Intelligent high-speed multi-source data fusion and multi-service collaborative processing methods and digital platforms

Abstract

This application discloses a method and digital platform for multi-source data fusion and multi-service collaborative processing of smart highways, relating to the field of smart highway technology. The method includes: acquiring multi-source operational data of the target highway network; mapping the multi-source operational data to a unified spatiotemporal benchmark of the road network to obtain standardized operational data; generating or updating multiple spatiotemporal twin objects to represent different operational objects based on the standardized operational data; constructing a spatiotemporal twin object map based on the spatiotemporal and business relationships between the spatiotemporal twin objects; determining event association information based on the spatiotemporal twin object map when an event is detected in the target road network; generating a multi-service collaborative processing strategy based on the event association information and outputting collaborative processing information. This can alleviate the problems of spatiotemporal inconsistency and business fragmentation of multi-source data, and improve the efficiency of road network event association analysis and collaborative processing.

Description

Technical Field

[0001] This application relates to the field of smart highway technology, and in particular to a method and digital platform for multi-source data fusion and multi-service collaborative processing of smart highways. Background Technology

[0002] With the continuous expansion of the expressway network and the advancement of digital and intelligent transportation construction, expressway operation and management are gradually evolving from traditional manual inspections, single-point monitoring, and decentralized business management towards comprehensive perception, real-time analysis, collaborative response, and digital twin visualization. In the context of smart expressways, the network's operational status involves not only traffic flow, traffic volume, and abnormal events, but also various other data, including data from toll stations, gantries, video surveillance, radar-visual fusion sensing equipment, meteorological monitoring equipment, electromechanical facilities, emergency supplies, road assets, and 3D geographic models. How to uniformly access, integrate, process, and collaboratively apply this multi-source, heterogeneous data, and intuitively present the network's operational status in a digital environment, has become a crucial technological direction for improving expressway operational efficiency, safety assurance capabilities, and emergency response capabilities.

[0003] In related technologies, intelligent highway management platforms typically integrate business data from video surveillance, toll collection systems, gantry systems, meteorological systems, and electromechanical facility monitoring systems to display road network operation information, query vehicle traffic, issue alarms for abnormal events, and perform some statistical analysis functions. Some platforms also incorporate two-dimensional geographic information systems or three-dimensional visualization scenes to display information such as traffic flow, event location, and facility status. However, these solutions are mostly based on a single business system or a single data layer. Differences exist in the collection frequency, data format, timestamp accuracy, spatial coordinate system, and business identification rules among various data types, easily leading to data silos and information fragmentation. Furthermore, the lack of a unified, object-oriented association between vehicles, events, road sections, facilities, meteorology, and emergency resources makes it difficult to automatically link surrounding videos, affected road sections, key vehicles, electromechanical facilities, and dispatchable emergency resources after an abnormal event occurs. In addition, existing platforms often emphasize status display and manual handling, making it difficult to generate collaborative handling strategies across multiple businesses such as monitoring, toll collection, maintenance, emergency response, and electromechanical control based on a unified digital foundation, affecting the accuracy of road network situation assessment and the timeliness of emergency response.

[0004] Therefore, in the construction of the intelligent highway digital platform, the difficulties in spatiotemporal alignment of multi-source heterogeneous data, the insufficient expression of correlations between vehicle event facilities and resources, and the weak multi-business collaborative processing capabilities have become urgent problems to be solved. Summary of the Invention

[0005] This application provides a method and digital platform for multi-source data fusion and multi-service collaborative processing of smart highways, aiming to solve the problems of spatiotemporal alignment difficulties of multi-source heterogeneous data, insufficient expression of correlation between vehicle event facility resources, and weak multi-service collaborative processing capabilities in the construction of smart highway digital platforms.

[0006] Firstly, a method for intelligent high-speed multi-source data fusion and multi-service collaborative processing, the method comprising: Acquire multi-source operational data of the target highway network; The multi-source operational data is mapped to a unified road network spatiotemporal reference to obtain standardized operational data. Multiple spatiotemporal twin objects are generated or updated based on the standardized operational data, and the multiple spatiotemporal twin objects are used to characterize different operational objects in the target highway network; Based on the spatiotemporal and business relationships between the spatiotemporal twin objects, a spatiotemporal twin object graph is constructed; Upon detecting a target road network event, event association information corresponding to the target road network event is determined based on the spatiotemporal twin object map. A multi-service collaborative handling strategy is generated based on the event association information, and collaborative handling information corresponding to the multi-service collaborative handling strategy is output.

[0007] Optionally, in the above scheme, acquiring multi-source operational data of the target highway network includes: Initial access data is obtained by accessing at least two types of data from the target highway network, including roadside perception data, video recognition data, toll station traffic data, gantry traffic data, meteorological monitoring data, electromechanical facility operation data, emergency resource data, road asset data, and 3D model data. The initial access data is parsed and format standardized to obtain the initial observation data; The multi-source operational data is generated based on the data source, collection time, and service type corresponding to the initial observation data.

[0008] Optionally, in the above scheme, mapping the multi-source operational data to a unified road network spatiotemporal reference to obtain standardized operational data includes: Based on the acquisition time, reception time, and data source delay information in the multi-source operational data, time correction is performed on the multi-source operational data to obtain unified time information; Based on road network topology information, equipment deployment location information, station information, lane information, and 3D scene coordinate information, construct the road network spatial mapping relationship; Based on the unified time information and the road network spatial mapping relationship, the multi-source operation data is spatiotemporally mapped to obtain the standardized operation data.

[0009] Optionally, in the above scheme, generating or updating multiple spatiotemporal twin objects based on the standardized operational data includes: Based on the category of the operational object corresponding to the standardized operational data, the standardized operational data is classified to obtain multiple object observation datasets; Based on the object observation datasets, object status information is extracted, including object spatiotemporal status, object business attributes, and data source information; Based on the object state information, at least one of the following is generated or updated: vehicle twin object, road segment twin object, event twin object, facility twin object, environment twin object, and resource twin object, to obtain the plurality of spatiotemporal twin objects.

[0010] Optionally, in the above scheme, generating or updating at least one of the following based on the object state information: vehicle twin object, road segment twin object, event twin object, facility twin object, environment twin object, and resource twin object, includes: The time matching information is determined based on the temporal proximity between the state information of the object to be merged and the already generated spatiotemporal twin object; Spatial matching information is determined based on the spatial proximity between the state information of the object to be fused and the generated spatiotemporal twin object. Attribute matching information is determined based on the degree of similarity of business attributes between the state information of the object to be merged and the generated spatiotemporal twin object; Based on the road network topology reachability status corresponding to the status information of the object to be merged, determine the topology matching information; Based on the reliability of the data source corresponding to the state information of the object to be merged, the data source credibility information is determined; Based on the time matching information, the spatial matching information, the attribute matching information, the topology matching information, and the data source credibility information, the object matching result is determined; Update the matched spatiotemporal twin objects based on the object matching results, or generate new spatiotemporal twin objects.

[0011] Optionally, in the above scheme, constructing a spatiotemporal twin object graph based on the spatiotemporal and business relationships between the spatiotemporal twin objects includes: Based on the spatial location and road network affiliation information of each spatiotemporal twin object, the spatial association results between each spatiotemporal twin object are determined; Based on the temporal state and state change sequence of each spatiotemporal twin object, the temporal correlation result between each spatiotemporal twin object is determined; Based on the object category and business attributes of each spatiotemporal twin object, determine the business association results between each spatiotemporal twin object; Based on the spatial association results, the temporal association results, and the business association results, an object association relationship is established, including at least one of the following: location relationship, passing relationship, influence relationship, observable relationship, schedulable relationship, controllable relationship, and disposal relationship, to obtain the spatiotemporal twin object map.

[0012] Optionally, in the above scheme, determining the event association information corresponding to the target road network event based on the spatiotemporal twin object map when a target road network event is detected includes: The target event object is determined in the spatiotemporal twin object map based on the target road network event; Based on the influence and transit relationships corresponding to the target event objects, the affected road segment objects and affected vehicle objects are determined; Based on the observable relationships corresponding to the target event objects, the observable facility objects are determined; Based on the schedulable and controllable relationships corresponding to the target event objects, determine the schedulable resource objects and controllable facility objects; The event association information is generated based on at least one of the target event object, the affected road segment object, the affected vehicle object, the observable facility object, the schedulable resource object, and the controllable facility object.

[0013] Optionally, in the above scheme, generating a multi-service collaborative handling strategy based on the event association information includes: Based on the event association information, the event type, event location, scope of impact, and risk level of the target road network event are determined to obtain the event handling scenario; Based on the event handling scenario and the associated objects in the event association information, multiple candidate business handling actions are generated. The candidate business handling actions include at least one of the following: video verification action, traffic guidance action, key vehicle monitoring action, emergency rescue action, electromechanical facility control action, toll station coordination action, and maintenance handling action. The timeliness, resource availability, traffic impact, and risk reduction effect of each candidate business action are evaluated to obtain the evaluation results. Based on the evaluation results, the target business action is determined from the multiple candidate business action actions to obtain the multi-business collaborative action strategy.

[0014] Optionally, in the above scheme, after outputting the collaborative processing information corresponding to the multi-service collaborative processing strategy, the scheme further includes: Obtain the processing execution data and road network recovery data corresponding to the multi-service collaborative processing strategy; The execution result of the disposal action is determined based on the disposal execution data; The event status change results and traffic status recovery results are determined based on the road network recovery data. Based on the execution results of the handling actions, the results of the event status changes, and the results of the traffic status recovery, update at least one of the following: data source credibility, object matching parameters, event association information determination parameters, and collaborative handling strategy generation parameters.

[0015] Secondly, a smart, high-speed, multi-source data fusion and multi-service collaborative processing digital platform, the digital platform comprising: The multi-source data access module is used to acquire multi-source operational data of the target highway network; The spatiotemporal reference mapping module is used to map the multi-source operational data to a unified road network spatiotemporal reference to obtain standardized operational data; The twin object management module is used to generate or update multiple spatiotemporal twin objects based on the standardized operational data. The multiple spatiotemporal twin objects are used to characterize different operational objects in the target highway network. The object graph construction module is used to construct a spatiotemporal twin object graph based on the spatiotemporal and business relationships between the spatiotemporal twin objects. The event association analysis module is used to determine the event association information corresponding to the target road network event based on the spatiotemporal twin object map when a target road network event is detected. The collaborative handling module is used to generate a multi-service collaborative handling strategy based on the event association information, and output collaborative handling information corresponding to the multi-service collaborative handling strategy; The 3D twin display module is used to generate 3D twin display information of the target highway network based on the spatiotemporal twin object, the spatiotemporal twin object map, the event association information, and the multi-service collaborative processing strategy. The feedback update module is used to update the spatiotemporal twin object, the spatiotemporal twin object map, and / or the strategy generation parameters of the collaborative processing module based on the processing feedback data corresponding to the multi-service collaborative processing strategy.

[0016] Compared with the prior art, this application has at least the following beneficial effects: Based on further analysis and research of existing technical problems, this application recognizes that existing technologies face difficulties in the spatiotemporal alignment of multi-source heterogeneous data, insufficient correlation expression between vehicle event facility resources, and weak multi-service collaborative processing capabilities in the construction of intelligent highway digital platforms. By acquiring multi-source operational data of the target highway network and mapping this multi-source operational data to a unified network spatiotemporal reference, highway operational data from different sources, in different formats, with different time granularities, and with different spatial representations can be converted into standardized operational data that can be uniformly calculated. This eliminates data fragmentation caused by temporal and spatial inconsistencies for subsequent correlation analysis. Furthermore, this application further generates or improves upon standardized operational data... The new spatiotemporal twin objects used to represent different operational objects enable operational elements such as vehicles, road segments, events, facilities, environment, and resources to no longer exist as isolated data records, but rather form unified object units capable of expressing the object's state. Subsequently, a spatiotemporal twin object map is constructed based on the spatiotemporal and business relationships between the various spatiotemporal twin objects, allowing for the formation of queryable and derivable association structures between different operational objects in the target highway network. When an event is detected in the target highway network, the event association information is determined based on the spatiotemporal twin object map, and multi-business collaborative handling strategies and corresponding collaborative handling information are generated accordingly, transforming event handling from single-point alarm display to collaborative handling based on associated objects. Thus, this application can solve the problems of spatiotemporal alignment difficulties of multi-source heterogeneous data, insufficient expression of relationships between vehicle events, facilities, and resources, and weak multi-business collaborative handling capabilities in intelligent highway digital platforms, improving the accuracy and timeliness of highway network operational situation awareness, event association analysis, and collaborative handling. Attached Figure Description

[0017] Figure 1 A flowchart illustrating a method for intelligent high-speed multi-source data fusion and multi-service collaborative processing provided in one embodiment of this application; Figure 2 This is a block diagram of the module architecture of a smart high-speed multi-source data fusion and multi-service collaborative processing digital platform provided in one embodiment of this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0019] In one embodiment, such as Figure 1As shown, a method for intelligent highway multi-source data fusion and multi-service collaborative processing is provided. This method can be executed by a digital platform for intelligent highway multi-source data fusion and multi-service collaborative processing. This digital platform can be deployed in highway operation and management centers, road network sub-centers, cloud platforms, or edge computing nodes, or it can be deployed in a cloud-edge collaborative manner to adapt to application scenarios of long-distance, multi-site, multi-device, and multi-service system parallel operation on highways.

[0020] In some embodiments, the intelligent highway multi-source data fusion and multi-service collaborative processing method includes: acquiring multi-source operational data of the target highway network; mapping the multi-source operational data to a unified road network spatiotemporal reference to obtain standardized operational data; generating or updating multiple spatiotemporal twin objects based on the standardized operational data, wherein the multiple spatiotemporal twin objects are used to represent different operational objects in the target highway network; constructing a spatiotemporal twin object map based on the spatiotemporal and business association relationships between the spatiotemporal twin objects; determining the event association information corresponding to the target highway network event based on the spatiotemporal twin object map when an event is detected in the target road network; generating a multi-service collaborative processing strategy based on the event association information, and outputting collaborative processing information corresponding to the multi-service collaborative processing strategy.

[0021] Specifically, the target highway network can be a single highway, a highway section, a section under the jurisdiction of a management sub-center, a highway network area, or a highway operation area including key areas such as tunnels, toll stations, service areas, interchanges, bridges, and slopes. Multi-source operational data can originate from roadside sensing equipment, video surveillance equipment, radar-visual fusion equipment, toll station systems, gantry systems, weather stations, electromechanical facility monitoring systems, emergency material management systems, road asset management systems, geographic information systems, and 3D model systems. The digital platform can access the aforementioned multi-source operational data through data interfaces, message queues, streaming data channels, batch synchronization interfaces, or device protocol gateways.

[0022] After acquiring multi-source operational data, the digital platform can perform unified spatiotemporal processing on this data. Because different devices or business systems may use different timestamps, spatial representations, and business codes when collecting data—for example, some data may represent location using latitude and longitude, some using station numbers, some using device numbers or toll station numbers, and some 3D model data using 3D scene coordinates—it is necessary to map the multi-source operational data to a unified road network spatiotemporal reference. This unified road network spatiotemporal reference can include unified time information, road segment information, driving direction information, station number information, lane information, and spatial location information. The standardized operational data obtained after mapping can be calculated, compared, and correlated within the same time and road network spatial semantics.

[0023] After obtaining standardized operational data, the digital platform can transform scattered data records into spatiotemporal twin objects based on the corresponding operational objects. For example, vehicle detection records, vehicle passage records, and vehicle trajectory fragments can be merged into vehicle twin objects; road segment traffic flow, average speed, congestion level, and capacity can be merged into road segment twin objects; information on accidents, disabled vehicles, road obstacles, abnormal parking, congestion, and the impact of severe weather can be merged into event twin objects; the operational status of cameras, radar equipment, information boards, toll collection facilities, and tunnel electromechanical equipment can be merged into facility twin objects; meteorological information collected by weather stations and environmental monitoring equipment can be merged into environmental twin objects; and information on emergency material warehouses, rescue vehicles, obstacle clearing equipment, and maintenance forces can be merged into resource twin objects. Through this processing, multi-source operational data no longer exists merely as isolated data records, but is transformed into spatiotemporal twin objects that can express the status of highway operational objects.

[0024] After generating or updating multiple spatiotemporal twin objects, the digital platform can further construct a spatiotemporal twin object graph based on the spatiotemporal and business relationships between these objects. Spatiotemporal relationships can include connections such as a vehicle being located on a road segment, a vehicle passing through a gantry, an event occurring within a certain mileage interval, equipment covering a road segment, and emergency resources being located in a service area or management station. Business relationships can include connections such as the road segment affected by the event, the vehicle associated with the event, the event being verifiable by nearby cameras, the event being handled by emergency resources, and facilities performing traffic guidance or electromechanical control. The spatiotemporal twin object graph integrates vehicles, events, road segments, facilities, the environment, and resources into a unified network of relationships.

[0025] Upon detecting a target road network event, the digital platform can locate the corresponding target event object in a spatiotemporal twin object map and determine event association information based on the relationships between the target event objects. Target road network events can include traffic accidents, vehicle malfunctions, abnormal parking, road obstructions, traffic congestion, abnormal behavior of key vehicles, low visibility, heavy rainfall, tunnel anomalies, toll station congestion, service area congestion, and electromechanical facility malfunctions. Event association information can include event location, event type, event impact range, affected road sections, affected vehicles, observable facilities, dispatchable resources, controllable facilities, and related business subsystems. Subsequently, the digital platform generates multi-business collaborative handling strategies based on the event association information. These strategies can include actions such as video verification, traffic guidance, key vehicle monitoring, emergency rescue, electromechanical facility control, toll station coordination, and maintenance, and can generate collaborative handling information for different business subsystems, such as alarm information, dispatch suggestions, linkage control commands, work order information, traffic guidance content, or 3D twin display information.

[0026] Through the above implementation methods, multi-source heterogeneous data is first mapped to a unified road network spatiotemporal reference, then transformed into spatiotemporal twin objects. Furthermore, cross-object and cross-business relationships are established through the spatiotemporal twin object graph, enabling the automatic determination of event correlation information and the generation of multi-business collaborative handling strategies after a target road network event occurs. This reduces data fragmentation between different business systems, enhances the correlation analysis capabilities between vehicles, events, road sections, facilities, environment, and resources, and improves the timeliness and accuracy of monitoring, early warning, and collaborative handling in smart highway scenarios.

[0027] In some embodiments, acquiring multi-source operational data of the target highway network includes: accessing at least two types of data from the target highway network, such as roadside sensing data, video recognition data, toll station traffic data, gantry traffic data, meteorological monitoring data, electromechanical facility operation data, emergency resource data, road asset data, and 3D model data, to obtain initial access data; performing field parsing and format standardization on the initial access data to obtain initial observation data; and generating multi-source operational data based on the data source, collection time, and service type corresponding to the initial observation data.

[0028] Specifically, roadside perception data can include information such as vehicle location, vehicle speed, lane number, vehicle type, target number, trajectory segments, traffic flow, and abnormal event identification results collected by radar-visual fusion equipment, millimeter-wave radar, lidar, traffic incident detection equipment, and vehicle flow detectors. Video recognition data can include information such as license plate, vehicle type, vehicle color, lane occupancy, traffic accidents, vehicle malfunctions, littering, pedestrian intrusion, and abnormal parking identified by video surveillance cameras or intelligent video analysis equipment. Toll station passage data can include the time of vehicle entry or exit from the toll station, toll station number, lane number, vehicle identification, vehicle type, and direction of travel. Gantry passage data can include ETC gantry number, vehicle passage time, vehicle identification, direction of travel, and road segment information. Meteorological monitoring data can include environmental information such as temperature, humidity, wind speed, wind direction, rainfall, snowfall, visibility, and road surface slipperiness.

[0029] Electromechanical facility operation data can include the online status, fault codes, operating parameters, and control status of facilities such as monitoring systems, toll collection systems, tunnel systems, information boards, lane indicators, lighting equipment, fans, broadcasting equipment, cameras, and radar-visual fusion equipment. Emergency resource data can include the location of emergency material warehouses, material types, inventory quantities, expiration dates, rescue vehicle locations, rescue equipment status, and maintenance force availability. Road asset data can include the category, location, specifications, usage status, and maintenance information of assets such as pavements, roadbeds, bridges, tunnels, slopes, guardrails, signs, landscaping, monitoring facilities, service areas, toll stations, and interchanges. 3D model data can include 3D spatial data of highways constructed based on satellite imagery, digital elevation models, laser scanning, oblique photography, real-scene image measurement, and building information models.

[0030] After initial data access, the digital platform can parse the fields within that data. For example, it can parse vehicle speed, vehicle type, device number, event type, location, and time fields uploaded from different devices into fields with unified semantics. For data in different formats, it can perform format standardization, such as converting text, structured tables, real-time streaming data, geospatial data, and 3D model data into data structures recognizable by the digital platform. For data containing null values, outliers, duplicates, or obvious errors, it can perform data cleaning, anomaly removal, missing value marking, or manual verification marking to obtain the initial observation data.

[0031] After obtaining initial observation data, the digital platform can generate multi-source operational data based on data source, acquisition time, and service type. Data source can identify whether the data originates from radar-visual fusion equipment, video systems, toll collection systems, gantry systems, meteorological systems, electromechanical systems, emergency systems, asset systems, or 3D model systems; acquisition time can be used for subsequent time alignment and time-series analysis; and service type can be used to differentiate between services such as vehicle operation, traffic flow, meteorological environment, facility status, emergency resources, asset management, and event alarms. Thus, the initial observation data possesses the fundamental attributes of traceable source, time alignment, and service classification before entering subsequent processing.

[0032] Through the above implementation methods, data scattered across different devices and business systems in highway operation and management can be uniformly accessed, and multi-source operation data can be formed through field parsing, format standardization, and source identification processing. This enables subsequent spatiotemporal mapping, twin object generation, and business collaboration to not rely on data from a single system, thereby improving data coverage and the completeness of road network operation status perception.

[0033] In some embodiments, mapping multi-source operational data to a unified road network spatiotemporal reference to obtain standardized operational data includes: performing time correction on the multi-source operational data based on the acquisition time, reception time, and data source delay information in the multi-source operational data to obtain unified time information; constructing a road network spatial mapping relationship based on road network topology information, equipment deployment location information, station number information, lane information, and three-dimensional scene coordinate information; and performing spatiotemporal mapping on the multi-source operational data based on the unified time information and the road network spatial mapping relationship to obtain standardized operational data.

[0034] Specifically, the acquisition cycle and upload method may differ for different data sources. For example, radar-visual fusion equipment can output vehicle target trajectories at a high frequency, video recognition systems may output data based on event or frame analysis results, toll stations and gantry systems typically generate passage records when vehicles pass, weather stations may upload data on a minute-by-minute basis, and electromechanical facility systems may upload operational status based on status changes or polling cycles. Therefore, the digital platform can perform time correction based on the acquisition time, data reception time, and data source latency information. Data source latency information can be determined based on equipment type, network link, transmission protocol, and historical reception delay. Through time correction, data from different sources within the same time window can be unified onto a consistent timeline, resulting in unified time information.

[0035] In terms of spatial processing, the digital platform can construct the spatial structure of the target highway network based on road network topology information. Road network topology information can include road segment connections, interchange merging and diverging relationships, toll station entrances and exits, service area entrances and exits, tunnel entrances and exits, bridge locations, number of lanes, and travel directions. Equipment deployment location information can include the installation locations, coverage areas, associated road segments, and associated station numbers of cameras, radar-visual fusion equipment, weather stations, gantries, information boards, electromechanical equipment, and emergency material warehouses. Station number information can be used to indicate the location along the longitudinal direction of the highway, lane information can be used to indicate the lane where a vehicle or event is located, and 3D scene coordinate information can be used to map the real-world road network location onto a 3D digital twin scene.

[0036] The digital platform can construct a spatial mapping relationship of the road network based on road network topology information, equipment deployment location information, station information, lane information, and 3D scene coordinate information. For data represented by latitude and longitude, it can be converted into corresponding road segments, driving directions, station numbers, and lanes; for data represented by equipment numbers, its corresponding road segments, station number intervals, and coverage areas can be determined based on the equipment deployment location information; for data represented by toll station, gantry, or service area numbers, it can be mapped to the corresponding road network nodes and travel directions; and for 3D model data, the real-world spatial location can be converted into 3D scene coordinates.

[0037] After obtaining unified time information and road network spatial mapping relationships, the digital platform can perform spatiotemporal mapping on multi-source operational data to obtain standardized operational data. Standardized operational data can include unified time, data source, business type, road segment number, driving direction, station interval, lane number, spatial coordinates, and business attributes. Through standardized operational data, it is possible to subsequently determine, on the same spatiotemporal basis, whether different observation data belong to the same vehicle, the same event, the same road segment status, or the same facility status.

[0038] The above implementation method can solve the problem that multi-source operational data are difficult to directly correlate due to different collection frequencies, timestamp accuracies, location representation methods and coordinate systems. It enables data such as radar-visual fusion, video, toll stations, gantries, meteorology, electromechanical facilities, emergency resources, road assets and 3D models to enter a unified road network spatiotemporal reference, providing a stable data foundation for subsequent object-oriented fusion and map construction.

[0039] In some embodiments, generating or updating multiple spatiotemporal twin objects based on standardized operational data includes: classifying the standardized operational data according to the operational object category corresponding to the standardized operational data to obtain multiple object observation datasets; extracting object status information based on each object observation dataset, the object status information including object spatiotemporal status, object business attributes, and data source information; and generating or updating at least one of vehicle twin objects, road segment twin objects, event twin objects, facility twin objects, environmental twin objects, and resource twin objects based on the object status information to obtain multiple spatiotemporal twin objects.

[0040] Specifically, the digital platform can determine the category of operational objects based on the business type, data source, and object identifier in standardized operational data. Operational object categories can include vehicle, road segment, event, facility, environmental, and resource categories. Vehicle data can include vehicle passage records, vehicle trajectory segments, vehicle identification results, vehicle speed, lane, vehicle type, license plate, vehicle color, and key vehicle types. Road segment data can include traffic flow, average speed, congestion status, lane occupancy, capacity, and traffic situation. Event data can include traffic accidents, vehicle malfunctions, road obstacles, abnormal parking, congestion, wrong-way driving, debris spillage, severe weather impacts, toll station queues, service area congestion, tunnel anomalies, and facility malfunctions. Facility data can include the status of facilities such as cameras, radar-guided equipment, information boards, toll collection equipment, tunnel ventilation fans, lighting equipment, broadcasting equipment, and lane control equipment. Environmental data can include temperature, humidity, wind speed, wind direction, rainfall, snowfall, visibility, and road surface conditions. Resource data can include the status of resources such as emergency supplies, rescue vehicles, obstacle clearing equipment, maintenance personnel, material warehouses, and emergency stations.

[0041] After classification, the digital platform can generate object observation datasets for each type of data. For example, a vehicle object observation dataset can be generated for multiple vehicle observation data that appear in a continuous time period for the same license plate or the same target number; a road segment object observation dataset can be generated for traffic flow, average speed, and congestion status of the same road segment within the same time window; an event object observation dataset can be generated for accidents or obstacles reported by multiple devices near the same location; and a facility object observation dataset can be generated for the operating status and fault information of the same device number.

[0042] Digital platforms can extract object status information from object observation datasets. The spatiotemporal status of an object can include the corresponding unified time, road segment, direction, station number, lane, and spatial coordinates. Object business attributes can be determined based on object category. For example, the business attributes of a vehicle twin object can include license plate, vehicle type, vehicle color, vehicle speed, key vehicle identification, and trajectory status; the business attributes of a road segment twin object can include traffic flow, average speed, congestion level, and capacity; the business attributes of an event twin object can include event type, event level, duration, affected lanes, and evidence source; the business attributes of a facility twin object can include facility type, online status, fault code, controllable status, and coverage; the business attributes of an environmental twin object can include meteorological elements and risk status; and the business attributes of a resource twin object can include resource type, inventory quantity, availability status, and dispatch location. Data source information can record which devices or business systems provided the object status for credibility assessment and traceability.

[0043] Based on object state information, the digital platform can generate new spatiotemporal twin objects or update existing spatiotemporal twin objects. For example, when a new vehicle is detected entering the target highway network, a vehicle twin object is generated; when the vehicle appears continuously in gantries, roadside radar-visual fusion devices, and video surveillance areas, the trajectory state of the vehicle twin object is updated. Similarly, when the speed on a road segment continuously decreases and multiple data sources indicate congestion, the congestion state of the road segment twin object is updated; when multiple cameras or radar-visual fusion devices detect an accident, an event twin object is generated or updated.

[0044] Through the above implementation methods, multi-source operational data can be transformed from scattered data records into spatiotemporal twin objects oriented towards highway operation objects, enabling vehicles, road sections, events, facilities, environment and resources to have a unified data expression and status update mechanism, thereby providing an object-level foundation for three-dimensional digital twin display, holographic vehicle archives, road network dynamic monitoring and cross-business collaborative handling.

[0045] In some embodiments, generating or updating at least one of vehicle twin objects, road segment twin objects, event twin objects, facility twin objects, environment twin objects, and resource twin objects based on object state information includes: determining time matching information based on the temporal proximity between the state information of the object to be merged and the generated spatiotemporal twin objects; determining spatial matching information based on the spatial proximity between the state information of the object to be merged and the generated spatiotemporal twin objects; determining attribute matching information based on the similarity of business attributes between the state information of the object to be merged and the generated spatiotemporal twin objects; determining topology matching information based on the road network topology reachability status corresponding to the state information of the object to be merged; determining data source credibility information based on the data source reliability corresponding to the state information of the object to be merged; determining object matching results based on time matching information, spatial matching information, attribute matching information, topology matching information, and data source credibility information; and updating the matched spatiotemporal twin objects or generating new spatiotemporal twin objects based on the object matching results.

[0046] Specifically, during multi-source data fusion, the status of the same vehicle, event, or facility may be repeatedly observed by multiple data sources, and inconsistent data may arise due to equipment false alarms, identification errors, positioning deviations, or transmission delays. Therefore, the digital platform can match and determine the status information of the object to be fused with the already generated spatiotemporal twin objects. The status information of the object to be fused can be one or more newly accessed or newly standardized object status information, while the already generated spatiotemporal twin objects can be vehicle twin objects, road segment twin objects, event twin objects, facility twin objects, environmental twin objects, or resource twin objects currently maintained by the digital platform.

[0047] For temporal matching information, the digital platform can compare the uniform time of the state information of the object to be fused with the difference between the most recent update time, expected arrival time, or state duration of the generated spatiotemporal twin object. For example, if a vehicle appears at the downstream gantry within a reasonable time range after appearing at the upstream radar-visual fusion device, the temporal proximity of the two is high. For spatial matching information, the digital platform can compare the spatial distance between the location of the state information of the object to be fused and the current location, coverage area, or predicted location of the generated spatiotemporal twin object. For example, if two event observations are both located near the same station interval and the same lane, the spatial proximity is high.

[0048] For attribute matching information, the digital platform can compare the similarity between business attributes. For example, vehicle twins can compare license plates, vehicle models, vehicle colors, driving directions, speed ranges, and key vehicle types; event twins can compare event types, event locations, affected lanes, severity, and associated vehicles; facility twins can compare equipment numbers, equipment types, coverage areas, and fault status. For topology matching information, the digital platform can determine whether two observations might belong to the same object based on the road network topology reachability of the target highway network. For example, whether two vehicle observation points are reachable along the same driving direction, whether there are interchange paths, and whether the distance that a vehicle can reach within the corresponding time difference is met; whether two event observations are located within the same accident impact area or adjacent road segments; and whether a facility can observe or control the road segment where the target event is located.

[0049] Regarding data source reliability information, the digital platform can determine the weight of different observation data in object matching based on the reliability of the data source. Data source reliability can be determined based on factors such as device type, device online status, historical false alarm rate, manual verification records, data completeness, time freshness, and multi-source consistency. For example, the reliability of data sources corresponding to cameras that have recently experienced frequent malfunctions can be reduced; while the reliability of event observation data supported by radar-visual fusion equipment, video recognition equipment, and manual verification can be increased.

[0050] The digital platform can determine object matching results based on temporal matching information, spatial matching information, attribute matching information, topological matching information, and data source credibility information. The object matching result indicates whether the state information of the object to be fused matches a previously generated spatiotemporal twin object, as well as the matching confidence level. When the matching confidence level meets preset matching conditions, the digital platform can merge the state information of the object to be fused into the already matched spatiotemporal twin object to update its state; when the preset matching conditions are not met, the digital platform can generate a new spatiotemporal twin object.

[0051] Through the above implementation methods, object fusion not only relies on the degree of proximity in time and space, but also incorporates the degree of similarity of business attributes, the reachability of road network topology, and the credibility of data sources. This can reduce duplicate vehicle filings, duplicate event alarms, road segment status conflicts, and erroneous associations of facility status, thereby improving the accuracy and reliability of spatiotemporal twin objects.

[0052] In some embodiments, a spatiotemporal twin object map is constructed based on the spatiotemporal and business association relationships between the spatiotemporal twin objects. This includes: determining the spatial association results between the spatiotemporal twin objects based on their spatial location and road network affiliation information; determining the temporal association results between the spatiotemporal twin objects based on their temporal state and state change sequence; determining the business association results between the spatiotemporal twin objects based on their object category and business attributes; and establishing at least one object association relationship, including location relationship, transit relationship, influence relationship, observable relationship, schedulable relationship, controllable relationship, and disposal relationship, based on the spatial association results, temporal association results, and business association results, to obtain the spatiotemporal twin object map.

[0053] Specifically, a spatiotemporal twin object graph can include multiple object nodes and object relationships. Object nodes can be vehicle twin objects, road segment twin objects, event twin objects, facility twin objects, environmental twin objects, and resource twin objects. Object relationships are used to express the spatial, temporal, and business-related relationships between different objects.

[0054] Digital platforms can determine spatial association results based on the spatial location and road network affiliation information of each spatiotemporal twin object. For example, if a vehicle twin object is located within the road segment corresponding to a road segment twin object, a locational relationship can be determined between the vehicle twin object and the road segment twin object; if an event twin object occurs within a certain station interval, a locational relationship can be determined between the event twin object and the corresponding road segment twin object; if a camera facility twin object covers a certain road segment, a coverage-related spatial association result can be determined between the facility twin object and the road segment twin object; and if an emergency resource twin object is located in a service area, management station, or material warehouse, a spatial association can be established with adjacent road segments.

[0055] Digital platforms can determine temporal correlation results based on the time status and the sequence of status changes of each spatiotemporal twin object. For example, a vehicle twin object can form a passing relationship by sequentially passing through toll stations, gantries, and the locations corresponding to multiple radar-visual fusion devices; after an event twin object occurs, the congestion level of the upstream road segment twin object increases and the average vehicle speed decreases, which can form a temporal correlation between the event and the road segment status changes; after a weather twin object shows a decrease in visibility, the event risk of a certain road segment increases, which can also form a temporal correlation.

[0056] Digital platforms can determine business association results based on object categories and business attributes. For example, an observable relationship can be established between an event twin object and a camera or radar-visual fusion device capable of observing the location of the event; a dispatchable relationship can be established between an event twin object and rescue vehicles, obstacle clearing equipment, or emergency supplies capable of reaching the location of the event; a controllable relationship can be established between an event twin object and facilities capable of performing control actions, such as information boards, lane indicators, tunnel ventilation fans, lighting equipment, and broadcasting equipment; and a disposal association relationship can be established between an event twin object and disposal actions such as video verification, traffic guidance, emergency rescue, electromechanical control, and maintenance work orders.

[0057] After establishing relationships such as location, transit, influence, observability, dispatchability, controllability, and handling, the digital platform obtains a spatiotemporal twin object map. This map can serve not only for 3D digital twin display but also for vehicle trajectory reconstruction, key vehicle monitoring, event playback, dynamic road network monitoring layer linkage, and cross-business collaborative handling. For example, when selecting an event object in a 3D scene, the digital platform can automatically display its affected road sections, affected vehicles, surrounding cameras, available emergency resources, and controllable facilities based on the map.

[0058] Through the above implementation methods, vehicles, road sections, events, facilities, environment and resources in highways can be transformed from individual display objects into object maps with calculable relationships, solving the problem of lack of unified relationship expression between different monitoring layers and business systems, thereby improving the efficiency of cross-layer analysis, event tracing and collaborative handling.

[0059] In some embodiments, when a target road network event is detected, determining event association information corresponding to the target road network event based on a spatiotemporal twin object map includes: determining a target event object in the spatiotemporal twin object map based on the target road network event; determining affected road segment objects and affected vehicle objects based on the influence and transit relationships corresponding to the target event object; determining observable facility objects based on the observable relationships corresponding to the target event object; determining schedulable resource objects and controllable facility objects based on the schedulable and controllable relationships corresponding to the target event object; and generating event association information based on at least one of the target event object, affected road segment objects, affected vehicle objects, observable facility objects, schedulable resource objects, and controllable facility objects.

[0060] Specifically, target road network events can be triggered by radar-visual fusion equipment, video event detection equipment, meteorological systems, electromechanical facility monitoring systems, toll station systems, manual reporting, or multi-source fusion analysis results. Target road network events can include traffic accidents, vehicle malfunctions, abnormal parking, road obstacles, congestion, wrong-way driving, debris spillage, severe weather risks, tunnel anomalies, toll station congestion, service area congestion, interchange merging and diverging anomalies, key vehicles deviating from their routes, key vehicles speeding, or electromechanical facility malfunctions.

[0061] Upon detecting a target road network event, the digital platform can identify the target event object within a spatiotemporal twin object map. The target event object can include event type, event location, event time, event level, affected lanes, event evidence, and ongoing status. Subsequently, the digital platform can determine the affected road segments and affected vehicles based on the influence and transit relationships corresponding to the target event object. Affected road segments can include the road segment where the target event occurs, upstream queuing spread segments, downstream capacity-restricted segments, interchange / diversion related segments, tunnel-related segments, toll station queuing impact areas, or service area entrance / exit impact areas. Affected vehicles can include vehicles near the target event, upstream queuing vehicles, priority vehicles, green channel vehicles, passenger and hazardous goods transport vehicles, oversized transport vehicles, and vehicles that may pass through the event's impact area.

[0062] The digital platform can identify observable facilities based on the observable relationships corresponding to the target event. Observable facilities can include cameras around the event point, upstream cameras, downstream cameras, tunnel cameras, radar-visual fusion equipment, video event detection equipment, and other monitoring facilities capable of covering the event location. By identifying observable facilities, the digital platform can provide direct data entry points for video verification, event playback, and command and dispatch.

[0063] The digital platform can also determine schedulable resource objects based on the schedulable relationships corresponding to the target event objects. Scheduling resource objects can include emergency supply warehouses, rescue vehicles, obstacle clearing equipment, maintenance personnel, snow removal agents, rescue equipment, and emergency stations located near the event location. For events requiring coordinated control, the digital platform can determine controllable facility objects based on controllable relationships. Controllable facility objects can include information boards, lane indicators, tollbooth lane equipment, tunnel ventilation fans, tunnel lighting, broadcasting equipment, and other electromechanical control equipment.

[0064] After obtaining the aforementioned objects, the digital platform can generate event association information. This information can include basic information about the target event object, the scope of its impact, affected road sections, affected vehicles, the status of key vehicles, observable facilities, dispatchable resources, controllable facilities, and handling priorities. This event association information can be used to subsequently generate multi-service collaborative handling strategies, as well as for layer highlighting, video linkage, event playback, and situational awareness displays in 3D digital twin scenes.

[0065] Through the above implementation methods, the target road network event is no longer just an isolated alarm, but can be automatically expanded into event-related information including affected objects, verification objects, scheduling objects, and control objects through the spatiotemporal twin object map, thereby improving the completeness of event situation analysis and the efficiency of response preparation.

[0066] In some embodiments, generating a multi-service collaborative handling strategy based on event association information includes: determining the event type, event location, impact range, and risk level of a target road network event based on the event association information to obtain an event handling scenario; generating multiple candidate service handling actions based on the event handling scenario and the associated objects in the event association information, the candidate service handling actions including at least one of video verification actions, traffic guidance actions, key vehicle monitoring actions, emergency rescue actions, electromechanical facility control actions, toll station collaborative actions, and maintenance handling actions; evaluating the handling timeliness, resource availability, traffic impact, and risk reduction effect of each candidate service handling action to obtain a handling evaluation result; and determining the target service handling action from the multiple candidate service handling actions based on the handling evaluation result to obtain a multi-service collaborative handling strategy.

[0067] Specifically, the digital platform can generate event handling scenarios based on the event type, location, impact range, and risk level in the event-related information. For example, if the target road network event is a traffic accident, the event handling scenario can include accident verification, lane closure reminders, upstream guidance, obstacle removal and rescue, and congestion relief; if the target road network event is low visibility weather risk, the event handling scenario can include speed limit reminders, information board announcements, key road section monitoring, and toll station diversion suggestions; if the target road network event is passenger vehicles or dangerous goods vehicles speeding or deviating from their routes, the event handling scenario can include key vehicle tracking, video verification, alarm push notifications, and regulatory linkage; if the target road network event is a tunnel electromechanical facility failure, the event handling scenario can include facility failure location, tunnel traffic situation monitoring, electromechanical linkage control, and maintenance work order generation.

[0068] After determining the incident handling scenario, the digital platform can generate multiple candidate business handling actions by combining the incident-related information, including affected road sections, affected vehicles, observable facilities, dispatchable resources, and controllable facilities. Video verification actions can include retrieving video footage from cameras around the incident point, upstream cameras, downstream cameras, or cameras inside the tunnel, and comparing it with incident playback or vehicle trajectory playback. Traffic guidance actions can include generating information board content, speed limit reminders, detour suggestions, lane occupancy reminders, and congestion reminders. Key vehicle monitoring actions can include trajectory tracking, anomaly alerts, and suggestions for key monitoring or interception verification of passenger vehicles, hazardous material transport vehicles, green channel vehicles, and oversized transport vehicles. Emergency rescue actions can include selecting rescue vehicles, clearing equipment, maintenance personnel, and emergency supplies, and generating dispatch routes or dispatch priorities. Electromechanical facility control actions can include suggestions for coordinated control of tunnel fans, lighting, broadcasting, lane indicators, and information boards. Toll station coordination actions can include suggestions for lane opening, queue management, traffic diversion reminders, and entrance control. Maintenance actions can include generating maintenance work orders and handling priorities for road obstacles, facility failures, slope risks, or pavement anomalies.

[0069] The digital platform can evaluate each candidate action. Action timeliness can be determined based on the action's initiation time, resource arrival time, video verification response time, or control command execution time. Resource availability can be determined based on the status of rescue vehicles, emergency supplies inventory, facility controllability, personnel scheduling, and road accessibility. Traffic impact can be determined based on the candidate action's effect on lane occupancy, toll station traffic, service area access, interchange diversion, and mainline traffic flow. Risk reduction effectiveness can be determined based on the expected reduction in secondary accident risk, congestion spread risk, key vehicle risk, or facility failure risk by the candidate action. Based on the above evaluations, the digital platform can obtain the action evaluation results.

[0070] After obtaining the evaluation results, the digital platform can determine the target business action from multiple candidate business actions and form a multi-business collaborative handling strategy. This strategy can include collaborative actions that need to be executed or confirmed by multiple business subsystems. For example, for traffic accidents, it could simultaneously include video verification, upstream information board dissemination, rescue vehicle dispatch, toll station diversion suggestions, and maintenance work order generation; for toll station congestion, it could simultaneously include lane opening suggestions, upstream traffic guidance, service area vehicle alerts, and key vehicle monitoring; for tunnel anomalies, it could simultaneously include tunnel video verification, lane control, broadcast reminders, fan or lighting linkage control, and emergency resource dispatch.

[0071] Through the above implementation methods, the digital platform can automatically form collaborative handling strategies across monitoring, traffic guidance, toll collection, emergency response, electromechanical systems, maintenance, and key vehicle supervision based on event-related information. Compared with simply displaying alarms or conducting manual analysis, this can improve the systematicness and timeliness of handling actions and reduce response delays caused by fragmented business systems.

[0072] In some embodiments, after outputting the collaborative handling information corresponding to the multi-service collaborative handling strategy, the method further includes: obtaining the handling execution data and road network recovery data corresponding to the multi-service collaborative handling strategy; determining the handling action execution result based on the handling execution data; determining the event state change result and traffic state recovery result based on the road network recovery data; and updating at least one of the following parameters based on the handling action execution result, event state change result, and traffic state recovery result: data source credibility, object matching parameter, event association information determination parameter, and collaborative handling strategy generation parameter.

[0073] Specifically, after the collaborative response information is output, the digital platform can continuously acquire response execution data. This data can include whether video verification is complete, whether information board content is published, whether electromechanical control commands are executed, whether rescue vehicles are dispatched, whether rescue vehicles have arrived at the incident location, whether maintenance work orders are issued, whether toll station lane adjustments are executed, whether key vehicle monitoring actions are completed, and manual confirmation results. This data can originate from the video platform, information board system, command and dispatch system, toll collection system, electromechanical control system, maintenance management system, emergency resource management system, or manual response terminals.

[0074] The digital platform can also acquire road network recovery data. This data can include event resolution time, vehicle speed recovery status, traffic flow recovery status, queue length changes, congestion level changes, capacity recovery status of affected road sections, status of key vehicle risk resolution, facility failure recovery status, and weather risk changes. Road network recovery data can be continuously provided by radar-visual fusion equipment, video recognition systems, toll station systems, gantry systems, meteorological systems, and electromechanical facility systems.

[0075] Based on the data from the handling and execution of actions, the digital platform can determine the results of these actions. The results can indicate whether each target business action was executed, when it was executed, whether the execution was successful, whether the target was correct, and whether there were any delays or failures. For example, a rescue vehicle dispatch action can correspond to the dispatch time, arrival time, and obstacle clearance completion time; an information board posting action can correspond to the posting time, posting content, and posting location; and a video verification action can correspond to the verification result and manual confirmation.

[0076] Based on road network recovery data, the digital platform can determine the results of event status changes and traffic recovery. Event status changes can indicate whether a target road network event has changed from unconfirmed to confirmed, from ongoing to pending, from pending to resolved, or whether the event severity has decreased. Traffic recovery results can indicate whether the average speed, traffic flow, congestion level, queue length, and capacity of the affected road segments have returned to their preset levels.

[0077] After obtaining the results of the response actions, the changes in event status, and the recovery of traffic conditions, the digital platform can update at least one of the following: data source credibility, object matching parameters, event association information determination parameters, and collaborative response strategy generation parameters. For example, if a data source generates multiple false alarms, or if its event identification results repeatedly fail after video verification, the data source credibility for that data source under the corresponding business type is reduced. If there are many errors in vehicle trajectory matching or event object matching, the relevant parameters for time matching, spatial matching, attribute matching, or topology matching are adjusted. If the actual impact range is significantly larger than the predicted impact range, the event association information determination parameters are adjusted. If a certain type of response action can significantly shorten the recovery time in similar event scenarios, the priority of strategy recommendations for that type of response action in similar scenarios is increased.

[0078] Through the above implementation methods, the digital platform can reverse the execution results and road network recovery status after the handling to the data fusion, object matching, event association and strategy generation processes. This allows the system to no longer rely on fixed rules for long-term operation, but to continuously adjust relevant parameters based on the actual handling effect, thereby improving the accuracy of subsequent event identification, situation analysis and collaborative handling.

[0079] In this embodiment, as Figure 2 As shown, a smart, high-speed, multi-source data fusion and multi-service collaborative processing digital platform is provided. This digital platform includes a multi-source data access module, a spatiotemporal benchmark mapping module, a twin object management module, an object graph construction module, an event correlation analysis module, a collaborative processing module, a 3D twin display module, and a feedback update module. Each module can be implemented as software programs, hardware devices, hardware-software hybrid modules, microservice components, or cloud-edge collaborative components.

[0080] The multi-source data access module is used to acquire multi-source operational data of the target highway network. Specifically, it can access roadside sensing data, video recognition data, toll station traffic data, gantry traffic data, meteorological monitoring data, electromechanical facility operation data, emergency resource data, road asset data, and 3D model data. The module supports both real-time and periodic data access. Real-time data can be accessed via message queues, stream processing channels, or device interfaces, while periodic data can be accessed via database synchronization, file import, or API calls. The module can also perform field parsing, format standardization, anomaly marking, and data source identification on the accessed data to obtain multi-source operational data that can be processed by subsequent modules.

[0081] The spatiotemporal reference mapping module is used to map multi-source operational data to a unified road network spatiotemporal reference, obtaining standardized operational data. Specifically, the module can perform time correction based on acquisition time, reception time, and data source delay information to obtain unified time information; it can also perform spatial mapping based on road network topology information, equipment deployment location information, station information, lane information, and 3D scene coordinate information to obtain unified spatial information corresponding to the target highway network. The standardized operational data output by the spatiotemporal reference mapping module can include unified time, road segment number, driving direction, station number, lane, spatial coordinates, service type, and data source.

[0082] The twin object management module is used to generate or update multiple spatiotemporal twin objects based on standardized operational data. Specifically, the twin object management module can generate vehicle twin objects, road segment twin objects, event twin objects, facility twin objects, environmental twin objects, and resource twin objects based on standardized operational data. For vehicle twin objects, the twin object management module can integrate data from toll stations, gantries, radar-visual fusion equipment, and video recognition to form basic vehicle information, dynamic status, and travel trajectory. For key vehicles, the twin object management module can identify passenger and hazardous goods vehicles, green channel vehicles, and oversized transport vehicles, and continuously update their location, speed, direction, and abnormal status. For road segment twin objects, the twin object management module can maintain traffic flow, average speed, congestion status, and traffic capacity. For event twin objects, the twin object management module can maintain the type, location, level, duration, and source of evidence for abnormal events. For facility twin objects, environmental twin objects, and resource twin objects, the twin object management module can respectively maintain the facility operation status, meteorological environment status, and emergency resource availability status.

[0083] The object graph construction module is used to construct a spatiotemporal twin object graph based on the spatiotemporal and business relationships between various spatiotemporal twin objects. Specifically, the object graph construction module can establish object relationships based on road segments traversed by vehicles, road segments affected by events, facility coverage, resource scheduling distance, and business handling rules. Object relationships can include location relationships, transit relationships, impact relationships, observable relationships, schedulable relationships, controllable relationships, and handling relationships. The spatiotemporal twin object graph output by the object graph construction module can provide a foundation for association queries for the event association analysis module and the 3D twin display module.

[0084] The event correlation analysis module is used to determine the event correlation information corresponding to the target road network event based on the spatiotemporal twin object map when a target road network event is detected. Specifically, the event correlation analysis module can locate the target event object based on the target road network event, and determine the affected road segment object, affected vehicle object, observable facility object, schedulable resource object, and controllable facility object, thereby generating event correlation information. The event correlation information can be used to support event playback, video linkage, impact range display, resource scheduling, and response strategy generation.

[0085] The collaborative response module generates multi-service collaborative response strategies based on event correlation information and outputs corresponding collaborative response information. Specifically, the module can generate video verification actions, traffic guidance actions, key vehicle monitoring actions, emergency rescue actions, electromechanical facility control actions, toll station coordination actions, and maintenance actions based on event type, location, impact scope, risk level, and associated objects. It then evaluates candidate actions to determine the final multi-service collaborative response strategy. Collaborative response information can be output to video platforms, information board systems, command and dispatch systems, toll collection systems, electromechanical control systems, maintenance management systems, and emergency resource management systems. It can also be output in the form of prompts, dispatch suggestions, linkage control commands, work orders, or visual alarms.

[0086] The 3D twin display module is used to generate 3D twin display information of the target highway network based on spatiotemporal twin objects, spatiotemporal twin object maps, event correlation information, and multi-service collaborative processing strategies. Specifically, the 3D twin display module can load the 3D model of the highway based on a real-time 3D engine and map vehicle twin objects, road segment twin objects, event twin objects, facility twin objects, environmental twin objects, and resource twin objects into the 3D scene. The 3D twin display module can display the digital effect of traffic flow, continuous lane-level positioning and driving trajectory of vehicles, traffic flow monitoring layers, meteorological monitoring layers, abnormal event monitoring layers, electromechanical facility monitoring layers, emergency material monitoring layers, key vehicle monitoring layers, and traffic situation awareness layers for key areas such as tunnels, toll stations, service areas, and interchanges. The 3D twin display module also supports driver's perspective, following vehicle patrol, lane patrol, event playback, vehicle trajectory tracing, and surrounding video linkage display.

[0087] The feedback update module is used to update the spatiotemporal twin objects, spatiotemporal twin object maps, and / or the strategy generation parameters of the collaborative handling module based on the handling feedback data corresponding to the multi-service collaborative handling strategy. Specifically, the feedback update module can obtain handling execution data, road network restoration data, manual confirmation results, and handling evaluation results, and update the data source credibility, object matching parameters, event association information determination parameters, and collaborative handling strategy generation parameters accordingly. The feedback update module can also synchronize the post-handling event status, road segment status, facility status, and resource status back to the twin object management module, enabling the 3D twin display module and object map construction module to reflect the latest operational status.

[0088] Through the aforementioned digital platform, the multi-source data access module, spatiotemporal reference mapping module, twin object management module, object graph construction module, event correlation analysis module, collaborative handling module, 3D twin display module, and feedback update module form a closed-loop structure from data access, spatiotemporal fusion, object modeling, graph correlation, event analysis, collaborative handling, 3D display to feedback optimization. This can enhance the intelligent highway digital platform's real-time perception of road network operation status, cross-business correlation analysis capabilities, 3D visualization command capabilities, and collaborative handling capabilities.

[0089] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

Claims

1. A method for intelligent high-speed multi-source data fusion and multi-service collaborative processing, characterized in that, The method includes: Acquire multi-source operational data of the target highway network; The multi-source operational data is mapped to a unified road network spatiotemporal reference to obtain standardized operational data. Multiple spatiotemporal twin objects are generated or updated based on the standardized operational data, and the multiple spatiotemporal twin objects are used to characterize different operational objects in the target highway network; Based on the spatiotemporal and business relationships between the spatiotemporal twin objects, a spatiotemporal twin object graph is constructed; Upon detecting a target road network event, event association information corresponding to the target road network event is determined based on the spatiotemporal twin object map. A multi-service collaborative handling strategy is generated based on the event association information, and collaborative handling information corresponding to the multi-service collaborative handling strategy is output.

2. The intelligent high-speed multi-source data fusion and multi-service collaborative processing method according to claim 1, characterized in that, The acquisition of multi-source operational data of the target highway network includes: Initial access data is obtained by accessing at least two types of data from the target highway network, including roadside perception data, video recognition data, toll station traffic data, gantry traffic data, meteorological monitoring data, electromechanical facility operation data, emergency resource data, road asset data, and 3D model data. The initial access data is parsed and format standardized to obtain the initial observation data; The multi-source operational data is generated based on the data source, collection time, and service type corresponding to the initial observation data.

3. The intelligent high-speed multi-source data fusion and multi-service collaborative processing method according to claim 2, characterized in that, The process of mapping the multi-source operational data to a unified road network spatiotemporal reference to obtain standardized operational data includes: Based on the acquisition time, reception time, and data source delay information in the multi-source operational data, time correction is performed on the multi-source operational data to obtain unified time information; Based on road network topology information, equipment deployment location information, station information, lane information, and 3D scene coordinate information, construct the road network spatial mapping relationship; Based on the unified time information and the road network spatial mapping relationship, the multi-source operation data is spatiotemporally mapped to obtain the standardized operation data.

4. The intelligent high-speed multi-source data fusion and multi-service collaborative processing method according to claim 1, characterized in that, The generation or updating of multiple spatiotemporal twin objects based on the standardized operational data includes: Based on the category of the operational object corresponding to the standardized operational data, the standardized operational data is classified to obtain multiple object observation datasets; Based on the object observation datasets, object status information is extracted, including object spatiotemporal status, object business attributes, and data source information; Based on the object state information, at least one of the following is generated or updated: vehicle twin object, road segment twin object, event twin object, facility twin object, environment twin object, and resource twin object, to obtain the plurality of spatiotemporal twin objects.

5. The intelligent high-speed multi-source data fusion and multi-service collaborative processing method according to claim 4, characterized in that, The generation or updating of at least one of the following based on the object state information: vehicle twin object, road segment twin object, event twin object, facility twin object, environment twin object, and resource twin object, includes: The time matching information is determined based on the temporal proximity between the state information of the object to be merged and the already generated spatiotemporal twin object; Spatial matching information is determined based on the spatial proximity between the state information of the object to be fused and the generated spatiotemporal twin object. Attribute matching information is determined based on the degree of similarity of business attributes between the state information of the object to be merged and the generated spatiotemporal twin object; Based on the road network topology reachability status corresponding to the status information of the object to be merged, determine the topology matching information; Based on the reliability of the data source corresponding to the state information of the object to be merged, the data source credibility information is determined; Based on the time matching information, the spatial matching information, the attribute matching information, the topology matching information, and the data source credibility information, the object matching result is determined; Update the matched spatiotemporal twin objects based on the object matching results, or generate new spatiotemporal twin objects.

6. The intelligent high-speed multi-source data fusion and multi-service collaborative processing method according to claim 4, characterized in that, The step of constructing a spatiotemporal twin object graph based on the spatiotemporal and business relationships between the spatiotemporal twin objects includes: Based on the spatial location and road network affiliation information of each spatiotemporal twin object, the spatial association results between each spatiotemporal twin object are determined; Based on the temporal state and state change sequence of each spatiotemporal twin object, the temporal correlation result between each spatiotemporal twin object is determined; Based on the object category and business attributes of each spatiotemporal twin object, determine the business association results between each spatiotemporal twin object; Based on the spatial association results, the temporal association results, and the business association results, at least one object association relationship is established, including the location relationship, the passing relationship, the influence relationship, the observable relationship, the schedulable relationship, the controllable relationship, and the disposal relationship, to obtain the spatiotemporal twin object map.

7. The intelligent high-speed multi-source data fusion and multi-service collaborative processing method according to claim 6, characterized in that, The step of determining the event association information corresponding to the target road network event based on the spatiotemporal twin object map when a target road network event is detected includes: The target event object is determined in the spatiotemporal twin object map based on the target road network event; Based on the influence and transit relationships corresponding to the target event objects, the affected road segment objects and affected vehicle objects are determined; Based on the observable relationships corresponding to the target event objects, the observable facility objects are determined; Based on the schedulable and controllable relationships corresponding to the target event objects, determine the schedulable resource objects and controllable facility objects; The event association information is generated based on at least one of the target event object, the affected road segment object, the affected vehicle object, the observable facility object, the schedulable resource object, and the controllable facility object.

8. The intelligent high-speed multi-source data fusion and multi-service collaborative processing method according to claim 7, characterized in that, The step of generating a multi-service collaborative handling strategy based on the event association information includes: Based on the event association information, the event type, event location, scope of impact, and risk level of the target road network event are determined to obtain the event handling scenario; Based on the event handling scenario and the associated objects in the event association information, multiple candidate business handling actions are generated. The candidate business handling actions include at least one of the following: video verification action, traffic guidance action, key vehicle monitoring action, emergency rescue action, electromechanical facility control action, toll station coordination action, and maintenance handling action. The timeliness, resource availability, traffic impact, and risk reduction effect of each candidate business action are evaluated to obtain the evaluation results. Based on the evaluation results, the target business action is determined from the multiple candidate business action actions to obtain the multi-business collaborative action strategy.

9. The intelligent high-speed multi-source data fusion and multi-service collaborative processing method according to claim 8, characterized in that, After outputting the collaborative processing information corresponding to the multi-service collaborative processing strategy, the method further includes: Obtain the processing execution data and road network recovery data corresponding to the multi-service collaborative processing strategy; The execution result of the disposal action is determined based on the disposal execution data; The event status change results and traffic status recovery results are determined based on the road network recovery data. Based on the execution results of the handling actions, the results of the event status changes, and the results of the traffic status recovery, update at least one of the following: data source credibility, object matching parameters, event association information determination parameters, and collaborative handling strategy generation parameters.

10. A smart, high-speed, multi-source data fusion and multi-service collaborative processing digital platform, characterized in that, include: The multi-source data access module is used to acquire multi-source operational data of the target highway network; The spatiotemporal reference mapping module is used to map the multi-source operational data to a unified road network spatiotemporal reference to obtain standardized operational data; The twin object management module is used to generate or update multiple spatiotemporal twin objects based on the standardized operational data. The multiple spatiotemporal twin objects are used to characterize different operational objects in the target highway network. The object graph construction module is used to construct a spatiotemporal twin object graph based on the spatiotemporal and business relationships between the spatiotemporal twin objects. The event association analysis module is used to determine the event association information corresponding to the target road network event based on the spatiotemporal twin object map when a target road network event is detected. The collaborative handling module is used to generate a multi-service collaborative handling strategy based on the event association information, and output collaborative handling information corresponding to the multi-service collaborative handling strategy; The 3D twin display module is used to generate 3D twin display information of the target highway network based on the spatiotemporal twin object, the spatiotemporal twin object map, the event association information, and the multi-service collaborative processing strategy. The feedback update module is used to update the spatiotemporal twin object, the spatiotemporal twin object map, and / or the strategy generation parameters of the collaborative processing module based on the processing feedback data corresponding to the multi-service collaborative processing strategy.

Images