, historical data field distributions, and historical decoding algorithms. The edge relationship between the historical data types and historical data field distributions and the historical decoding algorithms is obtained, and nodes are constructed based on the historical data types, historical data field distributions, and historical decoding algorithms. An undirected knowledge representation graph is constructed based on the constructed nodes and edge relationships. Model training is performed based on the embedded representation of the historical data types, historical data field distributions, and historical decoding algorithms combined with the undirected knowledge representation graph to construct a matching calculation model. Based on the matching calculation model, the data field distribution of each data type is matched with each decoding algorithm. The data field distribution of each data type is input into the matching calculation model to calculate the matching degree of the decoding algorithm. The decoding algorithm with the highest matching degree is used as the target decoding algorithm for the data field distribution of the corresponding data type. Different decoding schemes can be adopted for different data types to improve the reliability of data decoding. Based on the data field distribution of each data type and the corresponding target decoding algorithm, the corresponding decoder core is determined. Based on the decoder core, the data field distribution of each data type is subjected to data synchronous and parallel decoding using the corresponding target decoding algorithm combined with timestamp control. When performing data synchronous and parallel decoding, a comparison is performed based on the timestamp in the header information, and the latest time point in the timestamp of the first frame of data is selected as the benchmark. The data before the latest time point is decoded backward until the timestamps of all data are consistent with the latest time point. Decoding is stopped, thereby achieving time-synchronized decoding and obtaining the target readable physical quantity, that is, obtaining visualized physical quantities such as vehicle speed and acceleration. Data synchronous and parallel decoding can ensure the time synchronization of each data stream during decoding to restore the situation before and after the accident.

[0068] S14: comparing the target readable physical quantity with the accident vehicle status information generated by analyzing the accident scene evidence information to obtain a comparison result;

[0069] During the specific implementation of the present invention, the target readable physical quantity is compared with the accident vehicle status information generated by analyzing the accident scene evidence information to obtain a comparison result, including: extracting the accident scene evidence information, performing vehicle speed analysis based on the accident scene evidence information using uncertainty analysis to obtain target vehicle speed information; calculating the vehicle out-of-control displacement distance based on the accident scene evidence information using a correction coefficient and an adhesion coefficient; performing an accident simulation analysis based on the accident scene evidence information using an object finite element model to obtain accident simulation analysis data; and comparing the target readable physical quantity with the accident vehicle status information generated by the target vehicle speed information, the vehicle out-of-control displacement distance and the accident simulation analysis data to obtain a comparison result.

[0070] Specifically, accident scene evidence information is extracted, such as brake mark information, collision mark information, monitoring and transcripts, etc., uncertainty analysis is used to perform vehicle speed analysis based on the accident scene evidence information, historical accident scene evidence information is obtained, the historical accident scene evidence information is normalized to obtain normalized historical accident scene evidence information, posterior parameter analysis is performed based on the normalized historical accident scene evidence information using Bayesian theorem to obtain posterior parameters, and accident reconstruction based on uncertainty deduction is performed using empirical models and simulation models based on the posterior parameters to obtain several accident reconstruction results, and each accident reconstruction result is voted on through a voting algorithm, and accident reconstruction results with voting scores lower than a preset threshold are deleted, and vehicle speed analysis is performed based on the retained several accident reconstruction results to obtain vehicle speed information corresponding to the retained several accident reconstruction results, and an average is calculated based on each vehicle speed information, and the calculated average vehicle speed is used as the final vehicle speed information, that is, the target vehicle speed information is obtained. Based on the accident scene evidence information, the vehicle's loss of control displacement distance is calculated using a correction coefficient and an adhesion coefficient. The angular range of the road condition survey starting point is obtained based on the accident scene evidence information. The plastic deformation of the vehicle damage is analyzed using the adhesion coefficient based on the collision trace information in the accident scene evidence information. The spatiotemporal displacement distance of the vehicle is calculated based on the angular range and the plastic deformation of the vehicle damage using the target vehicle speed information combined with the correction coefficient. An accident simulation analysis is performed based on the accident scene evidence information using a finite element model of the object. The accident scene evidence information is used to calculate the stress and strain data of the vehicle at the time of the accident using the finite element model of the object, thereby obtaining accident simulation analysis data. The target readable physical quantity is compared with the accident vehicle status information generated by the target vehicle speed information, the vehicle out-of-control displacement distance and the accident simulation analysis data, the target vehicle speed information and the vehicle out-of-control displacement distance are optimized according to the accident simulation analysis data to obtain the optimized vehicle speed information and the optimized out-of-control displacement distance, the accident vehicle status information is generated according to the optimized vehicle speed information and the optimized out-of-control displacement distance, the out-of-control distance is analyzed according to the speed and braking status in the target readable physical quantity, the speed and out-of-control distance in the target readable physical quantity are compared with the optimized vehicle speed information and the optimized out-of-control displacement distance in the accident vehicle status information, and it is determined whether the difference between the data value in the readable physical quantity and the data value in the accident vehicle status information is within a preset allowable range; if it is within the preset allowable range, the target readable physical quantity will continue to be used as data for subsequent analysis and processing; if it is not within the preset allowable range, the data will be re-collected and processed until it reaches the preset allowable range to obtain a comparison result.

[0071] S15: generating vehicle state change information based on the comparison result, and constructing a cause-and-effect graph, performing accident scenario analysis and accident responsibility analysis based on the vehicle state change information using the cause-and-effect graph, and obtaining accident scenario analysis data and accident responsibility analysis data;

[0072] In the specific implementation process of the present invention, the vehicle state change information is generated based on the comparison result, and a causal graph is constructed. The accident scenario analysis and accident responsibility analysis are performed based on the vehicle state change information using the causal graph to obtain accident scenario analysis data and accident responsibility analysis data, including: optimizing the target readable physical quantity based on the comparison result to obtain the optimized readable physical quantity, and generating vehicle state change information based on the optimized readable physical quantity; obtaining investigation and analysis text data of historical vehicle accident events, and generating accident causal relationship event pairs based on the investigation and analysis text data; extracting event entities based on the investigation and analysis text data to obtain target event entities, and constructing a causal graph based on the accident causal relationship event pairs and target event entities using tuple mapping; performing scenario deduction based on the vehicle state change information based on the causal graph to obtain scenario deduction information, and performing accident scenario analysis based on the scenario deduction information using scenario association mechanism to obtain accident scenario analysis data; performing accident cause analysis based on the causal graph using the vehicle state change information to obtain accident cause information, and performing accident responsibility analysis based on the accident cause information to obtain accident responsibility analysis data.

[0073] Specifically, the target readable physical quantity is optimized based on the comparison result, that is, after the target readable physical quantity is compared with the accident vehicle status information, the target readable physical quantity and the numerical value in the accident vehicle status information are averaged, such as the speed in the target readable physical quantity and the speed in the accident vehicle status information are averaged to obtain the optimized readable physical quantity, and vehicle status change information is generated based on the optimized readable physical quantity, that is, the data in the optimized readable physical quantity is compared with the data stored in the automobile event recording system before the accident, and a change curve and chart are formed, which is the vehicle status change information. The investigation and analysis text data of historical vehicle accident events are obtained. The investigation and analysis text data include the responsibility for the accident, accident scene analysis, accident data, etc., and the accident causal relationship event pairs are generated based on the investigation and analysis text data. The verb closest to the causal relationship word in the investigation and analysis text data is used as the event trigger word. The causal relationship words include words such as cause, due to and because. The subject and object corresponding to the event trigger word are obtained, and the representation form of the event pair is determined according to the event trigger word and the corresponding subject and object. The double-layer model causal relationship extraction method based on the residual idea is used to extract the causal relationship event pairs. Two deep learning models are used for distribution processing to enhance the recognition of relationship boundaries. In the model In the internal structure, the Transformer-based bidirectional encoder representation model structure is used to extract rich semantic features, and then the features are linearly weighted fused based on the residual idea combined with the convolutional neural network and the bidirectional gated recurrent unit model to enhance the semantic representation ability. The causal semantic role information in the survey and analysis text data is obtained through the Transformer-based bidirectional encoder representation model, and the causal semantic role information is input into the convolutional neural network and the bidirectional gated recurrent unit model to divide the causal boundary. The conditional random field model is used to predict the probability of the label sequence to extract the causal relationship of the event, and the form of the causal event pair is converted into the representation form of the event pair. Event entity extraction is performed based on the investigation and analysis text data, and event entity extraction is performed on the investigation and analysis text data based on a supervised learning model to obtain a target event entity, and a causal graph is constructed based on the accident causal event pair and the target event entity using tuple mapping, and the accident causal event pair is event-tuple mapped with the accident domain ontology. Event tuple mapping is to calculate the similarity between the accident causal event pair and the tuple in the accident domain ontology, and the tuple with the highest similarity with the tuple with the highest similarity in the accident domain ontology is combined with the accident causal event pair and the target event entity to construct a causal graph, so that the constructed causal graph can mine the causal relationship of the accident.Based on the event graph, scenario deduction is performed using the vehicle state change information, and the deduction object and deduction path are determined according to the vehicle state change information. Scenario deduction is performed in the event graph according to the deduction object and deduction path, that is, the accident scenario deduction is performed to obtain scenario deduction information, and based on the scenario deduction information, the accident scenario analysis is performed using the scenario association mechanism. The scenario association mechanism is the association relationship of accident scenarios based on existing accident cases. The accident scenario is reconstructed according to the scenario association mechanism and the scenario deduction information, and the vehicle motion trajectory and motion angle before and after the accident are obtained according to the reconstructed accident scenario, that is, the accident scenario analysis data is obtained. Based on the event graph, the vehicle state change information is used to analyze the cause of the accident. Based on the event graph, the vehicle state change information is used to analyze the cause path of the accident. The cause information of the specific accident is determined based on the accident transmission path of different types of accidents combined with the analyzed accident cause path. For example, if a collision accident occurs due to excessive vehicle speed, the cause information of the accident is obtained. Based on the accident cause information, an accident responsibility analysis is performed, that is, the illegal vehicle is determined based on the accident cause information, and the illegal user is determined based on the illegal vehicle, to obtain accident responsibility analysis data.

[0074] S16: Generate an accident report based on the vehicle state change information, the accident scenario analysis data, and the accident responsibility analysis data, and visualize the accident report and the vehicle state change information.

[0075] In the specific implementation process of the present invention, an accident report is generated based on the vehicle state change information, accident scenario analysis data and accident responsibility analysis data, and the accident report and vehicle state change information are visualized, including: inputting the vehicle state change information, accident scenario analysis data and accident responsibility analysis data into a preset report template to obtain an accident report; obtaining a chart type corresponding to the accident report and the vehicle state change information, and determining a data format type corresponding to the chart type; performing rendering parameter analysis based on the chart type and data format type to obtain target rendering parameters; compressing the accident report and vehicle state change information based on hierarchical compression to obtain a compressed accident report and vehicle state change information; transmitting the compressed accident report and vehicle state change information to a visualization tool, and the visualization tool visually displays the compressed accident report and vehicle state change information based on the chart type, data format type and rendering parameters.

[0076] Specifically, the vehicle status change information, accident scenario analysis data and accident responsibility analysis data are input into a preset report template to obtain an accident report. The chart type corresponding to the accident report and the vehicle status change information is obtained. The corresponding chart type can be determined according to the data name corresponding to the accident report and the vehicle status change information, and the data format type corresponding to the chart type is determined. The data format type corresponding to the chart type is queried in the visualization list. Rendering parameter analysis is performed based on the chart type and data format type, and component configuration information and interactive event configuration information are determined according to the chart type and data format type. The component configuration information includes the configured components, such as tab components, event components and image components, etc. The interactive event configuration information includes the configuration of interactive operations, such as dragging and selecting operations on the visualization interface, etc. Corresponding listeners and triggers are generated according to the interactive event configuration information, and the rendering layout and component rendering pixels of the display window for visual display are determined according to the component configuration information. The target rendering parameters are composed of the corresponding listeners and triggers, as well as the rendering layout and component rendering pixels. The accident report and vehicle status change information are compressed based on hierarchical compression, and the accident report and vehicle status change information are sampled according to different sampling steps to form sampled data of different resolutions. The sampled data is the sampled accident report and vehicle status change information. The sampled data of different resolutions are sorted according to the resolution size to obtain hierarchical data. The hierarchical data is block-compressed using a block-based compression method to obtain compressed data, i.e., the compressed accident report and vehicle status change information. By performing hierarchical compression on the accident report and vehicle status change information, network transmission pressure can be reduced. The compressed accident report and vehicle status change information are transmitted to a visualization tool. The visualization tool visualizes the compressed accident report and vehicle status change information based on chart type, data format type, and rendering parameters. The visualization tool obtains a hierarchical compressed data list, decompresses the compressed accident report and vehicle status change information according to the hierarchical compressed data list, and determines a corresponding visualization layout based on the chart type, data format type, and rendering parameters. The decompressed accident report and vehicle status change information are displayed in the visualization layout, which can realize visualization of different types of data and improve visualization effect.

[0077] In an embodiment of the present invention, target accident data from corresponding vehicle event recording systems is collected based on an integrity verification method, ensuring the accuracy and integrity of the collected data. Data format analysis of the target accident data is performed based on field pattern recognition to obtain the data field distribution of each data type, improving the accuracy of data format analysis and adapting it to data format analysis in different vehicle event recording systems. Data is synchronously and parallelly decoded based on the data field distribution of each data type to obtain target readable physical quantities, ensuring data consistency and accuracy during the decoding process. Data from different vehicle event recording systems can be decoded into readable physical quantities without the need for specialized software. Vehicle state change information is generated based on the comparison results, and a causal graph is constructed. Accident scenario analysis and accident responsibility analysis are performed based on the vehicle state change information using the causal graph, making the obtained accident scenario analysis and accident responsibility analysis data more accurate. An accident report is generated based on the vehicle state change information, accident scenario analysis data, and accident responsibility analysis data, ensuring that the generated accident report more accurately reflects the specific accident situation. The accident report and vehicle state change information are visualized based on the chart type, data format type, and rendering parameters, improving the visualization effect and allowing users to clearly and intuitively understand the accident situation.

[0078] Example 2

[0079] See also Figure 2 , Figure 2 FIG. 1 is a flow chart of a method for generating an accident report of an automobile event recording system in another embodiment of the present invention, the method comprising:

[0080] S201: collecting target accident data from corresponding automobile event recording systems based on an integrity verification method;

[0081] S202: parsing the target accident data based on field pattern recognition to obtain data field distribution of each data type;

[0082] S203: Perform data synchronous parallel decoding based on the data field distribution of each data type to obtain the target readable physical quantity;

[0083] S204: comparing the target readable physical quantity with the accident vehicle status information generated by analyzing the accident scene evidence information to obtain a comparison result;

[0084] S205: Optimizing the target readable physical quantity based on the comparison result to obtain an optimized readable physical quantity, and generating vehicle state change information based on the optimized readable physical quantity;

[0085] S206: Acquire investigation and analysis text data of historical vehicle accident events, and generate accident causal relationship event pairs based on the investigation and analysis text data;

[0086] S207: extracting event entities based on the investigation and analysis text data to obtain target event entities, and constructing a cause-and-effect graph based on the accident causal relationship event pairs and the target event entities using tuple mapping;

[0087] S208: performing scenario deduction based on the event graph and the vehicle state change information to obtain scenario deduction information, and performing accident scenario analysis based on the scenario deduction information and the scenario association mechanism to obtain accident scenario analysis data;

[0088] S209: performing an accident cause analysis based on the event graph and the vehicle state change information to obtain accident cause information, and performing an accident responsibility analysis based on the accident cause information to obtain accident responsibility analysis data;

[0089] S210: Generate an accident report based on the vehicle state change information, the accident scenario analysis data, and the accident responsibility analysis data, and visualize the accident report and the vehicle state change information.

[0090] In an embodiment of the present invention, target accident data from corresponding vehicle event recording systems is collected based on an integrity verification method, ensuring the accuracy and integrity of the collected data. Data format analysis of the target accident data is performed based on field pattern recognition to obtain the data field distribution of each data type, improving the accuracy of data format analysis and adapting it to data format analysis in different vehicle event recording systems. Data is synchronously and parallelly decoded based on the data field distribution of each data type to obtain target readable physical quantities, ensuring data consistency and accuracy during the decoding process. Data from different vehicle event recording systems can be decoded into readable physical quantities without the need for specialized software. Vehicle state change information is generated based on the comparison results, and a causal graph is constructed. Accident scenario analysis and accident responsibility analysis are performed based on the vehicle state change information using the causal graph, making the obtained accident scenario analysis and accident responsibility analysis data more accurate. An accident report is generated based on the vehicle state change information, accident scenario analysis data, and accident responsibility analysis data, ensuring that the generated accident report more accurately reflects the specific accident situation. The accident report and vehicle state change information are visualized based on the chart type, data format type, and rendering parameters, improving the visualization effect and allowing users to clearly and intuitively understand the accident situation.

[0091] Example 3

[0092] See also Figure 3 , Figure 3: is a schematic diagram of the structure of an accident report generating device of an automobile event recording system in an embodiment of the present invention, the device comprising:

[0093] Data collection module 31: used for collecting target accident data from corresponding automobile event recording systems based on the integrity verification method;

[0094] Data format parsing module 32: used to perform data format parsing on the target accident data based on field pattern recognition to obtain data field distribution of each data type;

[0095] Data synchronization and parallel decoding module 33: used for performing data synchronization and parallel decoding based on the data field distribution of each data type to obtain the target readable physical quantity;

[0096] Data comparison module 34: used to compare the target readable physical quantity with the accident vehicle status information generated by analyzing the accident scene evidence information to obtain a comparison result;

[0097] Accident scenario analysis and accident responsibility analysis module 35: used to generate vehicle state change information based on the comparison results, and construct a cause-and-effect graph, and perform accident scenario analysis and accident responsibility analysis based on the vehicle state change information using the cause-and-effect graph to obtain accident scenario analysis data and accident responsibility analysis data;

[0098] Data visualization module 36: used to generate an accident report based on the vehicle state change information, accident scenario analysis data and accident responsibility analysis data, and visualize the accident report and vehicle state change information.

[0099] In the specific implementation process of the present invention, the specific implementation method of the device item can refer to the implementation method of the above-mentioned method item, which will not be repeated here.

[0100] In an embodiment of the present invention, target accident data from corresponding vehicle event recording systems is collected based on an integrity verification method, ensuring the accuracy and integrity of the collected data. Data format analysis of the target accident data is performed based on field pattern recognition to obtain the data field distribution of each data type, improving the accuracy of data format analysis and adapting it to data format analysis in different vehicle event recording systems. Data is synchronously and parallelly decoded based on the data field distribution of each data type to obtain target readable physical quantities, ensuring data consistency and accuracy during the decoding process. Data from different vehicle event recording systems can be decoded into readable physical quantities without the need for specialized software. Vehicle state change information is generated based on the comparison results, and a causal graph is constructed. Accident scenario analysis and accident responsibility analysis are performed based on the vehicle state change information using the causal graph, making the obtained accident scenario analysis and accident responsibility analysis data more accurate. An accident report is generated based on the vehicle state change information, accident scenario analysis data, and accident responsibility analysis data, ensuring that the generated accident report more accurately reflects the specific accident situation. The accident report and vehicle state change information are visualized based on the chart type, data format type, and rendering parameters, improving the visualization effect and allowing users to clearly and intuitively understand the accident situation.

[0101] An embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the method for generating an accident report for an automobile event recording system according to any of the above-mentioned embodiments is implemented. The computer-readable storage medium includes, but is not limited to, any type of disk (including floppy disks, hard disks, optical disks, CD-ROMs, and magneto-optical disks), ROM (Read-Only Memory), RAM (Random Access Memory), EPROM (Erasable Programmable Read-Only Memory), EEPROM (Electrically Erasable Programmable Read-Only Memory), flash memory, magnetic cards, or optical cards. In other words, a storage device includes any medium that can store or transmit information in a readable form by a device (e.g., a computer or a mobile phone), and can be a read-only memory, a disk, or an optical disk.

[0102] Example 4

[0103] See also Figure 4 , Figure 4 It is a schematic diagram of the structure of an electronic device in an embodiment of the present invention.

[0104] The embodiment of the present invention further provides an electronic device, such as Figure 4 As shown, the electronic device includes a memory 41, a processor 43, and a computer program 42 stored in the memory 41 and executable on the processor 43. It will be understood by those skilled in the art that Figure 3 The electronic devices shown do not constitute a limitation on all devices and may include more or fewer components than shown, or combinations of certain components. The memory 41 can be used to store the computer program 42 and various functional modules, and the processor 43 runs the computer program 42 stored in the memory 41, thereby executing various functional applications and data processing of the device. The memory can be internal memory or external memory, or include both internal memory and external memory. The internal memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, or random access memory. The external memory can include a hard disk, floppy disk, ZIP disk, USB flash drive, magnetic tape, etc. The processor 43 can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor, a single-chip microcomputer, or processor 43, or any conventional processor. The processor and memory disclosed in the present invention include but are not limited to these types of processors and memories. The processor and memory disclosed in the present invention are only examples and not limitations.

[0105] As an embodiment, the electronic device includes: one or more processors 43, a memory 41, and one or more computer programs 42, wherein the one or more computer programs 42 are stored in the memory 41 and are configured to be executed by the one or more processors 43, and the one or more computer programs 42 are configured to execute the accident report generation method of the automobile event recording system in any of the above-mentioned embodiments. For the specific implementation process, please refer to the above-mentioned embodiments and will not be repeated here.

[0106] In an embodiment of the present invention, target accident data from corresponding vehicle event recording systems is collected based on an integrity verification method, ensuring the accuracy and integrity of the collected data. Data format analysis of the target accident data is performed based on field pattern recognition to obtain the data field distribution of each data type, improving the accuracy of data format analysis and adapting it to data format analysis in different vehicle event recording systems. Data is synchronously and parallelly decoded based on the data field distribution of each data type to obtain target readable physical quantities, ensuring data consistency and accuracy during the decoding process. Data from different vehicle event recording systems can be decoded into readable physical quantities without the need for specialized software. Vehicle state change information is generated based on the comparison results, and a causal graph is constructed. Accident scenario analysis and accident responsibility analysis are performed based on the vehicle state change information using the causal graph, making the obtained accident scenario analysis and accident responsibility analysis data more accurate. An accident report is generated based on the vehicle state change information, accident scenario analysis data, and accident responsibility analysis data, ensuring that the generated accident report more accurately reflects the specific accident situation. The accident report and vehicle state change information are visualized based on the chart type, data format type, and rendering parameters, improving the visualization effect and allowing users to clearly and intuitively understand the accident situation.

[0107] In addition, the above is a detailed introduction to the accident report generation method and related devices of the automobile event recording system provided by the embodiment of the present invention. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for general technical personnel in this field, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A method for generating an accident report of an automobile event recording system, characterized in that: The method comprises: Collect target accident data from corresponding automobile event recording systems based on the integrity verification method; Performing data format analysis on the target accident data based on field pattern recognition to obtain data field distribution of each data type; Perform data synchronous and parallel decoding based on the data field distribution of each data type to obtain the target readable physical quantity; Comparing the target readable physical quantity with the accident vehicle status information generated by analyzing the accident scene evidence information to obtain a comparison result; Generate vehicle state change information based on the comparison result, and construct a cause-and-effect graph, and perform accident scenario analysis and accident responsibility analysis based on the vehicle state change information using the cause-and-effect graph to obtain accident scenario analysis data and accident responsibility analysis data; generating an accident report based on the vehicle state change information, the accident scenario analysis data, and the accident responsibility analysis data, and visualizing the accident report and the vehicle state change information; The data format parsing of the target accident data based on field pattern recognition to obtain data field distribution of each data type includes: performing header information analysis on the target accident data to obtain target header information; obtaining vehicle data patterns of each automobile event recording system, and identifying several target types to which the target accident data belongs using a recognition model based on the vehicle data patterns; obtaining target fields of the target accident data based on the several target types and target header information, and determining corresponding field patterns based on the target fields using field value arrangement; performing data format parsing on the target accident data based on the field patterns using association analysis and a preset format rule library to obtain data field distribution of each data type; The method of performing synchronous and parallel decoding of data based on the data field distribution of each data type to obtain a target readable physical quantity includes: obtaining historical data types, historical data field distributions and historical decoding algorithms in a historical decoding process, and constructing an undirected knowledge representation graph based on the historical data types, historical data field distributions and historical decoding algorithms, and constructing a matching calculation model based on the undirected knowledge representation graph; calculating the matching degree of the data field distribution of each data type and each decoding algorithm based on the matching calculation model, and using the decoding algorithm with the highest matching degree as the target decoding algorithm for the data field distribution of the corresponding data type; determining a corresponding decoder core based on the data field distribution of each data type and the corresponding target decoding algorithm, and performing synchronous and parallel decoding of the data field distribution of each data type based on the decoder core using the corresponding target decoding algorithm combined with timestamp control to obtain a target readable physical quantity.

2. The accident report generating method of the automobile event recording system according to claim 1, characterized in that: The target accident data collected from the corresponding automobile event recording systems based on the integrity verification method includes: Identifying the bus protocol of each vehicle based on a controller area network bus analyzer and an on-board diagnostic system second generation reader, and connecting to the electronic control unit of each vehicle based on the bus protocol; After connecting to the electronic control unit of each vehicle, the accident data of the vehicle event recording system in the electronic control unit of each vehicle is collected in real time; The accident data is integrity checked based on a preset identification data block and a preset data collection constraint to obtain an integrity check result, and target accident data in the automobile event recording system corresponding to each vehicle is determined based on the integrity check result.

3. The accident report generation method of the automobile event recording system according to claim 1, characterized in that: The step of comparing the target readable physical quantity with the accident vehicle status information generated by analyzing the accident scene evidence information to obtain a comparison result includes: Extracting accident scene evidence information, performing vehicle speed analysis based on the accident scene evidence information using uncertainty analysis to obtain target vehicle speed information; Calculating the vehicle's loss of control displacement distance using a correction coefficient and an adhesion coefficient based on the accident scene evidence information; Performing accident simulation analysis using an object finite element model based on the accident scene evidence information to obtain accident simulation analysis data; The target readable physical quantity is compared with the accident vehicle state information generated by the target vehicle speed information, the vehicle out-of-control displacement distance and the accident simulation analysis data to obtain a comparison result.

4. The accident report generation method of the automobile event recording system according to claim 1, characterized in that: Generating vehicle state change information based on the comparison result and constructing a cause-and-effect graph, performing accident scenario analysis and accident responsibility analysis based on the vehicle state change information using the cause-and-effect graph, and obtaining accident scenario analysis data and accident responsibility analysis data, including: optimizing the target readable physical quantity based on the comparison result to obtain an optimized readable physical quantity, and generating vehicle state change information based on the optimized readable physical quantity; Acquire investigation and analysis text data of historical vehicle accident events, and generate accident causal relationship event pairs based on the investigation and analysis text data; Extracting event entities based on the investigation and analysis text data to obtain target event entities, and constructing a cause-and-effect graph based on the accident causal relationship event pairs and the target event entities using tuple mapping; Based on the event graph, the vehicle state change information is used to perform scenario deduction to obtain scenario deduction information, and based on the scenario deduction information, the accident scenario analysis is performed using a scenario association mechanism to obtain accident scenario analysis data; Based on the event graph, the automobile state change information is used to perform an accident cause analysis to obtain accident cause information, and based on the accident cause information, an accident responsibility analysis is performed to obtain accident responsibility analysis data.

5. The accident report generation method of the automobile event recording system according to claim 1, characterized in that: Generating an accident report based on the vehicle state change information, the accident scenario analysis data, and the accident responsibility analysis data, and visualizing the accident report and the vehicle state change information, includes: Inputting the vehicle state change information, accident scenario analysis data, and accident responsibility analysis data into a preset report template to obtain an accident report; Obtaining a chart type corresponding to the accident report and the vehicle status change information, and determining a data format type corresponding to the chart type; Perform rendering parameter analysis based on the chart type and data format type to obtain target rendering parameters; compressing the accident report and the vehicle state change information based on hierarchical compression to obtain compressed accident report and vehicle state change information; The compressed accident report and vehicle status change information are transmitted to a visualization tool, and the visualization tool visualizes the compressed accident report and vehicle status change information based on a chart type, a data format type, and rendering parameters.

6. An accident report generating device for an automobile event recording system, characterized in that: The device comprises: Data acquisition module: used to collect target accident data from corresponding automobile event recording systems based on integrity verification method; Data format parsing module: used to perform data format parsing on the target accident data based on field pattern recognition to obtain data field distribution of each data type; Data synchronization and parallel decoding module: used to perform data synchronization and parallel decoding based on the data field distribution of each data type to obtain the target readable physical quantity; Data comparison module: used to compare the target readable physical quantity with the accident vehicle status information generated by analyzing the accident scene evidence information to obtain a comparison result; Accident scenario analysis and accident responsibility analysis module: used to generate vehicle state change information based on the comparison results, and construct a cause-and-effect graph, and use the cause-and-effect graph to perform accident scenario analysis and accident responsibility analysis based on the vehicle state change information, thereby obtaining accident scenario analysis data and accident responsibility analysis data; Data visualization module: used to generate an accident report based on the vehicle state change information, accident scenario analysis data and accident responsibility analysis data, and visualize the accident report and vehicle state change information; The data format parsing of the target accident data based on field pattern recognition to obtain data field distribution of each data type includes: performing header information analysis on the target accident data to obtain target header information; obtaining vehicle data patterns of each automobile event recording system, and identifying several target types to which the target accident data belongs using a recognition model based on the vehicle data patterns; obtaining target fields of the target accident data based on the several target types and target header information, and determining corresponding field patterns based on the target fields using field value arrangement; performing data format parsing on the target accident data based on the field patterns using association analysis and a preset format rule library to obtain data field distribution of each data type; The method of performing synchronous and parallel decoding of data based on the data field distribution of each data type to obtain a target readable physical quantity includes: obtaining historical data types, historical data field distributions and historical decoding algorithms in a historical decoding process, and constructing an undirected knowledge representation graph based on the historical data types, historical data field distributions and historical decoding algorithms, and constructing a matching calculation model based on the undirected knowledge representation graph; calculating the matching degree of the data field distribution of each data type and each decoding algorithm based on the matching calculation model, and using the decoding algorithm with the highest matching degree as the target decoding algorithm for the data field distribution of the corresponding data type; determining a corresponding decoder core based on the data field distribution of each data type and the corresponding target decoding algorithm, and performing synchronous and parallel decoding of the data field distribution of each data type based on the decoder core using the corresponding target decoding algorithm combined with timestamp control to obtain a target readable physical quantity.

7. An electronic device comprising a processor and a memory, characterized in that: The memory is used to store instructions, and the processor is used to call the instructions in the memory, so that the electronic device executes the accident report generation method of the automobile event recording system according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, which, when executed on an electronic device, enable the electronic device to execute the accident report generating method of the automobile event recording system according to any one of claims 1 to 5 .

Images