Wireless Vehicle Data Interaction System

Abstract

The present invention discloses a vehicle data wireless interaction system, specifically related to the field of wireless interaction technology, including: a node relationship model building module, which constructs a node model by analyzing the data flow and logical relationship among vehicle terminals, network-side nodes, cloud servers, and data servers; a multi-source data collection module, which collects data from each node, forms a multi-source data set and performs cleaning; an initial abnormal node extraction module, which screens out abnormal nodes by monitoring key indicators and calculating node abnormal indexes; an abnormal diagnosis module, which locates the cause of the fault by analyzing the abnormal feature correlation matrix; and further includes a diffusion risk assessment module, which calculates the abnormal diffusion risk coefficient and generates a network risk report by analyzing the risk factors and propagation paths of nodes, providing decision support for network management, and having an adaptive optimization mechanism, continuously optimizing the risk assessment and decision tree models according to the disposal effect, and improving the reliability and security of the system.

Description

Technical Field

[0001] The present invention relates to the field of wireless interaction technologies, and more specifically, to a vehicle data wireless interaction system. Background Art

[0002] With the popularization of vehicle networking technologies and the continuous improvement of autonomous driving functions, vehicles need to perform a large amount of data interaction with the external environment (including other vehicles, roadside units, and cloud platforms) during driving. These interaction data not only cover the operating parameters of the vehicle itself (such as speed, steering angle, battery status, etc.) and environmental perception information (such as the detection results of radars and cameras), but also include secure communication data and real-time collaboration instructions between the vehicle and the cloud. The in-vehicle terminal and the cloud achieve multi-directional transmission and fusion analysis through a wireless network, aiming to improve the driving safety and collaborative decision-making efficiency of the vehicle. However, in the face of the dynamic characteristics of the vehicle networking system and the massive real-time data stream, how to ensure accurate, efficient, and traceable information interaction between nodes has become an important research topic in the field of intelligent transportation.

[0003] In the process of vehicle wireless data interaction in the prior art, simple threshold monitoring is often only performed on a single node or a local link, and there is a lack of comprehensive consideration of the correlation between multiple nodes and multiple processes and the potential risk propagation mechanism. For example, when an abnormality occurs in the vehicle terminal encryption module or the cloud parsing service, it is difficult to determine in a timely manner whether it will have a chain impact on subsequent links (such as the fleet scheduling module or the assisted driving decision-making system). In addition, there are differences in the data formats and time bases of different nodes, resulting in the inability to achieve high-precision alignment of data between cross-nodes; when performing anomaly detection or fault troubleshooting, it often relies on manual analysis or decentralized logs, and it is difficult to form an adaptive and scalable linkage warning and closed-loop disposal capability. Summary of the Invention

[0004] To overcome the above-mentioned defects of the prior art, the present invention provides a vehicle data wireless interaction system to solve the problems raised in the above background art.

[0005] To achieve the above object, the present invention provides the following technical solution: A vehicle data wireless interaction system, including:

[0006] A node relationship model building module, configured to build a node relationship model for vehicle data interaction. The node types include vehicle terminals, network-side nodes, clouds, and data servers. Model each node of the vehicle data interaction, clarify the functions and data interaction modes of the vehicle terminal, roadside unit, cloud, and data server, and construct a node relationship model by analyzing the data flow and logical relationship between each node;

[0007] A multi-source data collection module collects data from each node to form a multi-source data set, and extracts the collected data to obtain the key indicators of each node;

[0008] An initial abnormal node extraction module monitors the key indicators of each node, inputs them into a node abnormality measurement function to calculate the node abnormality index of each node, determines whether the node abnormality index meets the threshold requirements, and filters to obtain an initial abnormal node set;

[0009] An abnormality diagnosis module extracts an abnormal feature correlation matrix from the initial abnormal node set. The abnormal feature correlation matrix includes the time feature, spatial position feature, and upstream and downstream dependency features of the initial abnormal node; performs fault diagnosis based on the abnormal feature correlation matrix; and takes measures based on the fault diagnosis result.

[0010] Preferably, the key indicators include common indicators and node function indicators. The common indicators at least include: data throughput, processing delay, resource occupancy rate, and log abnormality rate; the node function indicators reflect the performance of the node category. The node function indicators of the vehicle terminal node at least include: data packet packing accuracy rate, sensor data accuracy rate, sensor data integrity rate, and encryption success rate; the node function indicators of the network side node at least include: communication delay, packet loss rate, encryption / decryption pass rate; the node function indicators of the cloud node at least include: unpacking accuracy rate, parsing time consumption, data storage delay, and response time; the node function indicators of the data service node at least include: service processing accuracy rate, interface call success rate, and abnormal detection coverage rate.

[0011] Preferably, the node abnormality measurement function satisfies the following formula:

[0012] ;

[0013] ;

[0014] ;

[0015] Among them, m represents the total number of key indicators of the node, s represents the serial number of the key indicator; x s represents the actual value of the key indicator s; gw s represents the weight of the key indicator s; is used to quantify the abnormality degree of the key indicator x s relative to the threshold , α represents the adjustment coefficient; represents the abnormal deviation function, indicating the deviation degree between the current value x s and the average value μ s , represents the fluctuation coefficient of the key indicator s.

[0016] Preferably, the abnormal feature correlation matrix includes initial abnormal node time features, spatial location features, and upstream and downstream dependency features;

[0017] The time feature is obtained by collecting the timestamp information of the initial abnormal node, including the time of first occurrence of the abnormality, the duration, and the frequency of the abnormality occurrence;

[0018] The spatial location feature is obtained according to the node relationship model; if it is a physical node, the spatio-temporal location information of the node is obtained using geographical coordinates and position sensor data; if it is a data node, virtual space positioning is performed according to the server and network partition information where the node is located; the upstream and downstream dependency features include: data flow features, business process features, and interface call dependency features;

[0019] The data flow feature is obtained by starting from the initial abnormal node, tracing the data flow to the downstream nodes it connects, and collecting the abnormal conditions of the key indicators of the downstream nodes; or tracing the dependency source nodes of the initial abnormal node and evaluating the impact of upstream abnormalities on the current abnormality;

[0020] The business process feature is obtained by starting from the initial abnormal node, judging its position in the business process, and the impact on its upstream and downstream steps;

[0021] The interface call dependency feature is obtained by recording the interface call behavior of the initial abnormal node and evaluating the impact of the initial abnormal node on the interface.

[0022] Preferably, the fault diagnosis refers to establishing a mapping relationship between the abnormal node type and the abnormal feature correlation matrix based on the decision tree algorithm, constructing a decision tree model by inducing the time features, spatial location features, and upstream and downstream dependency features of the abnormal nodes in the historical data or simulation-annotated samples, and identifying the abnormal type and locating the root cause of the fault through the decision tree model.

[0023] Preferably, the system further includes:

[0024] A diffusion risk assessment module that performs correlation analysis on the propagation path of the initial abnormal node in the node relationship model and outputs the abnormal diffusion path and the abnormal diffusion risk coefficient;

[0025] A diffusion risk judgment and management module. If the abnormal diffusion risk coefficient exceeds the threshold, indicating a high probability of cascading abnormalities, it triggers linkage warnings and protection actions based on predefined linkage strategies, including switching to an alternative communication channel, enabling a redundant parsing module, and notifying the scheduling platform to pay attention to potential risks.

[0026] Preferably, the abnormal diffusion path is obtained by:

[0027] Take the initial abnormal node as the starting point of the path, and use the breadth-first search algorithm to traverse the node relationship model. Starting from the starting point, visit the neighbor nodes that meet the upstream and downstream logical relationships layer by layer. Each time during traversal, judge whether the new node meets the propagation probability threshold. If it meets the requirements, add the new node to the diffusion path and continue to traverse the lower layer;

[0028] When the set traversal times are reached, output the abnormal diffusion path.

[0029] Preferably, the acquisition method of the abnormal diffusion risk coefficient is as follows:

[0030] Calculate the initial risk index: Analyze the set of initial abnormal nodes, and obtain the risk factors of each initial abnormal node. The risk factors at least include: the information entropy of the node abnormal feature correlation matrix, the proportion of the influence of the node vehicle, the proportion of the influence of the node traffic, the node hardware failure rate, the node software failure rate, and the node protection level; perform weighted summation on the quantified risk factors and the corresponding weights, and output the initial risk index of each initial abnormal node;

[0031] Record the product of the probabilities of all connected nodes as the propagation probability of the abnormal diffusion path;

[0032] Based on the data transmission frequency and the security protection level, set a channel weight for each abnormal diffusion path;

[0033] Jointly analyze the propagation probability and the channel weight, and calculate the path risk index of each abnormal diffusion path;

[0034] Add up the initial risk indexes of all initial abnormal nodes to obtain the total node risk; add up the path risk indexes of all abnormal diffusion paths to obtain the total path risk; perform weighted summation on the total node risk and the path risk to calculate the abnormal diffusion risk coefficient;

[0035] Based on the abnormal diffusion risk coefficient, conduct a hierarchical assessment of the overall network risk level and generate a network risk report;

[0036] The network risk report includes: the risk level delimited according to the abnormal diffusion risk coefficient; the key information of the abnormal diffusion path, the length of the abnormal diffusion path, and the propagation probability.

[0037] Preferably, the acquisition method of the initial risk index is as follows:

[0038] Suppose there are Q risk factors, use p to represent the serial number of the risk factor, and record the value of the pth risk factor as f p ; The initial risk index Cf is calculated through the following formula: The initial risk index is calculated through the following formula:

[0039] ;

[0040] Among them, represents the weight of the p-th risk factor, which is set based on experience;

[0041] Let the starting point of each abnormal diffusion path be node i and the ending point be node j; the method for obtaining the channel weight is as follows:

[0042]

[0043] Among them, w ij represents the channel risk weight from node to ; F ij represents the data transmission frequency from node to , and the value range is from 0 to 1. The larger the value, the more frequent the data interaction; S ij represents the average security protection level from node to , and the range is [0, 1]. The larger the value, the higher the security protection ability.

[0044] Preferably, the system further includes:

[0045] A disposal effect adaptive optimization module, which tracks the effects of the linkage early warning and protection actions triggered by the diffusion risk assessment module, and continuously tracks the operating status and alarm indicators of the initial abnormal node; if the abnormality is alleviated or gradually disappears, the current abnormality and the disposal process are recorded in the case library for improving the risk model and policy thresholds; if the disposal effect does not meet the standard or new risks are derived, the latest initial abnormal node is re-incorporated into the abnormal diagnosis module, so as to continuously iterate and optimize the decision tree algorithm through the adaptive correction mechanism to ensure the reliability and security of the wireless vehicle data interaction process.

[0046] The technical effects and advantages of the present invention:

[0047] 1. The wireless vehicle data interaction system provided by the present invention realizes efficient data interaction and fault diagnosis between the vehicle and the cloud, network side, and data server side by constructing a node relationship model for vehicle data interaction and adopting multi-source data collection and abnormal node extraction technologies. The monitoring of key indicators of the receiving node and the calculation of the abnormal metric function can identify potential abnormal nodes in the system in real time, and perform fault diagnosis through the abnormal feature correlation matrix, solving the problem that the root cause of the fault cannot be accurately located in the traditional vehicle data management system. The system effectively improves the stability and reliability of vehicle data interaction, ensuring the vehicle use experience of the vehicle owner in various complex environments.

[0048] 2. The vehicle data wireless interaction system of the present invention realizes comprehensive early warning and protection for system anomalies through the introduction of a diffusion risk assessment and management module and based on the quantitative analysis of abnormal diffusion paths and risk coefficients. The system can give early warning of possible cascading failures according to the risk characteristics of nodes and the propagation probability of abnormal diffusion paths, and trigger redundant modules and standby communication paths through linkage strategies to ensure the continuous and stable operation of the vehicle data interaction system in case of failures. This technology effectively solves the problem that traditional systems lack the ability to predict and handle abnormal propagation, and greatly improves the fault tolerance and emergency response speed of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a block diagram of the structure of the vehicle data wireless interaction system of the present invention.

[0050] Figure 2 It is a block diagram of the structure for obtaining the abnormal diffusion risk coefficient of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0051] Hereinafter, exemplary embodiments of the present disclosure will be described in more detail with reference to the drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be fully conveyed to those skilled in the art.

[0052] At the same time, it should be understood that for the convenience of description, the dimensions of the various parts shown in the drawings are not drawn according to the actual proportional relationship.

[0053] The description of at least one exemplary embodiment below is actually only illustrative and in no way limits the present application and its application or use.

[0054] Technologies, methods, and devices known to those of ordinary skill in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and devices should be regarded as part of the specification.

[0055] BACKGROUND ART: In the embodiments of the present invention, a node can be understood as a functional entity or module divided by logical functions or physical deployments during the processes of data acquisition, transmission, processing, storage, analysis, etc.; a node refers to an entity or functional module that undertakes different functions and performs different operations in the entire process of data interaction; each node can be a physical device or a software / service process, as long as it plays an independent and recognizable functional role in the data link, it can be regarded as a node; the common node types and typical functions are as follows:

[0056] Vehicle terminal (vehicle side): Function examples: data collection, preprocessing, packaging, encryption, etc.;

[0057] Data types: Vehicle sensor outputs (such as speed, position, fuel consumption, vehicle status information, etc.), packaged logs, encrypted operation logs.

[0058] Roadside unit: Function examples: Perform short-range wireless communication with the vehicle terminal (such as V2X communication), assist in data relaying or preliminary processing; Data types: Relay communication logs, forwarding status information, network connection or fault information, etc.

[0059] Network-side node: Function examples: Intermediate nodes during data transmission in the public network / special network (such as base stations, gateways, encryption / decryption units), perform functions such as encryption, routing, or forwarding; Data types: Encryption / decryption success rate, communication delay, packet loss information, network load, etc.

[0060] Terminal platform; Function examples: Receive data transmitted from the vehicle side or network side, perform unpacking, parsing, storage, analysis, etc.; Provide interfaces for external services or data analysis modules; Data types: Unpacking and parsing results, parsing time-consuming, database write logs, response logs.

[0061] Other data service modules: Function examples: Modules that provide higher-level data mining, visualization, or external services outside or above the cloud platform; Data types: Business processing logs, data call interface logs, exception handling records, etc.

[0062] Example 1, refer to Figure 1 the structural block diagram of the vehicle data wireless interaction system in

[0063] Node relationship model building module, used to build a node relationship model for vehicle data interaction. The node types include vehicle terminals, network-side nodes (such as roadside units, transmission units), the cloud, and data service ends. Model each node of vehicle data interaction, clarify the functions and data interaction modes of vehicle terminals, roadside units, the cloud, and data service ends, and construct a node relationship model by analyzing the data flow and logical relationships between each node;

[0064] Explanation: The node relationship model includes: Node function identifiers used to describe node functions, and the node functions at least include vehicle terminal preprocessing, network encryption, and cloud unpacking; Data interaction channels used to record data transmission paths and connection methods between nodes; Upstream and downstream logical relationships used to describe the sequence and logical association of each node in the data interaction process;

[0065] The multi-source data collection module collects data from each node to form a multi-source data set, extracts the collected data to obtain the key indicators of each node; performs cleaning and normalization processing on the multi-source data set, including timestamp alignment, field formatting, and missing data filling; according to the node relationship model, attaches node function identifiers, data interaction channels, and upstream and downstream logical relationships to the cleaned data, and stores it as a retrievable basic data set;

[0066] Specifically, collecting data from each node includes: collecting sensor output data and packaging logs from the vehicle terminal, collecting encryption / transmission status information and communication logs from the network side, collecting unpacking and parsing results and response logs from the cloud; collecting logs from the data server;

[0067] The initial abnormal node extraction module: monitors the key indicators of each node, inputs the node abnormal metric function to calculate the node abnormal index of each node, determines whether the node abnormal index meets the threshold requirements, and screens to obtain the initial abnormal node set;

[0068] Explanation: The node abnormal metric function focuses on the abnormal performance of a single node and outputs the node abnormal index of each node to identify the initial abnormal nodes;

[0069] The abnormal diagnosis module extracts the abnormal feature correlation matrix from the initial abnormal node set. The abnormal feature correlation matrix includes the time feature, spatial location feature, and upstream and downstream dependency feature of the initial abnormal node; performs fault diagnosis based on the abnormal feature correlation matrix; takes measures based on the fault diagnosis result.

[0070] In the embodiments of the present invention, it needs to be further explained that the key indicators include common indicators and node function indicators. The common indicators at least include: data throughput, processing delay, resource occupancy rate, and log exception rate; the node function indicators reflect the performance of node categories. The node function indicators of the vehicle terminal node at least include: data packaging correct rate, sensor data accuracy rate, sensor data integrity rate, and encryption success rate; the node function indicators of the network side nodes (such as roadside units, transmission units) at least include: communication delay, packet loss rate, encryption / decryption pass rate; the node function indicators of the cloud node at least include: unpacking correct rate, parsing time consumption, data storage delay, and response time; the node function indicators of the data service node at least include: business processing correct rate, interface call success rate, and exception detection coverage rate.

[0071] In the embodiments of the present invention, it needs to be further explained that the node abnormal metric function satisfies the following formula:

[0072] ;

[0073] ;

[0074] ;

[0075] where m represents the total number of key indicators of the node, and s represents the sequential number of the key indicator; x s represents the actual value of the key indicator s; gw s represents the weight of the key indicator s (assigned according to business importance or historical statistical data); is used to quantify the key indicator x s relative to the threshold of the degree of abnormality, and α represents the adjustment coefficient; represents the abnormal deviation function, indicating the current value x s and the deviation degree from the average value μ s ; represents the fluctuation coefficient of the key indicator s.

[0076] It should be further explained that in the embodiments of the present invention, the abnormal feature association matrix includes initial abnormal node time features, spatial location features, and upstream and downstream dependency features;

[0077] The acquisition method of the time feature is: collecting the timestamp information of the initial abnormal node, including the time of first occurrence of the abnormality, the duration, and the frequency of the abnormality occurrence;

[0078] The acquisition method of the spatial location feature is: obtained according to the node relationship model; if it is a physical node, the spatio-temporal location information of the node is obtained by using geographical coordinates and position sensor data; if it is a data node, virtual space positioning is performed according to the server and network partition information where the node is located; the upstream and downstream dependency features include: data flow features, business process features, and interface call dependency features;

[0079] The acquisition method of the data flow feature is: starting from the initial abnormal node, tracing the data flow to the downstream nodes it connects, and collecting the abnormal conditions of the key indicators of the downstream nodes; or tracing the dependency source nodes of the initial abnormal node and evaluating the impact of upstream abnormalities on the current abnormality;

[0080] The acquisition method of the business process feature is: starting from the initial abnormal node, judging its position in the business process and the impact on its upstream and downstream steps;

[0081] The acquisition method of the interface call dependency feature is: recording the interface call behavior of the initial abnormal node and evaluating the impact of the initial abnormal node on the interface.

[0082] In the embodiments of the present invention, it needs to be further explained that the fault diagnosis refers to establishing a mapping relationship between the abnormal node type and the abnormal feature association matrix based on the decision tree algorithm, constructing a decision tree model by inducing the time features, spatial location features, and upstream and downstream dependency features of abnormal nodes in historical data or simulation-annotated samples, and identifying the abnormal type and locating the root cause of the fault through the decision tree model. For example, if the model determines that there is a positioning sensor fault in a vehicle terminal, the fault type of the vehicle terminal perception is output; if it is determined that there is a data packet loss in the encryption / transmission module of a network-side node, the fault cause of the network encryption module or the transmission module is output; if a significant deviation is found between the rule or model and the actual situation during the diagnosis process, the decision tree model needs to be updated and improved by adding new abnormal samples or correcting feature annotations, so as to continuously improve the diagnostic accuracy of the abnormal diagnosis module.

[0083] Furthermore, the decision tree model selects the best splitting point for training by calculating the gain value or information gain of the feature, and evaluates the accuracy of the model through the cross-validation method. During the diagnosis process, if a significant deviation is found between the rule or model and the actual situation, the decision tree model needs to be updated and improved by adding new abnormal samples or correcting feature annotations to improve the diagnostic accuracy.

[0084] In the embodiments of the present invention, it needs to be further explained that the measures taken based on the fault diagnosis results include:

[0085] Taking measures such as abnormal isolation, linkage remediation, risk control, fault repair and recovery, and case recording and model optimization. These measures can quickly locate abnormalities, reduce the impact of faults, improve system stability, and achieve system intelligence and high reliability through continuous optimization.

[0086] Explanation: The abnormal isolation measure refers to quickly isolating the abnormal node or path according to the fault diagnosis result to prevent the fault from spreading to other nodes or modules, including node isolation and path isolation. Node isolation means removing the diagnosed abnormal node from the data interaction process, for example, temporarily suspending its data transmission function and reconnecting it to the system after the problem is solved. Path isolation means that if the diagnosis result shows that there is an abnormality in a certain transmission path, the data flow is adjusted to bypass the fault point through other paths.

[0087] The linkage remediation measure refers to implementing remediation actions on the abnormal node and related paths according to the predefined linkage strategy to reduce the impact of the abnormality on the overall system, including enabling the redundancy mechanism, data rollback and retransmission, and dynamically adjusting the threshold.

[0088] The above-mentioned fault impact assessment and risk control refer to combining the fault diagnosis results, evaluating the impact scope of the current fault on the overall system, and implementing risk control measures; including impact assessment, early warning mechanism, and risk level classification; the risk level classification refers to classifying the fault impact scope and potential consequences according to the severity of the diagnosis results, such as low risk, medium risk, and high risk, and implementing hierarchical control measures;

[0089] Fault repair and recovery refer to taking repair and recovery operations according to the fault type and root cause in the diagnosis results; including software system update, hardware repair or replacement, and automatic reset;

[0090] Fault case recording and model optimization refer to recording the fault diagnosis process and the measures taken into the system case library for optimizing the anomaly detection and diagnosis model; including case library recording, model optimization, and linkage strategy iteration.

[0091] Summary: By constructing a node relationship model, collecting multi-source data and cleaning and processing it, calculating the node anomaly index, generating an anomaly feature correlation matrix, and performing fault diagnosis based on the decision tree algorithm, not only can the fault nodes and anomaly types be quickly located, but also the effective prevention of fault diffusion and the reduction of system risks can be achieved through measures such as anomaly isolation, linkage remedy, risk control, fault repair and recovery. In addition, through case recording and model optimization, the system has the ability of adaptive learning, can continuously optimize the fault diagnosis and response strategies, thereby improving the stability, reliability, and intelligent level of system operation.

[0092] Embodiment 2. The difference between the embodiment of the present invention and Embodiment 1 is that the system further includes:

[0093] A diffusion risk assessment module that performs correlation analysis on the propagation path of the initial abnormal node in the node relationship model and outputs the abnormal diffusion path and the abnormal diffusion risk coefficient;

[0094] A diffusion risk judgment and management module. If the abnormal diffusion risk coefficient exceeds the threshold, indicating that there is a high probability of a chain anomaly occurring, then based on the pre-defined linkage strategy, linkage early warning and protection actions are triggered, including switching to a backup communication channel, enabling a redundant parsing module, and notifying the dispatching platform to pay attention to potential risks.

[0095] In the embodiment of the present invention, it needs to be further explained that the acquisition method of the above-mentioned abnormal diffusion path is:

[0096] Taking the initial abnormal node as the path starting point, using the breadth-first search algorithm to traverse the node relationship model, starting from the starting point, layer by layer visiting the neighbor nodes that meet the upstream and downstream logical relationships. Each time during traversal, it is judged whether the new node meets the propagation probability threshold. If the requirement is met, the new node is added to the diffusion path and continue to traverse downwards;

[0097] When the set diffusion threshold or the number of traversals is reached, an abnormal diffusion path is output.

[0098] In the embodiments of the present invention, what needs to be further explained is to refer to Figure 2 the structural block diagram for obtaining the abnormal diffusion risk coefficient. The obtaining method of the abnormal diffusion risk coefficient is as follows:

[0099] Calculate the initial risk index: Analyze the initial abnormal node set, and obtain the risk factors of each initial abnormal node. The risk factors at least include: the information entropy of the node abnormal feature correlation matrix, the proportion of the influence of the node vehicle, the proportion of the influence of the node traffic (the ratio of the node data traffic to the total data traffic), the node hardware failure rate, the node software failure rate, and the node protection level; perform weighted summation on the quantified risk factors and the corresponding weights, and output the initial risk index of each initial abnormal node;

[0100] Record the product of the probabilities of all connected nodes as the propagation probability of the abnormal diffusion path;

[0101] Based on the data transmission frequency and the security protection level, set the channel weight for each abnormal diffusion path;

[0102] Jointly analyze the propagation probability and the channel weight, and calculate the path risk index of each abnormal diffusion path;

[0103] Add up the initial risk indices of all initial abnormal nodes to obtain the total node risk; add up the path risk indices of all abnormal diffusion paths to obtain the total path risk; perform weighted summation on the total node risk and the path risk to calculate the abnormal diffusion risk coefficient;

[0104] Based on the abnormal diffusion risk coefficient, conduct a hierarchical assessment of the overall network risk level and generate a network risk report;

[0105] Explanation: The network risk report includes: the risk level (low, medium, high, extremely high, etc.) delimited according to the abnormal diffusion risk coefficient and a detailed description of the current risk level, pointing out the high-risk areas or nodes and their reasons; the key information of the abnormal diffusion path, such as path nodes, length, and propagation probability, and pointing out the risk contribution of each node and the most critical influencing node; clarifying the risk level and potential threats of the initial abnormal nodes; specific protection plans for high-risk paths and nodes; in possible embodiments, the network risk report will also predict the future diffusion trend, provide early warning information, and simulate potential abnormal propagation, visually display the risk diffusion process with charts and path diagrams, and provide a basis for deploying preventive measures in advance.

[0106] In the embodiments of the present invention, what needs to be further explained is that the obtaining method of the initial risk index is as follows:

[0107] Suppose there are Q risk factors, and p represents the sequential number of the risk factor. Denote the value of the p-th risk factor as f p ; The initial risk index Cf is calculated through the following formula: The initial risk index is calculated through the following formula:

[0108] ;

[0109] where represents the weight of the p-th risk factor, which is set based on experience;

[0110] In the embodiments of the present invention, it needs to be further explained that let the starting point of each abnormal diffusion path be node i and the ending point be node j; The acquisition method of the channel weight is as follows:

[0111] ;

[0112] where, w ij represents the channel risk weight from node to , F ij represents the data transmission frequency from node to , and the value range is from 0 to 1. The larger the value, the more frequent the data interaction; S ij represents the average security protection level from node to , and the range is [0, 1]. The larger the value, the higher the security protection ability.

[0113] In the embodiments of the present invention, it needs to be further explained that the system further includes:

[0114] A disposal effect adaptive optimization module that tracks the effects of the linkage early warning and protection actions triggered by the diffusion risk assessment module, and continuously tracks the operating status and alarm indicators of the initial abnormal node; If the abnormality is alleviated or gradually disappears, record the current abnormality and the disposal process in the case library for improving the risk model and policy thresholds; If the disposal effect does not meet the standard or new risks are derived, re-incorporate the latest initial abnormal node into the abnormal diagnosis module, so as to continuously iterate and optimize the decision tree algorithm through the adaptive correction mechanism to ensure the reliability and security of the wireless data interaction process of the vehicle.

[0115] Summary: By adding a diffusion risk assessment module and a diffusion risk judgment and management module, the embodiments of the present invention can achieve comprehensive analysis and dynamic management of abnormal diffusion paths and their risks; obtain abnormal diffusion paths based on the breadth-first search algorithm, combine the initial risk index, propagation probability, and channel weight, accurately calculate the abnormal diffusion risk coefficient, and generate a detailed network risk report. Through risk grading assessment and the formulation of protection plans, the system can timely identify high-risk areas and key nodes, provide predictive warnings and protection measures, and significantly improve the risk resistance ability, abnormal management efficiency, and overall security of the system in a complex network environment.

[0116] Finally: The above are only the preferred embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A vehicle data wireless interaction system, characterized in that, Including: A node relationship model building module, which is used to build a node relationship model for vehicle data interaction. The node types include vehicle terminals, network-side nodes, cloud servers, and data servers. It models each node in the vehicle data interaction, and constructs a node relationship model by analyzing the data flow and logical relationships between the nodes; A multi-source data collection module, which collects data from each node to form a multi-source data set, and extracts key indicators of each node from the collected data; An initial abnormal node extraction module: monitors the key indicators of each node, inputs them into a node abnormality measurement function to calculate the node abnormality index of each node, determines whether the node abnormality index meets the threshold requirements, and filters to obtain an initial set of abnormal nodes; An abnormality diagnosis module, which extracts an abnormal feature correlation matrix from the initial set of abnormal nodes. The abnormal feature correlation matrix includes the time features, spatial location features, and upstream and downstream dependency features of the initial abnormal nodes; Performs fault diagnosis based on the abnormal feature correlation matrix; takes measures based on the fault diagnosis results; A diffusion risk assessment module, which conducts a correlation analysis on the propagation paths of the initial abnormal nodes in the node relationship model, and outputs the abnormal diffusion paths and the abnormal diffusion risk coefficients. The method for obtaining the abnormal diffusion risk coefficients is as follows: Calculates the initial risk index: analyzes the initial set of abnormal nodes, obtains the risk factors of each initial abnormal node. The risk factors at least include: the information entropy of the node abnormal feature correlation matrix, the proportion of vehicle influence of the node, the proportion of traffic influence of the node, the hardware failure rate of the node, the software failure rate of the node, and the protection level of the node; performs a weighted sum of the quantified risk factors and the corresponding weights, and outputs the initial risk index of each initial abnormal node; records the product of the probabilities of all connected nodes as the propagation probability of the abnormal diffusion path; sets channel weights for each abnormal diffusion path based on the data transmission frequency and security protection level; jointly analyzes the propagation probability and the channel weights, and calculates the path risk index of each abnormal diffusion path; adds up the initial risk indices of all initial abnormal nodes to obtain the total node risk; adds up the path risk indices of all abnormal diffusion paths to obtain the total path risk; performs a weighted sum of the total node risk and the path risk to calculate the abnormal diffusion risk coefficient; conducts a hierarchical assessment of the overall network risk level based on the abnormal diffusion risk coefficient, and generates a network risk report; The network risk report includes: the risk level defined according to the abnormal diffusion risk coefficient, the key information of the abnormal diffusion path, the length of the abnormal diffusion path, and the propagation probability; A diffusion risk judgment and management module. If the abnormal diffusion risk coefficient exceeds the threshold, indicating that there is a high probability of a chain abnormality occurring, it triggers linkage warnings and protection actions based on predefined linkage strategies, including switching to a backup communication channel, enabling a redundant parsing module, and notifying the dispatching platform to pay attention to potential risks.

2. The vehicle data wireless interaction system according to claim 1, wherein The key indicators include common indicators and node function indicators. The common indicators at least include: data throughput, processing delay, resource occupancy rate, and log anomaly rate. The node function indicators reflect the performance of node categories. The node function indicators of vehicle terminal nodes at least include: data packet packing accuracy rate, sensor data accuracy rate, sensor data integrity rate, and encryption success rate. The node function indicators of network-side nodes at least include: communication delay, packet loss rate, encryption / decryption passing rate. The node function indicators of cloud nodes at least include: unpacking accuracy rate, parsing time-consuming, data storage delay, and response time. The node function indicators of data service nodes at least include: business processing accuracy rate, interface call success rate, and anomaly detection coverage rate.

3. The vehicle data wireless interaction system according to claim 1, characterized in that The node anomaly metric function satisfies the following formula: ； ； ； Among them, m represents the total number of key indicators of the node, and s represents the serial number of the key indicator; represents the actual value of the key indicator s; represents the weight of the key indicator s; is used to quantify the key indicator relative to the threshold of the degree of abnormality, and α represents the adjustment coefficient; represents the abnormal deviation function, represents the current value and the average value of the degree of deviation, represents the fluctuation coefficient of the key indicator s.

4. The vehicle data wireless interaction system according to claim 1, wherein The anomaly feature correlation matrix includes the time feature, spatial location feature, and upstream and downstream dependency features of the initial anomaly node. The acquisition method of the time feature is: collect the timestamp information of the initial anomaly node, including the time of first occurrence of the anomaly, the duration, and the frequency of the anomaly occurrence. The acquisition method of the spatial location feature is: obtained according to the node relationship model. If it is a physical node, use geographical coordinates and position sensor data to obtain the spatio-temporal location information of the node. If it is a data node, perform virtual space positioning according to the server where the node is located and network partition information. The upstream and downstream dependency features include: data flow feature, business process feature, and interface call dependency feature. The acquisition method of the data flow feature is: starting from the initial anomaly node, trace the data flow to the downstream nodes it connects, and collect the anomaly situations of the key indicators of the downstream nodes. Or trace the dependency source node of the initial anomaly node and evaluate the impact of upstream anomalies on the current anomaly. The acquisition method of the business process feature is: starting from the initial anomaly node, judge its position in the business process and the impacts on its upstream and downstream steps. The acquisition method of the interface call dependency feature is: record the interface call behavior of the initial anomaly node and evaluate the impact of the initial anomaly node on the interface.

5. The vehicle data wireless interaction system according to claim 1, wherein The fault diagnosis refers to establishing a mapping relationship between the anomaly node type and the anomaly feature correlation matrix based on the decision tree algorithm. By inducing the time feature, spatial location feature, and upstream and downstream dependency features of the anomaly nodes in historical data or simulation-annotated samples, construct a decision tree model, and identify the anomaly type and locate the root cause of the fault through the decision tree model.

6. The vehicle data wireless interaction system according to claim 1, wherein The acquisition method of the anomaly diffusion path is: Take the initial anomaly node as the starting point of the path, and use the breadth-first search algorithm to traverse the node relationship model. Starting from the starting point, visit the neighbor nodes that satisfy the upstream and downstream logical relationships layer by layer. Each time during the traversal, judge whether the new node meets the propagation probability threshold. If it meets the requirements, add the new node to the diffusion path and continue to traverse the lower layer. When the set traversal times are reached, output the anomaly diffusion path.

7. The vehicle data wireless interaction system according to claim 1, characterized in that The acquisition method of the initial risk index is: Suppose there are Q risk factors, and p represents the sequential number of a risk factor. Denote the value of the p-th risk factor as ; The initial risk index is calculated by the following formula : The initial risk index is calculated by the following formula: ； Among them, represents the weight of the p-th risk factor, which is set based on experience; Let the starting point of each anomaly diffusion path be node i and the ending point be node j. The acquisition method of the channel weight is: ； Among them, represents the channel risk weight from node i to j, represents the data transmission frequency from node i to j, and the value range is from 0 to 1. The larger the value, the more frequent the data interaction; represents the average security protection level from node i to j, with a range of [0, 1]. The larger the value, the higher the security protection ability.

8. The vehicle data wireless interaction system according to claim 1, wherein, It also includes: The disposal effect adaptive optimization module tracks the effects of the linkage early warning and protection actions triggered by the diffusion risk assessment module, continuously monitors the operating status and warning indicators of the initial abnormal nodes; if the abnormality is alleviated or gradually disappears, the current abnormality and the disposal process are recorded in the case library to improve the risk model and policy thresholds; if the disposal effectiveness does not meet the standards or new risks are derived, the latest initial abnormal nodes are re-incorporated into the abnormal diagnosis module, so as to continuously iterate and optimize the decision tree algorithm through the adaptive correction mechanism to ensure the reliability and security of the wireless vehicle data interaction process.

Images