CN-122020716-A - Intelligent traffic control system of big data traffic Internet of things

CN122020716ACN 122020716 ACN122020716 ACN 122020716ACN-122020716-A

Abstract

The invention discloses an intelligent traffic control system of a big data traffic Internet of things, and relates to the technical field of traffic control. The system comprises a multi-source data acquisition module, a co-building sharing center, a dynamic authority management and control module, a cross-domain collaborative decision module, a safety protection and operation and maintenance module, wherein the multi-department heterogeneous traffic data is accessed through a standardized interface and preprocessed to generate standardized data, a coalition chain is adopted to record a full life cycle log of data, multi-department data combined model training is realized based on transverse federal learning, the dynamic authority management and control module adopts an RBAC+ABAC hybrid access control model and combines data desensitization and access audit functions to realize data grading authorization and safety management and control, and the cross-domain collaborative decision module generates a cross-department linkage treatment scheme, and the safety protection and operation and maintenance module guarantees data safety and ensures stable operation of a system through intelligent operation and maintenance and data life cycle management. The invention can realize accurate perception, intelligent research and judgment and efficient treatment of traffic running states and provides a brand new solution for urban traffic management.

Inventors

GAO QIANG
ZHU CONGRUI
WANG LIJUN
ZHAO YUYU
DU XIAOCUI
LI YING
GUO JIARONG
DENG YUXUAN
DENG XIN
MAO HAIJIAO
TIAN ZHOU
QIU JIE
ZHANG JINSONG
CHENG HAORAN
DU KAI
SONG JINGNI
CHEN DAN
JIN YINLI
HE XINYING
MA JINGBO
LIU HUIYANG

Assignees

长安大学

Dates

Publication Date: 20260512
Application Date: 20260129

Claims (6)

1. An intelligent traffic management and control system of big data traffic thing networking, which is characterized by comprising: The multi-source data acquisition module is accessed into multi-department heterogeneous traffic data through a standardized interface, and standardized data is generated through format conversion and preprocessing; the co-building sharing center adopts a alliance chain to record a full life cycle log of data, and realizes multi-department data joint model training based on transverse federal learning; The dynamic authority management and control module adopts an RBAC+ABAC hybrid access control model, combines data desensitization and access audit functions, and realizes data hierarchical authorization and security management and control; the cross-domain collaborative decision-making module generates a cross-department linkage treatment scheme through a multi-scene decision-making model library and a collaborative scheduling engine; The safety protection and operation and maintenance module adopts transmission encryption and storage encryption to ensure the safety of data, and ensures the stable operation of the system through intelligent operation and maintenance and data life cycle management.
2. The intelligent traffic control system of the big data traffic Internet of things according to claim 1, wherein, The multi-department data federation model includes a federation chain network, a file system network, and a federation learning unit.
3. The intelligent traffic control system of the big data traffic internet of things according to claim 2, wherein, The multi-department data joint model training comprises: the multi-department data joint model training comprises: Performing qualification examination on all nodes of which the intention forms a alliance, dynamically dividing the nodes passing the examination into a plurality of subgroups according to a consensus algorithm, selecting one node in each subgroup as an aggregation node, and forming an alliance chain network among the subgroups; each node completes identity verification based on an intelligent contract unit in the alliance chain network, and each node is used as a request user to upload task needs and task information to the alliance chain network; The method comprises the steps that a request user is used as a data request party and provides a unified federation learning unit and a sharing model, the rest or all nodes are used as data providers, local data processing and model testing are carried out through the federation learning unit and the sharing model, and the data providers issue results of the data processing and the model testing to a federation chain network; The method comprises the steps that a user is requested to announce all data providers to start training tasks, the data providers train a local model in a local data set, after training is completed, local model parameters are uploaded to an interstellar file system network, a return value is obtained, the return value is uploaded to a alliance chain network, and when the number of the uploaded data providers reaches a threshold value, an aggregation node sends an aggregation task request to an intelligent contract unit; The aggregation node downloads local model parameters uploaded by other data providers in the group to the body, and performs an intra-group aggregation task according to a weighted federal average algorithm based on the reputation value; After all the sink nodes complete the aggregation task in the group, a polling mechanism is adopted to select one node from the sink nodes to bear the aggregation task between groups, local model parameters uploaded by all the groups are aggregated into new model parameters, and a shared model is updated; And after the shared model is updated in the round, evaluating the reputation value of the data provider according to the historical behavior and the current behavior, and finishing the reputation value update of the data provider until the shared model is converged.
4. The intelligent traffic control system of the big data traffic Internet of things according to claim 1, wherein, The RBAC+ABAC hybrid access control model includes: Defining entity class of authority model, associating entity class with object of service resource to be controlled to form association information; acquiring defined control item information according to the model entity class, setting corresponding control conditions for each defined control item, and generating condition information; And calling different condition resolvers to analyze according to the condition information, obtaining an analysis result, and giving the user permission based on the analysis result.
5. The intelligent traffic control system of the big data traffic Internet of things according to claim 1, wherein, Generating the cross-department linkage treatment plan includes: the system comprises a multi-scene decision model library, a decision scheme set and a traffic data processing module, wherein the multi-scene decision model library is used for storing traffic data; performing simulation deduction on each decision scheme in the decision scheme set by utilizing a collaborative scheduling engine to obtain a simulation result of each decision scheme; And calculating index data of each decision scheme according to the simulation result to generate evaluation results of different decision schemes, and obtaining an optimal scheme.
6. The intelligent traffic control system of the big data traffic Internet of things according to claim 1, wherein, The security of the data transmission channel is ensured by a TLS/SSL protocol, the security of the static data and the sensitive fields is ensured by adopting high-strength encryption storage, and the intelligent operation and maintenance is realized by monitoring and early warning, log analysis and fault self-healing.

Description

Intelligent traffic control system of big data traffic Internet of things Technical Field The invention relates to the technical field of traffic control, in particular to an intelligent traffic control system of a big data traffic Internet of things. Background With the acceleration of the urban process and the rapid increase of the conservation quantity of motor vehicles, the problems of traffic jam, accident frequency, low management efficiency and the like become key bottlenecks for restricting the urban development. The traditional traffic control system has difficulty in adapting to the requirements of modern traffic control due to the defects of single data source, insufficient cooperation of departments, weak safety protection and the like. Therefore, how to provide an intelligent traffic control system of the internet of things for big data traffic to solve the difficulties existing in the prior art is a problem that needs to be solved by those skilled in the art. Disclosure of Invention In view of the above, the invention provides an intelligent traffic control system of the big data traffic Internet of things, which can realize accurate perception, intelligent research and judgment and efficient treatment of traffic running states and provides a brand new solution for urban traffic control. In order to achieve the above purpose, the present invention adopts the following technical scheme: an intelligent traffic management and control system of big data traffic internet of things, comprising: The multi-source data acquisition module is accessed into multi-department heterogeneous traffic data through a standardized interface, and standardized data is generated through format conversion and preprocessing; the co-building sharing center adopts a alliance chain to record a full life cycle log of data, and realizes multi-department data joint model training based on transverse federal learning; The dynamic authority management and control module adopts an RBAC+ABAC hybrid access control model, combines data desensitization and access audit functions, and realizes data hierarchical authorization and security management and control; the cross-domain collaborative decision-making module generates a cross-department linkage treatment scheme through a multi-scene decision-making model library and a collaborative scheduling engine; The safety protection and operation and maintenance module adopts transmission encryption and storage encryption to ensure the safety of data, and ensures the stable operation of the system through intelligent operation and maintenance and data life cycle management. Optionally, the multi-department data federation model includes a federation chain network, a file system network, and a federation learning unit. Optionally, the multi-department data federation model training includes: Performing qualification examination on all nodes of which the intention forms a alliance, dynamically dividing the nodes passing the examination into a plurality of subgroups according to a consensus algorithm, selecting one node in each subgroup as an aggregation node, and forming an alliance chain network among the subgroups; each node completes identity verification based on an intelligent contract unit in the alliance chain network, and each node is used as a request user to upload task needs and task information to the alliance chain network; The method comprises the steps that a request user is used as a data request party and provides a unified federation learning unit and a sharing model, the rest or all nodes are used as data providers, local data processing and model testing are carried out through the federation learning unit and the sharing model, and the data providers issue results of the data processing and the model testing to a federation chain network; The method comprises the steps that a user is requested to announce all data providers to start training tasks, the data providers train a local model in a local data set, after training is completed, local model parameters are uploaded to an interstellar file system network, a return value is obtained, the return value is uploaded to a alliance chain network, and when the number of the uploaded data providers reaches a threshold value, an aggregation node sends an aggregation task request to an intelligent contract unit; The aggregation node downloads local model parameters uploaded by other data providers in the group to the body, and performs an intra-group aggregation task according to a weighted federal average algorithm based on the reputation value; After all the sink nodes complete the aggregation task in the group, a polling mechanism is adopted to select one node from the sink nodes to bear the aggregation task between groups, local model parameters uploaded by all the groups are aggregated into new model parameters, and a shared model is updated; And after the shared model is updated in the round, evaluating the reputation