Method for comprehensively scoring vehicle driving based on multiple factors

Abstract

The invention discloses a method for comprehensively scoring vehicle driving based on multiple factors, and relates to the technical field of vehicle insurance, and the system comprises the following components: a data collection module, a risk scoring model module, a scoring system module, a technical framework support module and a safety guarantee module. By integrating a plurality of data collection modules, including an intelligent sensor and a multi-source data fusion technology, driving records, driving experience, basic information, vehicle types and value data of drivers can be comprehensively collected, and meanwhile, the risk scoring model module applies an improved multi-factor weighted risk assessment algorithm, comprehensively considers a plurality of factors, and improves the risk assessment efficiency. And the psychological and health condition data of the driver is introduced, so that the risk assessment is more accurate and comprehensive, and the insurance company and the traffic management department can identify the high-risk driver and vehicle more accurately, thereby taking corresponding risk management measures.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle insurance, and in particular to a method for comprehensively scoring vehicle driving based on multiple factors. Background Art

[0002] With the rapid development of the transportation industry, the number of vehicles has increased dramatically, and traffic safety issues have become increasingly prominent. In order to effectively manage traffic risks, insurance companies and traffic management departments need to accurately assess the risk levels of drivers and vehicles.

[0003] In traditional technologies, risk assessment of drivers and vehicles mainly relies on limited data sources and simple assessment methods. For example, risk assessment is based only on the driver's driving record or vehicle type, ignoring the impact of multiple factors such as the driver's psychological state, health status, and real-time traffic conditions on risk. In addition, the traditional static threshold division method may not be able to adapt to the actual conditions in different regions and different time periods, resulting in inaccurate and incomplete risk assessment results.

[0004] In summary, traditional technologies have obvious shortcomings in assessing the risk levels of drivers and vehicles, mainly reflected in the limited data sources, simple assessment methods and the limitations of static threshold division methods. Therefore, it is particularly important to develop a method for comprehensive scoring of vehicle driving based on multiple factors. Summary of the Invention

[0005] The purpose of this invention is to make up for the shortcomings of the existing technology and provide a method for comprehensive vehicle driving scoring based on multiple factors. It can integrate multiple data collection modules, use intelligent sensors and multi-source data fusion technology to comprehensively collect driver and vehicle information, and use an improved multi-factor weighted risk assessment algorithm and introduce driver psychological and health status data to make risk assessment more accurate and comprehensive, helping insurance companies and traffic management departments to accurately identify high-risk objects and implement effective risk management.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: a method for comprehensively scoring vehicle driving based on multiple factors, the system comprising the following components: a data collection module, a risk scoring model module, a scoring system module, a technical framework support module, and a safety assurance module;

[0007] The data collection module collects data related to the driver and vehicle from multiple sources, including driving records, driving experience, basic driver information, vehicle type and value;

[0008] The risk scoring model module uses statistical models to analyze and evaluate the collected data and calculate the risk scores of drivers and vehicles;

[0009] The scoring system module converts the calculated risk scores into grades or scores according to the standards and divides them into several categories to represent different risk levels;

[0010] The technical framework support module: uses the .NetCore cross-platform, high-performance open source framework to build the system, utilizes the ABP framework and toolset to implement a modular application infrastructure, uses container technology to package and deploy applications, selects the MySQL open source relational database management system to store structured data, introduces message queues to implement asynchronous communication within the system, uses distributed cache to improve data reading speed, utilizes OSS cloud services to store and retrieve large amounts of unstructured data, builds scalable network applications based on Node.js, and combines the Vue3 progressive JavaScript framework to build modern web applications. AntDesign, a Vue-based UI component library, is used to improve interface design and interactive experience, and adopts a front-end and back-end separation architecture.

[0011] The security assurance module introduces a tenant isolation strategy to ensure that data from different tenants are isolated from each other, and adopts multi-level security measures in architectural design, including access control, encryption, and monitoring.

[0012] Furthermore, the statistical model used by the risk scoring model module is an improved multi-factor weighted risk assessment algorithm, which comprehensively considers multiple factors of the driver and vehicle to obtain a more accurate risk score. The specific algorithm formula is as follows:

[0013]

[0014] Where R is the final risk score, n is the number of factors considered, and w i is the weight of the i-th factor, and x i is the actual value of the i-th factor, f i (x i ) is the risk mapping function of the i-th factor. For the driving record factor, such as the number of accidents x1, its risk mapping function f1(x1) can be designed as:

[0015]

[0016] where x 1max The historical maximum number of accidents. This function maps the number of accidents to the range of 0-100. The greater the number of accidents, the higher the risk mapping value.

[0017] For the driving experience x2, the risk mapping function f2(x2) can be designed as:

[0018]

[0019] where x 2max is the historical maximum value of driving years. Generally speaking, the longer the driving years, the lower the risk. Therefore, the risk mapping value of this function decreases with the increase of driving years.

[0020] For vehicle type and value factors, assuming the vehicle purchase price is x3, its risk mapping function f3(x3) can be designed as:

[0021]

[0022] where x 3avg is the average purchase price of similar vehicles. The higher the vehicle price, the higher the maintenance cost may be, and the risk will also increase accordingly. i The weights are obtained by training the random forest algorithm in machine learning on a large amount of historical insurance data. The random forest algorithm analyzes the data by constructing multiple decision trees. Each decision tree is trained on different data sets and feature subsets, and the weight of each feature is finally determined according to its importance in the decision tree. During the training process, the random forest algorithm will automatically adjust the weight of each factor according to the degree of influence of different factors in the historical data on the occurrence of insurance accidents, making risk assessment more accurate.

[0023] Furthermore, the data collection module uses smart sensors and multi-source data fusion to collect data. A variety of smart sensors are also installed on the vehicle. The accelerometer can monitor the acceleration and deceleration of the vehicle in real time. By analyzing the acceleration data, it can be determined whether the driver's driving style is aggressive. The gyroscope sensor can detect the steering situation of the vehicle. Combined with the vehicle's driving speed, it can be determined whether the driver complies with safety rules when turning. The camera can record the road conditions and the driver's behavior during the vehicle's driving process, and whether there is distracted driving.

[0024] Furthermore, the scoring system module uses a dynamic thresholding method to convert risk scores into grades or scores. Traditional static thresholding methods may not adapt to the actual conditions in different regions and time periods. The dynamic thresholding method of this system dynamically adjusts the threshold based on real-time insurance market data, regional traffic conditions, and seasonal factors.

[0025] Specifically, the system will collect data on insurance accident rates and traffic congestion in different regions. For areas with higher insurance accident rates, the threshold for high-risk levels will be appropriately lowered to more rigorously assess the risks of drivers and vehicles in the area. In some large cities with complex traffic conditions and frequent accidents, the original risk score of 61-100 points will be adjusted to 55-100 points as a high-risk level;

[0026] Seasonal factors can also affect risk assessment. In winter, due to the cold weather and the easy icing of the road surface, the driving risk is relatively high. The system will appropriately increase the thresholds of each risk level in winter, so that more drivers and vehicles are evaluated as having a higher risk level. The range of the medium risk level is adjusted from 31 - 60 points to 25 - 55 points.

[0027] To achieve dynamic threshold division, the system adopts a decision-making algorithm based on fuzzy logic. Fuzzy logic can process information with uncertainty and ambiguity. The system uses insurance market data, regional traffic conditions, and seasonal factors as input variables. Each input variable is divided into different fuzzy sets according to its value range. Then, according to the preset fuzzy rules, new thresholds are obtained through fuzzy reasoning. Finally, the results obtained from fuzzy reasoning are defuzzified to obtain specific threshold values.

[0028] Furthermore, the message queue in the technical framework support module adopts an improved priority message queue algorithm. In a traditional message queue, messages are usually processed in the order of first in first out. However, in this system, different types of messages may have different priorities.

[0029] The improved priority message queue algorithm introduces a message priority classification and dynamic adjustment mechanism. First, messages are divided into different priority categories: high priority, medium priority, and low priority. For high-priority messages, the system will process them first.

[0030] The dynamic adjustment mechanism dynamically adjusts the priority of messages according to the system load situation and the urgency of the messages. When the system load is high, some originally medium-priority messages may be temporarily downgraded in priority to ensure the timely processing of high-priority messages.

[0031] The specific implementation of this algorithm is as follows:

[0032] Let the initial priority of message m be P0(m), the system load factor be L (0 < L < 1, the larger L is, the higher the system load), and the message urgency factor be E (0 < E < 1, the larger E is, the more urgent the message). Then the formula for the dynamic priority P(m) of message m is:

[0033] P(m) = P0(m) × (1 - L) + E × L

[0034] In this way, the system can flexibly adjust the processing order of messages according to the actual situation, improving the performance and response ability of the system.

[0035] Furthermore, the access control in the security module adopts the role-based and attribute-based access control model (RBAC+ABAC). The traditional role-based access control (RBAC) model only considers the user's role, but ignores the user's attributes and environmental factors. This system combines attribute-based access control (ABAC) to control user access to system resources in a more fine-grained manner.

[0036] In the RBAC section, the system defines different roles, such as administrator, insurance agent, and general user. Each role has a different set of permissions. For example, administrators can operate all system functions and data, insurance agents can view and process customers' insurance business information, and general users can only view their own insurance information and assessment results.

[0037] In the ABAC part, the system considers user attributes (such as age, years of work experience, department, etc.), resource attributes (such as data sensitivity level, data creation time, etc.), and environmental factors (such as access time and access location, etc.). For example, for some highly sensitive data, only users who meet specific attribute conditions (such as more than five years of work experience, belonging to a specific department) and under specific environmental conditions (such as being on the company's internal network and during working hours) can access it.

[0038] The specific access control decision algorithm is as follows:

[0039] Assume that the attribute set of user U is A U , the attribute set of resource R is A R , the set of environmental factors is E, the set of access policies is S, and for each access policy s∈S, define a Boolean function f s (A U ,A R ,E),When the function returns true, it means that the access policy allows the user to access the resource. The final access control decision D is:

[0040]

[0041] That is, as long as there is a policy that allows access, the user is allowed to access resources. In this way, the system can achieve more flexible and secure access control.

[0042] Furthermore, the risk scoring model module also considers the driver's psychological factors and health status when calculating the risk score. The driver's psychological state and health status will have a significant impact on driving behavior;

[0043] In order to obtain data on the driver's psychological factors and health status, the system can interact with wearable devices. Wearable devices can monitor the driver's heart rate, blood pressure, and sleep quality physiological indicators in real time. By analyzing these indicators, the driver's psychological state and health status can be judged.

[0044] In the risk assessment algorithm, psychological and health risk factors are introduced. Assuming the driver's heart rate is x4, blood pressure is x5, and sleep quality score is x6, the psychological and health risk mapping function f4(x4,x5,x6) is defined as:

[0045]

[0046] Among them, α, β, and γ are weight coefficients, and α+β+γ=1, x 4max 、x 5max 、x 6max They are the historical maximum values of heart rate, blood pressure, and sleep quality score;

[0047] When calculating the final risk score, psychological and health risk factors are taken into account. The updated risk score calculation formula is:

[0048] R′=(1-w4)×R+w4×f4(x4,x5,x6)

[0049] Where R is the risk score when psychological and health factors are not considered, w4 is the weight of psychological and health factors. The weight w4 is determined by analyzing a large number of accident cases and combining medical research results to determine the degree of influence of psychological and health factors on the occurrence of accidents, thereby determining their weights.

[0050] Furthermore, the distributed cache in the technical framework support module adopts a multi-level cache architecture and an adaptive cache update strategy. The multi-level cache architecture includes local cache and distributed cache servers. The local cache stores the most recently accessed data to reduce the number of visits to the distributed cache server and improve the system's response speed. The distributed cache server stores a large amount of shared data and provides data sharing services for multiple application instances.

[0051] Furthermore, the encryption technology in the security assurance module adopts an encryption scheme based on the combination of elliptic curve cryptography and homomorphic encryption. Elliptic curve cryptography has higher security and smaller key length. At the same security level, the key length required by ECC is much shorter than that of the traditional RSA algorithm, thereby reducing the computational complexity of encryption and decryption and improving system performance.

[0052] Compared with the existing technology, this method of comprehensive scoring of vehicle driving based on multiple factors has the following beneficial effects:

[0053] 1. By integrating multiple data collection modules, including intelligent sensors and multi-source data fusion technology, this system can comprehensively collect drivers' driving records, driving experience, basic information, as well as vehicle type and value data. At the same time, the risk scoring model module uses an improved multi-factor weighted risk assessment algorithm to comprehensively consider multiple factors and introduce driver psychological and health status data, making risk assessment more accurate and comprehensive. This helps insurance companies and traffic management departments more accurately identify high-risk drivers and vehicles, so that they can take appropriate risk management measures.

[0054] 2. This system adopts a dynamic threshold division method to dynamically adjust the risk assessment threshold based on real-time insurance market data, regional traffic conditions and seasonal factors. This dynamic adjustment mechanism enables the system to better adapt to the actual conditions in different regions and time periods, and improve the pertinence and effectiveness of risk assessment. In addition, the technical framework support module adopts a high-performance open source framework and advanced technical means, such as message queues and distributed cache, which enhances the scalability and flexibility of the system, allowing the system to be customized and optimized according to different needs to meet application requirements in different scenarios.

[0055] Other advantages, objects and features of the present invention will be described in part in the following description and, in part, will be apparent to those skilled in the art based on an examination of the following or may be learned from the practice of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] 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. Those skilled in the art can also derive other drawings based on these drawings without inventive effort.

[0057] Figure 1 A flowchart for the function implementation of a method for comprehensive scoring of vehicle driving based on multiple factors;

[0058] Figure 2 The figure is a flowchart of the overall architecture of a method for comprehensive scoring of vehicle driving based on multiple factors. DETAILED DESCRIPTION

[0059] In order to further illustrate the technical means and effects adopted by the present invention to achieve the predetermined purpose of the invention, the specific implementation methods, structures, features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments.

[0060] Example 1

[0061] This embodiment describes an insurance company's introduction of a comprehensive assessment system to more accurately assess the risk of auto insurance policyholders and thus rationally set premiums. The insurance company needs to analyze a large amount of driver and vehicle data to determine the risk level of each policyholder.

[0062] The data collection module collects data through intelligent sensors and multi-source data fusion. Accelerometers, gyroscopes, and cameras are installed on the insured's vehicle. The accelerometer detected that Driver A frequently accelerated and decelerated suddenly within a month, an average of five times a day. The gyroscope showed that he exceeded the safe speed multiple times when turning. The camera recorded that Driver A had distracted driving behavior, such as looking at his mobile phone while driving, which occurred three times within a week. At the same time, the basic information of Driver A, including his five-year driving experience, and vehicle information were collected. The vehicle was purchased for 200,000 yuan and was a mid-size SUV.

[0063] The risk scoring model module uses an improved multi-factor weighted risk assessment algorithm, assuming that the four factors (n=4) of driving record, driving years, vehicle purchase price, psychological and health factors are considered. Assume that after training, the weights of each factor are: driving record weight w1=0.4, driving years weight w2=0.2, vehicle purchase price weight w3=0.2, psychological and health factor weight w4=0.2. In terms of driving record, assuming that the historical maximum number of accidents in the area is x1max=10 times, and the number of accidents of driver A is x1=2 times, according to the risk mapping function but In terms of driving experience, assuming that the historical maximum driving experience in the region is x 2max = 30 years, driver A’s driving experience x2 = 5 years, according to the risk mapping function but In terms of vehicle purchase price, assuming that the average purchase price of similar medium-sized SUV vehicles is x3avg = 150,000 yuan, and the purchase price of driver A's vehicle is xs = 200,000 yuan, according to the risk mapping function but

[0064] In terms of psychological and health factors, it is assumed that driver A's wearable device monitors heart rate x4 = 80 beats / minute, blood pressure x5 = 120 / 80 mmHg (systolic pressure 120, diastolic pressure 80, here calculated based on systolic pressure), sleep quality score x6 = 7 points (out of 10 points), historical maximum heart rate x4max = 100 beats / minute, historical maximum blood pressure x 5max =140mmHg, the historical maximum value of sleep quality score x 6max =10 points, weight coefficients α=0.4, β=0.3, γ=0.3.

[0065] Mapping functions based on psychological and health risks but Risk score without considering psychological and health factors. Risk score after considering psychological and health factors.

[0066] The scoring system module uses a dynamic threshold division method to convert risk scores into levels based on real-time insurance market data, regional traffic conditions and seasonal factors. Assuming the current market and regional conditions, risk scores of 0-30 are low risk, 31-60 are medium risk, and 61-100 are high risk. Driver A's risk score of 31.068 is classified as medium risk.

[0067] The technical framework support module adopts the .NetCore cross-platform, high-performance open source framework to build the system, uses the ABP framework and tool set to implement a modular application infrastructure, uses container technology to package and deploy applications, selects the MySQL open source relational database management system to store structured data, introduces message queues to implement asynchronous communication within the system, uses distributed cache to improve data reading speed, uses OSS cloud services to store and retrieve large amounts of unstructured data, builds scalable network applications based on Node.js, combines the Vue3 progressive JavaScript framework to build modern Web applications, uses the AntDesign Vue-based UI component library to improve interface design and interactive experience, adopts a front-end and back-end separation architecture, and introduces a tenant isolation strategy in the security assurance module to ensure that data from different insurance companies are isolated from each other. Access control adopts a role-based and attribute-based access control model (RBAC+ABAC), and encryption technology adopts an encryption scheme based on elliptic curve cryptography and homomorphic encryption.

[0068] Example 2

[0069] This example describes a large logistics fleet responsible for nationwide cargo transportation. The fleet owns hundreds of various vehicles and numerous drivers with diverse experience and driving styles. As the business scale continues to expand, ensuring transportation safety and reducing operating costs have become top priorities for fleet management. To achieve refined management, the fleet has introduced a comprehensive vehicle driving scoring method based on multiple factors. By assessing the risks of both drivers and vehicles, a scientific and reasonable management strategy is formulated.

[0070] The data collection module collects multi-dimensional data through intelligent sensors. On the vehicle, the accelerometer is like a sensitive "sensory antenna", recording in detail every acceleration and deceleration action of Driver B. In the transportation missions in the past week, the sensor showed that Driver B accelerated and decelerated sharply a total of 20 times. This data shows that his driving style is relatively aggressive, which may increase vehicle damage and the risk of accidents. The gyroscope sensor closely monitors the steering of the vehicle. Combined with the vehicle's driving speed, it was found that Driver B failed to slow down as required when turning many times. This dangerous driving behavior seriously threatens transportation safety. At the same time, the camera installed in the car also played an important role. It clearly recorded that Driver B experienced fatigue driving twice in a week, such as closing his eyes for a long time and yawning frequently. This is undoubtedly a huge hidden danger on the transportation road.

[0071] In addition to vehicle driving data, the personal information of Driver B was also collected. He has been driving for 8 years and is an experienced driver in the industry. However, judging from recent driving data, there are still major problems with his driving habits. The vehicle he drives is a heavy truck, which was purchased for 350,000 yuan. Such vehicles usually carry a large amount of cargo. Once an accident occurs, the loss will be very heavy. In addition, the fleet equipped the driver with a wearable device, which monitors Driver B's heart rate, blood pressure, sleep quality and other health data. The data showed that his heart rate was 85 beats / minute, blood pressure was 130 / 85 mmHg, and his sleep quality score was only 6 points. These data reflect that Driver B's physical condition may affect his driving state and need to be taken seriously.

[0072] The risk scoring model module uses an improved multi-factor weighted risk assessment algorithm to comprehensively consider various factors of the driver and vehicle. In terms of driving records, due to frequent sudden acceleration and deceleration, not slowing down when turning, and fatigue driving, it shows that Driver B has a greater safety risk during driving. This factor has a greater impact on the risk score. Although he has 8 years of driving experience, his bad driving habits have prevented the advantage of experience from being fully reflected. However, compared with novice drivers, he still has a certain stability, so the impact on the risk score is moderate. The high purchase price of the vehicle means high maintenance costs and potential losses, which also account for a certain proportion in the risk score. Driver B's psychological and health data show that his physical condition may affect his driving performance, which is also an important factor included in the risk assessment. After comprehensive evaluation, it is concluded that Driver B's risk score is at a relatively high level.

[0073] The scoring system module adopts a dynamic threshold classification method based on the characteristics of the logistics industry and the actual situation of the fleet. Taking into account the characteristics of logistics transportation in different seasons (such as icy roads in winter, heavy rains in summer and other severe weather conditions that increase transportation risks), traffic conditions in different regions (such as narrow roads and heavy traffic in some areas, which are prone to congestion and accidents) and real-time market demand changes (such as heavy transportation tasks in peak seasons, which place higher demands on drivers and vehicles), the system divides risk scores into different levels. In the current assessment, Driver B's risk score is classified as medium risk. Although it has not yet reached the high risk level, it is close to the critical value, which has attracted the attention of the fleet management.

[0074] Based on the risk assessment results, the fleet took a series of targeted management measures. For Driver B, a professional safe driving training course was arranged, focusing on strengthening safe driving awareness and correcting bad driving habits, such as teaching him how to accelerate and decelerate smoothly, slow down in advance when turning, and pay attention to road conditions. At the same time, he was required to arrange rest time reasonably and ensure adequate sleep to improve his physical condition and reduce the risk of fatigue driving. In terms of vehicle scheduling, the frequency of his long-distance transportation tasks was temporarily reduced, and priority was given to some transportation routes with shorter distances and relatively good road conditions to reduce risks during transportation. In addition, the fleet also strengthened the daily inspection and maintenance of the vehicle, increased the frequency of inspections, ensured that the vehicle was always in good operating condition, and reduced the possibility of accidents due to vehicle failure.

[0075] The technical framework module uses the cross-platform, high-performance open source .NetCore framework to build the system, ensuring stable operation on different operating systems and meeting the fleet's diverse business needs. The ABP framework and toolset implement a modular application infrastructure, making system development, maintenance, and expansion more convenient and efficient. Container technology is used for application packaging and deployment, improving deployment flexibility and efficiency while facilitating system management and monitoring. The MySQL open source relational database management system is used to store structured data, ensuring data security and consistency. Message queues are introduced to implement asynchronous communication within the system, improving system responsiveness and processing capabilities, particularly when processing large amounts of data and concurrent requests. Distributed caching improves data access speed, reduces query time, and enhances overall system performance. OSS cloud services are used to store and retrieve large amounts of unstructured data, such as vehicle surveillance videos and driver training materials, facilitating data storage and management. Scalable network applications are built using Node.js, combined with the Vue3 progressive JavaScript framework to build modern web applications. AntDesign, a Vue-based UI component library, enhances interface design and interactive experience, providing fleet managers with a concise, intuitive, and easy-to-use interface. The separation of front-end and back-end architecture improves system security and maintainability.

[0076] The security module introduces a tenant isolation strategy to ensure that data from different fleets are isolated from each other to prevent data leakage and cross-interference. Access control adopts a role-based and attribute-based access control model (RBAC+ABAC). Fine-grained access control is performed based on the employee's role (such as fleet manager, dispatcher, driver) and attributes (such as years of work, department), as well as resource attributes (such as data sensitivity level, vehicle information confidentiality) and environmental factors (such as access time and access location). For example, the fleet manager can operate all functions and data of the system, the dispatcher can only view and process information related to vehicle scheduling, and the driver can only view his own tasks and vehicle status information. The encryption technology uses elliptic curve cryptography. The encryption scheme combined with homomorphic encryption encrypts the transmitted and stored data to ensure data security and privacy. The message queue adopts an improved priority message queue algorithm to dynamically adjust the priority of the message according to factors such as the urgency of the transportation task and the severity of the vehicle fault information to ensure that important information can be processed in a timely manner. The distributed cache adopts a multi-level cache architecture and an adaptive cache update strategy to store the most recently accessed data in the local cache, reduce the number of visits to the distributed cache server, and improve the data reading speed. At the same time, the cache is automatically updated according to the changes in the data to ensure data consistency. Through these technical frameworks and security measures, solid technical support and security guarantees are provided for the risk assessment and management of the fleet.

[0077] The above description is merely a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Although the present invention has been disclosed as above in terms of a preferred embodiment, it is not intended to limit the present invention. Any person skilled in the art can, without departing from the scope of the technical solution of the present invention, make some changes or modifications to equivalent embodiments using the technical contents disclosed above. However, any brief modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. A method for comprehensively scoring vehicle driving based on multiple factors, characterized in that: The system includes the following components: data collection module, risk scoring model module, scoring system module and technical framework support module, as well as security assurance module; The data collection module collects data related to the driver and vehicle from multiple sources, including driving records, driving experience, basic driver information, vehicle type and value; The risk scoring model module uses statistical models to analyze and evaluate the collected data and calculate the risk scores of drivers and vehicles; The scoring system module converts the calculated risk scores into grades or scores according to the standards and divides them into several categories to represent different risk levels; The technical framework support module: uses the .NetCore cross-platform, high-performance open source framework to build the system, utilizes the ABP framework and toolset to implement a modular application infrastructure, uses container technology to package and deploy applications, selects the MySQL open source relational database management system to store structured data, introduces message queues to implement asynchronous communication within the system, uses distributed cache to improve data reading speed, utilizes OSS cloud services to store and retrieve large amounts of unstructured data, builds scalable network applications based on Node.js, and combines the Vue3 progressive JavaScript framework to build modern web applications. AntDesign, a Vue-based UI component library, is used to improve interface design and interactive experience, and adopts a front-end and back-end separation architecture. The security assurance module introduces a tenant isolation strategy to ensure that data from different tenants are isolated from each other, and adopts multi-level security measures in architectural design, including access control, encryption, and monitoring.

2. The method for comprehensively scoring vehicle driving based on multiple factors according to claim 1, characterized in that: The statistical model used in the risk scoring model module is an improved multi-factor weighted risk assessment algorithm, which comprehensively considers multiple factors of the driver and the vehicle. The specific algorithm formula is as follows: Where R is the final risk score, n is the number of factors considered, and w i is the weight of the i-th factor, and x i is the actual value of the i-th factor, f i (x i ) is the risk mapping function of the i-th factor. For the driving record factor, such as the number of accidents x1, its risk mapping function f1(x1) can be designed as: where x 1max This is the highest number of accidents in history; For the driving experience x2, the risk mapping function f2(x2) can be designed as: where x 2max The historical maximum number of years of driving experience; For vehicle type and value factors, assuming the vehicle purchase price is x3, its risk mapping function f3(x3) can be designed as: where x 3avg is the average purchase price of similar vehicles. The higher the vehicle price, the higher the maintenance cost may be, and the risk will also increase accordingly. i During the training process, the random forest algorithm will automatically adjust the weight of each factor according to the impact of different factors on the occurrence of insurance accidents in historical data.

3. The method for comprehensively scoring vehicle driving based on multiple factors according to claim 1, characterized in that: The data collection module uses smart sensors and multi-source data fusion to collect data. A variety of smart sensors are also installed on the vehicle. The acceleration sensor can monitor the acceleration and deceleration of the vehicle in real time. By analyzing the acceleration data, it can be determined whether the driver's driving style is aggressive. The gyroscope sensor can detect the steering situation of the vehicle. Combined with the vehicle's driving speed, it can be determined whether the driver complies with safety rules when turning. The camera can record the road conditions and the driver's behavior during the vehicle's driving process, and whether there is distracted driving.

4. The method for comprehensively scoring vehicle driving based on multiple factors according to claim 1, characterized in that: The scoring system module adopts a dynamic threshold division method when converting risk scores into grades or scores. The traditional static threshold division method may not be able to adapt to the actual conditions in different regions and different time periods. The dynamic threshold division method of this system dynamically adjusts the threshold based on real-time insurance market data, regional traffic conditions and seasonal factors.

5. The method for comprehensively scoring vehicle driving based on multiple factors according to claim 1, characterized in that: The message queue in the technical framework support module adopts an improved priority message queue algorithm. In traditional message queues, messages are usually processed in a first-in-first-out order, but in this system, different types of messages may have different priorities. The improved priority message queue algorithm introduces a message priority classification and dynamic adjustment mechanism. First, messages are divided into different priority categories: high priority, medium priority, and low priority. The system will give priority to high-priority messages. The dynamic adjustment mechanism dynamically adjusts the priority of messages based on the system load and the urgency of the message. When the system load is high, some messages that were originally medium priority may be temporarily lowered in priority. The specific implementation of the algorithm is as follows: Assume that the initial priority of message m is P0(m), the system load factor is L, and the urgency factor of the message is E. The dynamic priority P(m) of message m is calculated as follows: P(m)=P0(m)×(1-L)+E×L In this way, the system can flexibly adjust the message processing order according to actual conditions, thereby improving the system's performance and responsiveness.

6. The method for comprehensively scoring vehicle driving based on multiple factors according to claim 1, characterized in that: The access control in the security module adopts a role-based and attribute-based access control model. The traditional role-based access control model only considers the user's role, but ignores the user's attributes and environmental factors. This system combines attribute-based access control to control user access to system resources in a more fine-grained manner. In the RBAC section, the system defines different roles, such as administrator, insurance agent, and general user. Each role has a different set of permissions. For example, administrators can operate all system functions and data, insurance agents can view and process customers' insurance business information, and general users can only view their own insurance information and assessment results. In the ABAC part, the system takes into account user attributes, resource attributes, and environmental factors. For some highly sensitive data, only users who meet specific attribute conditions and under specific environmental conditions can access it. The specific access control decision algorithm is as follows: Assume that the attribute set of user U is A U , the attribute set of resource R is A R , the set of environmental factors is E, the set of access policies is S, and for each access policy s∈S, define a Boolean function f s (A U ,A R ,E),When the function returns true, it means that the access policy allows the user to access the resource. The final access control decision D is: That is, as long as there is a policy allowing access, the user is allowed to access the resource.

7. The method for comprehensively scoring vehicle driving based on multiple factors according to claim 1, characterized in that: The risk scoring model module also considers the driver's psychological factors and health status when calculating the risk score. The driver's psychological state and health status have a significant impact on driving behavior; In order to obtain data on the driver's psychological factors and health status, the system can interact with wearable devices. Wearable devices can monitor the driver's heart rate, blood pressure, and sleep quality physiological indicators in real time. By analyzing these indicators, the driver's psychological state and health status can be judged. In the risk assessment algorithm, psychological and health risk factors are introduced. Assuming the driver's heart rate is x4, blood pressure is x5, and sleep quality score is x6, the psychological and health risk mapping function f4(x4,x5,x6) is defined as: Among them, α, β, and γ are weight coefficients, and α+β+γ=1, x 4max 、x 5max 、x 6max They are the historical maximum values of heart rate, blood pressure, and sleep quality score; When calculating the final risk score, psychological and health risk factors are taken into account. The updated risk score calculation formula is: R′=(1-w4)×R+w4×f4(x4,x5,x6) Where R is the risk score without considering psychological and health factors, and w4 is the weight of psychological and health factors.

8. The method for comprehensively scoring vehicle driving based on multiple factors according to claim 1, characterized in that: The distributed cache in the technical framework support module adopts a multi-level cache architecture and an adaptive cache update strategy. The multi-level cache architecture includes a local cache and a distributed cache server. The local cache stores the most recently accessed data to reduce the number of visits to the distributed cache server. The distributed cache server stores a large amount of shared data and provides data sharing services for multiple application instances.

9. The method for comprehensively scoring vehicle driving based on multiple factors according to claim 1, characterized in that: The encryption technology in the security assurance module adopts an encryption scheme based on the combination of elliptic curve cryptography and homomorphic encryption. Elliptic curve cryptography has high security and a small key length. At the same security level, the key length required by ECC is much shorter than that of the traditional RSA algorithm, thereby reducing the computational complexity of encryption and decryption and improving system performance.

Images