Automatic traffic accident disposal system based on large model and block chain technology

Abstract

The invention relates to a traffic accident automatic handling system based on a large model and a block chain technology. The traffic accident automatic handling system comprises a traffic accident analysis intelligent module, a traffic accident liability confirmation intelligent module, a traffic accident report auditing intelligent module and a whole process data storage and processing module based on the block chain technology. The three intelligent modules cooperate with one another by simulating the thinking modes of different roles in the real world disposal process to jointly complete automatic identification of traffic accidents. According to the invention, the automation of traffic accident disposal is realized, the accuracy and fairness of traffic accident identification are improved, and the traceability of accident original data and the transparency of the processing process are ensured.

Description

Technical Field

[0001] The present invention relates to an automatic traffic accident handling system based on a large model and blockchain technology, belonging to the technical field of traffic management. Background Art

[0002] With the rapid development of road traffic and the growth of motor vehicle ownership in my country, improving traffic management capabilities has become a key task for transportation departments. Accelerating the traffic accident handling process is a key aspect of improving traffic management capabilities. It not only quickly alleviates traffic congestion caused by traffic accidents, but also effectively reduces the occurrence of secondary accidents.

[0003] The existing traffic accident handling process mostly relies on traffic police arriving at the scene to manually identify the accident. This method is not only inefficient but also requires significant time and labor costs. In recent years, automated traffic accident handling methods based on video call technology have been proposed. After a traffic accident occurs, the parties involved can report the incident online through video and conduct online identification. This method saves traffic police time and improves handling efficiency to a certain extent. However, it still relies on manual labor and requires a significant investment of manpower.

[0004] Therefore, developing an automated traffic accident handling system that does not rely on manual labor will be able to fundamentally solve the problem of low efficiency in traffic accident handling from the two dimensions of time and manpower. Summary of the Invention

[0005] In order to overcome the shortcomings of low efficiency and high labor costs in existing traffic accident handling, the present invention provides an automatic traffic accident handling system based on large models and blockchain technology, which realizes the automation of traffic accident handling without relying on manual labor, and fundamentally solves the problem of low efficiency in traffic accident handling.

[0006] A traffic accident automatic handling system based on a large model and blockchain technology includes the following steps:

[0007] Traffic accident analysis intelligent module, used to analyze traffic accidents and extract core feature information;

[0008] Traffic accident responsibility confirmation intelligent module, used to determine the specific responsibility for traffic accidents;

[0009] Traffic accident report review intelligent module, used to review and confirm the traffic accident responsibility determination and generate a report after review and confirmation;

[0010] A full-process data storage and processing module based on blockchain technology is used to store accident-related feature data, full-process reasoning data of the intelligent module system, and confirmation report data of the accident responsible party on the chain;

[0011] Among them, the traffic accident analysis intelligent module, traffic accident responsibility confirmation intelligent module and traffic accident report review intelligent energy body simulate the thinking mode of different roles in the real-world handling process, cooperate with each other, and jointly complete the automated identification of traffic accidents.

[0012] The enhancement method of the traffic accident analysis intelligent module includes: constructing a role-playing mechanism based on the thinking of traffic police, simulating the on-site investigation perspective of traffic police, and improving the observation ability of the traffic accident analysis intelligent module in traffic accident analysis; constructing a large model fine-tuning method to enhance the accident calculation and analysis capabilities, and improving the traffic accident analysis intelligent module's quantitative analysis capability of the severity of accidents in traffic accident analysis; and constructing an analysis method based on retrieval enhancement judgment of a historical accident knowledge base, providing auxiliary information for the large model intelligent body module in traffic accident analysis based on the knowledge base and through a retrieval enhancement generation method, and improving the decision-making ability of the large model intelligent body module to capture key features of traffic accidents.

[0013] The role-playing mechanism based on traffic police thinking specifically includes: constructing the role background of the agent module and building a reasoning chain of accident investigation. The role background of the agent module is constructed by setting background task instructions to guide the large model to conduct subsequent reasoning analysis using the thinking method of a traffic police officer. The reasoning chain of accident investigation begins by constructing multiple rounds of reasoning prompts, sequentially analyzing vehicle trajectories, pedestrian behavior, and weather data, and then verifying logical contradictions through multiple rounds of reasoning.

[0014] Preferably, the large-model fine-tuning method for enhancing the accident calculation and analysis capabilities specifically includes: utilizing momentum conservation, collision angle calculation, and other methods to fine-tune the parameters of the traffic accident analysis intelligent agent module based on traffic accident physical model data, focusing on optimizing the large-model's quantitative calculation capabilities during the collision accident analysis process, and improving the ability to judge the severity of the accident and the collision mode.

[0015] Preferably, the historical accident knowledge base in the analysis method for enhancing judgment through retrieval based on a historical accident knowledge base is composed of historical traffic accident data, including real-time video or photographic data of traffic accidents, weather data, and descriptions of key accident features. The historical accident knowledge base is searched using a key feature similarity matching method to provide reference auxiliary information for the analysis method for enhancing judgment through retrieval based on the historical accident knowledge base.

[0016] The enhancement method for the traffic accident liability determination intelligent module includes: constructing a knowledge graph for traffic accident liability determination, utilizing information from the knowledge graph to assist the large-scale model in decision-making. The nodes of the knowledge graph encompass provisions of the Road Traffic Safety Law, local regulations, and typical precedents, with edges defining liability association rules. Furthermore, a dual-perspective judgment mechanism based on "legal experts" and "insurance adjusters" is implemented. The "legal expert" judgment mechanism utilizes the traffic accident liability determination knowledge graph to determine legal liability, while the "insurance adjuster" judgment mechanism uses the analysis results of the "legal expert" judgment mechanism and combines them with insurance clause reasoning and analysis to determine financial liability and assess the compensation ratio. The insurance clauses are provided to the large-scale model intelligent agent module as contextual input information.

[0017] The enhancement method of the traffic accident report review intelligent module includes: building a standardized information integrity review workflow, establishing an information extraction question-and-answer mechanism based on a large model, guiding the traffic accident responsibility confirmation intelligent module to conduct information integrity review, and the review content includes information on the parties involved, responsibility ratio, physical evidence list, etc.; and building a standardized information logic review workflow, using the thinking chain inquiry technology of the large model, and establishing a three-level reasoning review process, including a primary review to verify the correctness of data logical constraints, an intermediate review to compare responsibility determination and legal consistency, and a final review to evaluate the social impact of the report.

[0018] The standardized information logic review workflow specifically includes: using the thinking chain inquiry technology of the large model to establish a three-level reasoning review process, including the primary review to verify the correctness of the data logic constraints (such as the total responsibility must be 100%), the intermediate review to compare the responsibility judgment with the legal consistency, and the final review to evaluate the social impact of the report (if special vehicles are involved, additional marking is required).

[0019] The storage and chain-up mechanism of the full-process data storage and processing module based on blockchain technology includes: building a main chain-sub-chain architecture, the main chain stores the core responsibility identification report hash value, the sub-chain stores the original accident data by region, and realizes cross-chain data verification through Merkle tree; and designs differentiated storage strategies for different types of data, with structured data fully chained and unstructured data stored in IPFS distributed storage, with only file hash values and metadata chained, introducing data sharding technology, and encapsulating single accident-related data into independent data blocks.

[0020] A method for automatically handling traffic accidents based on a large model and blockchain technology, comprising the following steps:

[0021] Construct a traffic accident handling intelligent module system based on a large model, which consists of three intelligent modules: traffic accident analysis intelligent module, traffic accident responsibility confirmation intelligent module, and traffic accident report review intelligent module;

[0022] Construct and enhance a traffic accident analysis intelligent module based on a large model, which analyzes traffic accidents and extracts core feature information;

[0023] Construct and enhance a traffic accident responsibility confirmation intelligent module based on a large model, which performs specific responsibility determination for traffic accidents;

[0024] Building and enhancing a traffic accident report review intelligent module based on a large model, which reviews and confirms the determination of traffic accident responsibility and generates a report after review and confirmation; and

[0025] Build a full-process data storage and processing mechanism based on blockchain technology, and achieve penetrating supervision of the entire disposal process through blockchain technology, ensuring that every link from raw data collection to final responsibility determination is auditable and cannot be tampered with.

[0026] Compared with the prior art, the present invention has the following beneficial effects:

[0027] The present invention proposes a traffic accident handling intelligent module system based on a large model, comprising a traffic accident analysis intelligent module, a traffic accident responsibility confirmation intelligent module, a traffic accident report review intelligent module, and a full-process data storage and processing module based on blockchain technology. The three intelligent modules collaborate with each other by simulating the thinking of different roles in the real-world handling process to jointly complete the automated identification of traffic accidents. The traffic accident analysis intelligent module uses a role-playing mechanism based on the thinking of traffic police, a large model fine-tuning method to enhance accident calculation and analysis capabilities, and an analysis method based on retrieval and enhanced judgment based on a historical accident knowledge base; the traffic accident responsibility confirmation intelligent module determines responsibility by constructing a responsibility determination knowledge graph and a dual-perspective determination mechanism; the traffic accident report review intelligent module conducts review and confirmation through a standardized information integrity and logical review workflow; and blockchain technology enables penetrating supervision of the entire handling process. This achieves the automation of traffic accident handling without relying on manual labor, fundamentally solving the problem of low traffic accident handling efficiency.

[0028] The present invention proposes a method for constructing an intelligent module for traffic accident analysis, which specifically includes: constructing a role-playing mechanism based on the thinking of traffic police to ensure the rationality of the intelligent module in the accident analysis process; based on large-scale model fine-tuning, enhancing the quantitative analysis capability in the accident severity analysis process to ensure the accuracy of the intelligent module in accident analysis; based on a historical accident knowledge base, constructing an analysis method based on retrieval-enhanced judgment to ensure the authenticity of the intelligent module in accident analysis.

[0029] The present invention proposes a method for constructing an intelligent module for determining traffic accident liability, which specifically includes: constructing a knowledge graph based on traffic accident liability determination to enhance the professionalism of the intelligent module in the liability determination process; and constructing a dual-perspective determination mechanism based on "legal experts" and "insurance adjusters" to enhance the comprehensiveness of the intelligent module's analysis of the liability determination process.

[0030] The present invention proposes a method for constructing an intelligent module for traffic accident report review, specifically constructing a standardized information integrity and information logic review workflow to ensure the practicality and compliance of accident reports.

[0031] The present invention proposes a multi-agent module collaborative working mechanism in a traffic accident analysis scenario, which reasonably divides the tasks of the multi-agent module system in this scenario. The three intelligent modules cooperate with each other, improving the fairness and efficiency of traffic accident identification.

[0032] The present invention proposes a full-process data storage and processing mechanism based on blockchain technology, which specifically includes: blockchain-based storage technology, which ensures the traceability of accident original data and the transparency of the intelligent module system processing process; based on the main chain-sub-chain architecture and differentiated storage strategies designed for different types of data, it not only ensures the credibility of the system, but also improves the system's operating efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] 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.

[0034] Figure 1 This is the overall architecture diagram of the automatic traffic accident handling system based on the big model and blockchain technology of the present invention.

[0035] Figure 2 This is a diagram of the composition of the automatic traffic accident handling system based on the big model and blockchain technology of the present invention.

[0036] Figure 3 This is an architectural diagram of the full-process data storage and processing mechanism based on blockchain technology in the present invention.

[0037] Figure 4 This is a flow chart of the method for automatically handling traffic accidents based on big models and blockchain technology of the present invention. DETAILED DESCRIPTION

[0038] 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 creative efforts are within the scope of protection of the present invention.

[0039] like Figure 1 As shown, the automatic traffic accident handling system based on big models and blockchain technology provided by an embodiment of the present invention includes a traffic accident analysis intelligent module 10, a traffic accident responsibility confirmation intelligent module 20, a traffic accident report review intelligent module 30 and a full-process data storage and processing module 40 based on blockchain technology.

[0040] The traffic accident analysis intelligent module 10 is used to analyze traffic accidents and extract core traffic accident feature information. Specifically, the traffic accident analysis intelligent module 10 receives traffic accident video or photo data as input and outputs traffic accident core feature data. The core traffic accident features include vehicle trajectory, pedestrian behavior, and weather data.

[0041] The traffic accident responsibility confirmation intelligent module 20 is used to determine the specific responsibility of the traffic accident. Specifically, the traffic accident responsibility confirmation intelligent module 20 receives the traffic accident characteristic data output by the traffic accident analysis intelligent module 10 as input and outputs a traffic accident responsibility preliminary analysis report.

[0042] The traffic accident report review intelligent module 30 is used to review and confirm the traffic accident responsibility determination and generate a report after the review and confirmation. Specifically, the traffic accident report review intelligent module 30 receives the traffic accident responsibility pre-analysis report output by the traffic accident responsibility confirmation intelligent module 20 as input and outputs a traffic accident report.

[0043] The full-process data storage and processing module 40 based on blockchain technology is used to store accident-related feature data, full-process reasoning data of the intelligent module system, and confirmation report data of the accident responsible party on the chain.

[0044] Among them, the traffic accident analysis intelligent module 10, the traffic accident responsibility confirmation intelligent module 20 and the traffic accident report review intelligent module 30 simulate the thinking mode of different roles in the real-world handling process, cooperate with each other, and jointly complete the automated identification of traffic accidents.

[0045] like Figure 2 As shown, the enhancement method of the traffic accident analysis intelligent module 10 includes:

[0046] Build a role-playing mechanism based on traffic police thinking 101, simulate the traffic police's on-site investigation perspective, and improve the observation ability of the traffic accident analysis intelligent module in traffic accident analysis;

[0047] Constructing a large-scale model fine-tuning method 102 to enhance the accident computation and analysis capabilities, and improving the traffic accident analysis intelligent module's ability to quantitatively analyze accident severity in traffic accident analysis; and

[0048] Construct an analysis method 103 for retrieval-enhanced judgment based on a historical accident knowledge base, provide auxiliary information for the large-model intelligent agent module in traffic accident analysis based on the knowledge base and through a retrieval-enhanced generation method, and improve the decision-making ability of the large-model intelligent agent module to capture key features of traffic accidents.

[0049] Specifically, the role-playing mechanism 101 based on traffic police thinking includes: constructing the role background of the agent module and establishing a reasoning chain for accident investigation. The role background of the agent module is constructed by setting background task instructions to guide the large model to conduct subsequent reasoning analysis using the thinking mode of a traffic police officer. The reasoning chain for accident investigation begins by constructing multiple rounds of reasoning prompts and inquiries, sequentially analyzing vehicle trajectories, pedestrian behavior, and weather data, and then verifying logical contradictions through multiple rounds of reasoning.

[0050] Specifically, the large-scale model fine-tuning method 102 for enhancing the accident calculation and analysis capabilities includes: utilizing momentum conservation, collision angle calculation, and other methods based on traffic accident physical model data to fine-tune the parameters of the traffic accident analysis intelligent module, focusing on optimizing the large-scale model's quantitative calculation capabilities during the collision accident analysis process, and improving the ability to judge the severity of the accident and the collision mode.

[0051] The historical accident knowledge base described in the analysis method 103 for retrieval-enhanced judgment based on a historical accident knowledge base consists of historical traffic accident data, including real-time video or photographic data of traffic accidents, weather data, and descriptions of key accident features. This historical accident knowledge base is searched using a key feature similarity matching method to provide reference information for the analysis method for retrieval-enhanced judgment based on the historical accident knowledge base.

[0052] like Figure 2 As shown, the enhancement method of the traffic accident responsibility confirmation intelligent module 20 includes:

[0053] Constructing a knowledge graph 201 for determining responsibility for traffic accidents, and using the information in the knowledge graph to assist the large model in making decisions; and

[0054] Judgment mechanism based on the dual perspectives of "legal experts" and "insurance adjusters"202.

[0055] The nodes of the traffic accident responsibility determination knowledge graph 201 cover the provisions of the "Road Traffic Safety Law", local regulations, typical precedents (such as full responsibility for lane change, failure to give way at intersections), and the edge relationships define the responsibility association rules (such as the priority right of way logic chain).

[0056] The determination mechanism 202 based on the dual perspectives of "legal experts" and "insurance adjusters" includes: a determination mechanism based on "legal experts" and a determination mechanism based on "insurance adjusters."

[0057] The judgment mechanism based on "legal experts" uses the traffic accident responsibility judgment knowledge graph 201 to judge legal responsibility;

[0058] The "insurance adjuster"-based judgment mechanism uses the analysis results of the "legal expert" judgment mechanism and combines them with insurance clause reasoning and analysis to determine financial liability and assess the compensation ratio. The insurance clauses are provided as contextual input to the large model agent module.

[0059] like Figure 2 As shown, the enhancement method of the traffic accident report review intelligent module 30 includes:

[0060] Establishing a standardized information integrity review workflow 301; and

[0061] Build a standardized information logic review workflow 302.

[0062] Specifically, the standardized information integrity review workflow 301 includes: establishing an information extraction question-and-answer mechanism based on a large model, guiding the traffic accident responsibility confirmation intelligent module to conduct an information integrity review, and the review content includes information on the parties involved, the proportion of responsibility, a list of physical evidence, etc.

[0063] Specifically, the standardized information logic review workflow 302 includes: using the thinking chain inquiry technology of the big model to establish a three-level reasoning review process, including the primary review to verify the correctness of the data logic constraints (such as the total responsibility must be 100%), the intermediate review to compare the responsibility judgment with the legal consistency, and the final review to evaluate the social impact of the report (if special vehicles are involved, additional marking is required).

[0064] like Figure 3 As shown, the storage and chain-up mechanism of the full-process data storage and processing module 40 based on blockchain technology includes:

[0065] Build a main chain-subchain architecture 401, with the main chain storing the core responsibility identification report hash value, and the subchain storing the original accident data by region, and implementing cross-chain data verification through Merkle trees; and

[0066] Differentiated storage strategies 402 are designed for different types of data. Structured data is fully uploaded to the chain, and unstructured data is stored using IPFS distributed storage, with only file hash values and metadata uploaded to the chain. Data sharding technology is introduced, and data related to a single accident is encapsulated into independent data blocks.

[0067] like Figure 4 As shown, an embodiment of the present invention further provides a method for automatically handling traffic accidents based on a large model and blockchain technology, comprising the following steps:

[0068] Step S1: constructing a traffic accident handling intelligent module system based on a large model, wherein the system is composed of three intelligent modules: a traffic accident analysis intelligent module, a traffic accident responsibility confirmation intelligent module, and a traffic accident report review intelligent module;

[0069] Step S2: constructing and enhancing a traffic accident analysis intelligent module based on a large model, wherein the traffic accident analysis intelligent module analyzes traffic accidents and extracts core feature information;

[0070] Step S3: constructing and enhancing a traffic accident responsibility confirmation intelligent module based on a large model, wherein the traffic accident responsibility confirmation intelligent module performs specific responsibility determination for traffic accidents;

[0071] Step S4, constructing and enhancing a traffic accident report review intelligent module based on a large model, wherein the traffic accident report review intelligent module reviews and confirms the traffic accident responsibility determination and generates a report after the review and confirmation; and

[0072] Step S5: Build a full-process data storage and processing mechanism based on blockchain technology, and use blockchain technology to achieve penetrating supervision of the entire disposal process, ensuring that every link from raw data collection to final responsibility determination is auditable and cannot be tampered with.

[0073] In step S2, the method for enhancing the traffic accident analysis intelligent module includes: constructing a role-playing mechanism based on traffic police thinking, simulating the traffic police's on-site investigation perspective, and improving the observation ability of the traffic accident analysis intelligent module in traffic accident analysis;

[0074] Develop a large-scale model fine-tuning method to enhance the accident calculation and analysis capabilities, and improve the traffic accident analysis intelligent module's ability to quantitatively analyze accident severity in traffic accident analysis; and

[0075] Construct an analysis method for retrieval-enhanced judgment based on the historical accident knowledge base. Based on the knowledge base and through the retrieval-enhanced generation method, provide auxiliary information for the large-model intelligent agent module in traffic accident analysis, and improve the decision-making ability of the large-model intelligent agent module to capture key features of traffic accidents.

[0076] In step S3, the method for enhancing the traffic accident responsibility confirmation intelligent module includes:

[0077] Construct a knowledge graph for traffic accident liability determination, and use the information in the knowledge graph to assist the big model in making decisions; and

[0078] A judgment mechanism based on the dual perspectives of "legal experts" and "insurance adjusters".

[0079] In step S4, the method for enhancing the traffic accident report review intelligent module includes:

[0080] Establishing a standardized information integrity review workflow; and

[0081] Build a standardized information logic review workflow.

[0082] Among them, in step S5, the whole process data storage and processing mechanism based on blockchain technology includes: confirming the key data information collected and generated by the system, including accident-related feature data, full-process reasoning data of the intelligent module system, and confirmation report data of the accident responsible party; and

[0083] Confirm the storage and on-chain mechanism of key data, including building a main chain-sub-chain architecture and designing differentiated storage strategies for different types of data.

[0084] An embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which implements the above-mentioned method when the program is executed by a processor.

[0085] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited to the described embodiments. It is apparent to those skilled in the art that various changes, modifications, substitutions, and variations to these embodiments may be made without departing from the principles and spirit of the present invention, and these changes and modifications still fall within the scope of protection of the present invention.

Claims

1. A traffic accident automatic handling system based on big model and blockchain technology, characterized by: include: Traffic accident analysis intelligent module, used to analyze traffic accidents and extract core feature information; Traffic accident responsibility confirmation intelligent module, used to determine the specific responsibility for traffic accidents; Traffic accident report review intelligent module, used to review and confirm the traffic accident responsibility determination and generate a report after review and confirmation; A full-process data storage and processing module based on blockchain technology is used to store accident-related feature data, full-process reasoning data of the intelligent module system, and confirmation report data of the accident responsible party on the chain; Among them, the traffic accident analysis intelligent module, traffic accident responsibility confirmation intelligent module and traffic accident report review intelligent energy body simulate the thinking mode of different roles in the real-world handling process, cooperate with each other, and jointly complete the automated identification of traffic accidents.

2. The automatic traffic accident handling system based on a large model and blockchain technology according to claim 1 is characterized by: The method for enhancing the traffic accident analysis intelligent module includes: Build a role-playing mechanism based on traffic police thinking, simulate the traffic police's on-site investigation perspective, and improve the observation ability of the traffic accident analysis intelligent module in traffic accident analysis; Construct a large-scale model fine-tuning method to enhance the accident calculation and analysis capabilities, and improve the traffic accident analysis intelligent module's ability to quantitatively analyze the severity of accidents in traffic accident analysis; Construct an analysis method for retrieval-enhanced judgment based on the historical accident knowledge base. Based on the knowledge base and through the retrieval-enhanced generation method, provide auxiliary information for the large-model intelligent agent module in traffic accident analysis, and improve the decision-making ability of the large-model intelligent agent module to capture key features of traffic accidents.

3. The automatic traffic accident handling system based on a large model and blockchain technology according to claim 1 is characterized by: The method for enhancing the traffic accident responsibility confirmation intelligent module includes: Construct a knowledge graph for traffic accident responsibility determination, and use the information in the knowledge graph to assist the big model in making decisions. The judgment mechanism is based on the dual perspectives of "legal experts" and "insurance adjusters". The judgment mechanism based on "legal experts" uses the knowledge graph based on traffic accident liability judgment to determine legal liability, while the judgment mechanism based on "insurance adjusters" uses the analysis results of the "legal expert" judgment mechanism and matches it with the insurance clause reasoning analysis to determine economic liability and evaluate the compensation ratio.

4. The automatic traffic accident handling system based on a large model and blockchain technology according to claim 1 is characterized by: The method for enhancing the traffic accident report review intelligent module includes: Build a standardized information integrity review workflow, establish an information extraction question-and-answer mechanism based on a large model, and guide the traffic accident responsibility confirmation intelligent module to conduct information integrity review. The review content includes information on the parties involved, the proportion of responsibility, and the list of physical evidence; Build a standardized information logic review workflow, use the thinking chain query technology of the big model, and establish a three-level reasoning review process, including primary review to verify the correctness of data logical constraints, intermediate review to compare responsibility determination and legal consistency, and final review to evaluate the social impact of the report.

5. The automatic traffic accident handling system based on big model and blockchain technology according to claim 1 is characterized by: The storage and chain-up mechanism of the full-process data storage and processing module based on blockchain technology includes: building a main chain-sub-chain architecture, the main chain stores the hash value of the core responsibility identification report, the sub-chain stores the original accident data by region, and realizes cross-chain data verification through the Merkle tree; and designs differentiated storage strategies for different types of data, with structured data fully chained and unstructured data stored in IPFS distributed storage, file hash values and metadata chained, and data sharding technology introduced to encapsulate single accident-related data into independent data blocks.

6. A method for automatically handling traffic accidents based on a large model and blockchain technology according to claims 1-5, characterized in that: The following steps are involved: Build a traffic accident handling intelligent system based on a large model, which consists of three intelligent modules: traffic accident analysis intelligent module, traffic accident responsibility confirmation intelligent module, and traffic accident report review intelligent module; Construct and enhance a traffic accident analysis intelligent module based on a large model, which analyzes traffic accidents and extracts core feature information; Construct and enhance a traffic accident responsibility confirmation intelligent module based on a large model, which performs specific responsibility determination for traffic accidents; Construct and enhance a traffic accident report review intelligent module based on a large model, which reviews and confirms the determination of traffic accident responsibility and generates a report after review and confirmation; Build a full-process data storage and processing mechanism based on blockchain technology, and achieve penetrating supervision of the entire disposal process through blockchain technology, ensuring that every link from raw data collection to final responsibility determination is auditable and cannot be tampered with.

Images