Construction scene traffic diversion prediction system and method based on digital twinborn technology

Abstract

The invention discloses a construction scene traffic diversion prediction system and method based on a digital twinborn technology, and belongs to the technical field of intelligent traffic. A data acquisition module acquires historical traffic flow data, road network data around a construction area and data of a plurality of use sensors deployed in the construction area; the feature extraction module performs time dimension extraction on historical flow data to obtain a time dimension sequence and a trip chain sequence, the flow prediction module trains a traffic flow prediction model, predicts future traffic flow by using the trained model, updates a pre-constructed digital twinborn model according to prediction data, and obtains a traffic flow prediction result; and a traffic diversion strategy is formulated. Prediction is carried out based on real-time flow data and vehicle data, the time dimension and the space dimension of the real-time flow data are combined, real-time dynamic accurate prediction is achieved, meanwhile, the digital twin model is adopted to carry out visual display on field data, and the digital twin model is dynamically updated in real time based on environment data and the flow data.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent transportation technology, and specifically relates to a construction scene traffic diversion prediction system and method based on digital twin technology. Background Art

[0002] With the acceleration of urbanization and the continued rise in motor vehicle ownership, road congestion and traffic jams are becoming increasingly severe. This is especially true during road maintenance and construction work, where fencing is often erected to close off sections of the road, exacerbating congestion. However, traditional diversion methods rely heavily on manual control, which is overly reliant on individual experience, lacks scientific accuracy, and lacks real-time, dynamic forecasting capabilities, making it difficult to respond to emergencies. Furthermore, the development and adjustment of diversion plans is inefficient, making it difficult to adapt to rapidly changing traffic demands.

[0003] While the development of artificial intelligence and big data technologies has enabled intelligent methods to divert traffic around construction sites, these methods still have drawbacks, such as low real-time dynamic prediction accuracy and a lack of visual operation. Therefore, there is an urgent need for an effective technical solution to address these issues with existing technologies. Summary of the Invention

[0004] The purpose of the present invention is to provide a construction scene traffic diversion prediction system and method based on digital twin technology to solve the above-mentioned problems existing in the prior art.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] In the first aspect, the present invention provides a construction scene traffic diversion prediction system based on digital twin technology, including a data acquisition module, a feature extraction module, a traffic flow prediction module and a model update module:

[0007] The data acquisition module is used to obtain historical traffic flow data, road network data around the construction area, and multiple real-time sensor data deployed in the construction area, the historical traffic flow data including historical traffic flow data and historical vehicle data, the real-time sensor data including real-time traffic flow data, real-time vehicle data, and real-time environmental data, upload the historical traffic flow data and historical vehicle data to the feature extraction module, upload the road network data around the construction area, real-time traffic flow data, and real-time vehicle data to the traffic prediction module, and upload the real-time traffic flow data and real-time environmental data to the model update module;

[0008] The feature extraction module is used to extract the time dimension of historical traffic data to obtain a time dimension sequence, obtain a vehicle trip chain based on the historical vehicle data, and use the vehicle trip chain to represent the location and time stamp set of sensors that the vehicle passes through during driving. The vehicle trip chains are sorted according to chronological order to obtain a trip chain sequence, and the time dimension sequence and trip chain sequence are uploaded to the traffic prediction module;

[0009] The traffic flow prediction module is used to input the time dimension sequence, the trip chain sequence and the road network data around the construction area into the traffic flow prediction model for training, obtain the trained traffic flow prediction model, obtain the real-time trip chain sequence based on the real-time vehicle data, input the real-time traffic flow data and the real-time trip chain sequence into the trained traffic flow prediction model for prediction, output the predicted traffic flow, and upload the predicted traffic flow to the model update module;

[0010] The model update module is used to obtain real-time road data, update the pre-built digital twin model based on the real-time road data, real-time environmental data and predicted traffic, and formulate a traffic diversion strategy based on the real-time traffic data and predicted traffic.

[0011] In one possible design, a model building module is also included for obtaining aerial photography information and construction area drawings, and building a digital twin model based on road network data around the construction area, aerial photography information and construction area drawings, wherein the road network data around the construction area includes construction area road information.

[0012] In one possible design, the data acquisition module is also used to perform data cleaning, missing value filling, standardization and normalization on the historical traffic flow data, the road network data around the construction area and the multiple real-time sensor data deployed in the construction area after acquiring the historical traffic flow data, the road network data around the construction area and the multiple real-time sensor data deployed in the construction area.

[0013] In one possible design, a vehicle travel chain is obtained based on historical vehicle data, including:

[0014] Obtain historical travel chains of different vehicles based on historical vehicle data;

[0015] The SOM algorithm is used to cluster the historical travel chains of different vehicles to obtain multiple clustered historical travel chains;

[0016] The clustered historical travel chains are classified to obtain vehicle travel chains.

[0017] In one possible design, the traffic flow prediction model is constructed based on the spatiotemporal convolution algorithm. The traffic flow prediction model includes an attention module and a spatiotemporal convolution module. The attention module is used to extract time dimension features, and the spatiotemporal convolution module is used to extract space dimension features and combine the time dimension features and space dimension features.

[0018] In one possible design, the time dimension sequence includes an adjacent time dimension sequence, a daily time dimension sequence, and a weekly time dimension sequence; and inputting the time dimension sequence, the trip chain sequence, and the road network data around the construction area into the traffic flow prediction model for training includes:

[0019] Feature extraction is performed on the adjacent time dimension series, daily time dimension series, weekly time dimension series, trip chain series, and road network data around the construction area to obtain time dimension features and spatial dimension features. The time dimension features include adjacent time features, daily time features, and weekly time features, and the spatial dimension features include trip chain features and road features.

[0020] Based on the graph convolution operation, the time dimension features and the spatial dimension features are combined to obtain the combined features, and the convolution operation is performed on the combined features to output the feature map;

[0021] The feature map is input into the traffic flow prediction model for training.

[0022] In one possible design, traffic diversion strategies are developed based on real-time traffic data and predicted traffic, including:

[0023] Obtain the preset diversion category table, input the predicted traffic into the trained random forest model, and obtain the traffic flow label;

[0024] Query the preset diversion category table based on the traffic flow label to obtain the corresponding diversion method;

[0025] Develop traffic diversion strategies based on corresponding diversion measures and real-time traffic data.

[0026] In a second aspect, the present invention provides a construction scene traffic diversion prediction method based on digital twin technology, comprising:

[0027] Acquire historical traffic flow data, road network data around the construction area, and multiple real-time sensor data deployed in the construction area, wherein the historical traffic flow data includes historical traffic flow data and historical vehicle data, and the real-time sensor data includes real-time traffic flow data, real-time vehicle data, and real-time environmental data;

[0028] The time dimension of historical traffic data is extracted to obtain a time dimension sequence. Based on the historical vehicle data, a vehicle trip chain is obtained. The vehicle trip chain is used to represent the location and time stamp set of sensors that the vehicle passes while driving. The vehicle trip chains are sorted according to the chronological order to obtain a trip chain sequence.

[0029] Input the time dimension sequence, trip chain sequence and road network data around the construction area into the traffic flow prediction model for training, obtain the trained traffic flow prediction model, obtain the real-time trip chain sequence based on the real-time vehicle data, input the real-time traffic flow data and the real-time trip chain sequence into the trained traffic flow prediction model for prediction, and output the predicted traffic flow;

[0030] Obtain real-time road data, update the pre-built digital twin model based on real-time road data, real-time environmental data and predicted traffic, and formulate traffic diversion strategies based on real-time traffic data and predicted traffic.

[0031] In the third aspect, the present invention provides a construction scene traffic diversion prediction device based on digital twin technology, comprising a memory, a processor and a transceiver that are communicatively connected in sequence, wherein the memory is used to store computer programs, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the construction scene traffic diversion prediction method based on digital twin technology as described in the second aspect.

[0032] In a fourth aspect, the present invention provides a computer program product comprising instructions, which, when executed on a computer, enables the computer to execute the construction scene traffic diversion prediction method based on digital twin technology as described in the second aspect.

[0033] The beneficial effects of the present invention are as follows:

[0034] The present invention discloses a construction scene traffic diversion prediction system and method based on digital twin technology, wherein the data acquisition module acquires historical traffic flow data, road network data around the construction area, and data from multiple sensors deployed in the construction area; the feature extraction module extracts the time dimension of the historical traffic data to obtain a time dimension sequence; the historical vehicle data is sorted by travel chains to obtain a travel chain sequence; the traffic prediction module inputs the time dimension sequence, the travel chain sequence, and the road network data around the construction area into a traffic flow prediction model for training; the trained traffic flow prediction model is used to predict future traffic flow; the pre-built digital twin model is updated according to the prediction results; and a traffic diversion strategy is formulated according to the prediction results. The present invention adopts a traffic flow prediction model to make predictions based on real-time traffic data and vehicle data, combines the time dimension and spatial dimension of real-time traffic data, and realizes real-time dynamic and accurate predictions. At the same time, a digital twin model is used to visualize the on-site data, and the digital twin model is updated in real time based on environmental data and traffic data, so that the digital twin model is more in line with the actual situation and has broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 A module diagram of a construction scene traffic diversion prediction system based on digital twin technology provided in the first aspect of the embodiment;

[0036] Figure 2 A flowchart of a construction scene traffic diversion prediction method based on digital twin technology is provided for the second aspect of the embodiment. DETAILED DESCRIPTION

[0037] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the present invention will be briefly introduced below in conjunction with the drawings and the description of the embodiments or the prior art. Obviously, the following description of the structure of the drawings is only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work. It should be noted that the description of these embodiments is used to help understand the present invention, but does not constitute a limitation of the present invention.

[0038] Example:

[0039] like Figure 1 As shown, the first aspect of this embodiment provides a construction scene traffic diversion prediction system based on digital twin technology, including a data acquisition module, a feature extraction module, a traffic flow prediction module and a model update module;

[0040] The data acquisition module is used to obtain historical traffic flow data, road network data around the construction area, and multiple real-time sensor data deployed in the construction area, the historical traffic flow data including historical traffic flow data and historical vehicle data, the real-time sensor data including real-time traffic flow data, real-time vehicle data, and real-time environmental data, upload the historical traffic flow data and historical vehicle data to the feature extraction module, upload the road network data around the construction area, real-time traffic flow data, and real-time vehicle data to the traffic prediction module, and upload the real-time traffic flow data and real-time environmental data to the model update module;

[0041] The feature extraction module is used to extract the time dimension of historical traffic data to obtain a time dimension sequence, obtain a vehicle trip chain based on the historical vehicle data, and use the vehicle trip chain to represent the location and time stamp set of sensors that the vehicle passes through during driving. The vehicle trip chains are sorted according to chronological order to obtain a trip chain sequence, and the time dimension sequence and trip chain sequence are uploaded to the traffic prediction module;

[0042] The traffic flow prediction module is used to input the time dimension sequence, the trip chain sequence and the road network data around the construction area into the traffic flow prediction model for training, obtain the trained traffic flow prediction model, obtain the real-time trip chain sequence based on the real-time vehicle data, input the real-time traffic flow data and the real-time trip chain sequence into the trained traffic flow prediction model for prediction, output the predicted traffic flow, and upload the predicted traffic flow to the model update module;

[0043] The model update module is used to obtain real-time road data, update the pre-built digital twin model based on the real-time road data, real-time environmental data and predicted traffic, and formulate a traffic diversion strategy based on the real-time traffic data and predicted traffic.

[0044] Furthermore, the sensors deployed in the construction area include but are not limited to license plate recognition sensors, speed sensors, environmental sensors and geomagnetic sensors. Among them, the license plate recognition sensor is used to identify license plate information and distinguish different vehicles; the speed sensor is used to measure the driving speed of passing vehicles, and the driving speed is inversely proportional to the traffic flow; the environmental sensor is used to measure various parameters of the environment, among which the measured parameters include temperature, humidity, air pressure, wind speed, wind direction and precipitation; the geomagnetic sensor is used to detect whether there is a vehicle and identify the vehicle model, etc.

[0045] In one possible design, a model building module is also included for obtaining aerial photography information and construction area drawings, and building a digital twin model based on road network data around the construction area, aerial photography information and construction area drawings, wherein the road network data around the construction area includes construction area road information.

[0046] Specifically, by constructing a digital twin model, the construction area and the scenes around the construction area are visualized. The digital twin model presents a visual interface, which enables users to obtain vehicle flow information in the construction area more intuitively. The road signs in the visual interface are marked with corresponding colors according to the size of the vehicle flow. In this embodiment, green, yellow and red can be used as examples, indicating small flow, moderate flow and large flow, respectively, so as to implement diversion strategies for relevant roads.

[0047] In one possible design, the data acquisition module is also used to perform data cleaning, missing value filling, standardization and normalization on the historical traffic flow data, the road network data around the construction area and the multiple real-time sensor data deployed in the construction area after acquiring the historical traffic flow data, the road network data around the construction area and the multiple real-time sensor data deployed in the construction area.

[0048] Specifically, before performing data cleaning on the above data, receive the formulated data cleaning rules, delete or fill missing values, delete or replace outliers, and standardize and normalize the crawled results according to the data cleaning rules to ensure the integrity and accuracy of the data and improve the generalization ability of subsequent models. Standardization and normalization processing helps to improve the accuracy and reliability of the data.

[0049] In one possible design, a vehicle travel chain is obtained based on historical vehicle data, including:

[0050] Obtain historical travel chains of different vehicles based on historical vehicle data;

[0051] The SOM algorithm is used to cluster the historical travel chains of different vehicles to obtain multiple clustered historical travel chains;

[0052] The clustered historical travel chains are classified to obtain vehicle travel chains.

[0053] Specifically, the SOM algorithm (Self-Origanizing Maps) is an unsupervised learning algorithm. Its algorithm principle is to map the input data into a two-dimensional or three-dimensional grid structure to show the topological relationship and clustering structure between the data.

[0054] In one possible design, the traffic flow prediction model is constructed based on the spatiotemporal convolution algorithm. The traffic flow prediction model includes an attention module and a spatiotemporal convolution module. The attention module is used to extract time dimension features, and the spatiotemporal convolution module is used to extract space dimension features and combine the time dimension features and space dimension features.

[0055] Specifically, the Space-Time Convolutional Neural Networks (ST-CNNs) is a neural network model that captures the temporal and spatial variation characteristics of data by simultaneously applying convolution operations in both time and space dimensions.

[0056] In one possible design, the time dimension sequence includes an adjacent time dimension sequence, a daily time dimension sequence, and a weekly time dimension sequence; and inputting the time dimension sequence, the trip chain sequence, and the road network data around the construction area into the traffic flow prediction model for training includes:

[0057] Feature extraction is performed on the adjacent time dimension series, daily time dimension series, weekly time dimension series, trip chain series, and road network data around the construction area to obtain time dimension features and spatial dimension features. The time dimension features include adjacent time features, daily time features, and weekly time features, and the spatial dimension features include trip chain features and road features.

[0058] Based on the graph convolution operation, the time dimension features and the spatial dimension features are combined to obtain the combined features, and the convolution operation is performed on the combined features to output the feature map;

[0059] The feature map is input into the traffic flow prediction model for training.

[0060] In one possible design, a traffic diversion strategy is formulated based on real-time traffic data and predicted traffic, including but not limited to the following steps:

[0061] Obtain the preset diversion category table, input the predicted traffic into the trained random forest model, and obtain the traffic flow label;

[0062] Query the preset diversion category table based on the traffic flow label to obtain the corresponding diversion method;

[0063] Develop traffic diversion strategies based on corresponding diversion measures and real-time traffic data.

[0064] Furthermore, the preset diversion category table is manually set and includes diversion methods equipped according to different traffic flow information. For example, when the predicted traffic volume is medium, alternative routes can be recommended to the driver on various navigation apps. By adjusting the vehicle's navigation route, the construction area can be bypassed to achieve the purpose of reducing vehicle flow. When the predicted traffic volume is high, the construction area and the vicinity of the construction area can be adjusted by adjusting traffic signals, and temporary traffic signs can be set up in the construction area to notify relevant personnel to manually evacuate vehicles, thereby achieving the purpose of diversion. Combining the above diversion methods with real-time traffic data can improve the accuracy of the predicted traffic flow data. At the same time, the diversion methods can be adjusted in real time based on real-time traffic data, improving the practicality of the traffic diversion strategy.

[0065] This embodiment discloses a construction scene traffic diversion prediction system based on digital twin technology, including a data acquisition module, a feature extraction module, a traffic prediction module and a model update module. The data acquisition module acquires historical traffic flow data, road network data around the construction area and multiple real-time sensor data deployed in the construction area. The feature extraction module extracts the historical traffic flow data to obtain a time dimension sequence and a vehicle travel chain. The vehicle travel chain is used to characterize the location and time stamp set of sensors passed by the vehicle during driving. The vehicle travel chain is sorted according to the chronological order of the timestamps to obtain a travel chain sequence. The traffic prediction module inputs the time dimension sequence, the travel chain sequence and the road network data around the construction area into a traffic flow prediction model for training to obtain a trained traffic flow prediction model. A real-time travel chain sequence is obtained based on real-time vehicle data. Prediction is performed based on the real-time traffic data and the real-time travel chain sequence to obtain a prediction result. The pre-built digital twin model is updated according to the real-time road data, real-time environmental data and predicted traffic, and a traffic diversion strategy is formulated. In this embodiment, the traffic flow prediction model is constructed through a spatiotemporal convolution algorithm, which can combine data in the time dimension with data in the space dimension. The trained traffic flow prediction model makes predictions based on real-time traffic data and real-time travel chain sequences to obtain real-time dynamic and accurate prediction results. The digital twin model is used to visualize the construction area and dynamically update it in real time based on environmental data, thereby enhancing the user's practical experience and allowing users to more intuitively see the changes in traffic in the construction area.

[0066] like Figure 2 As shown, the second aspect of this embodiment provides a construction scene traffic diversion prediction method based on digital twin technology, including but not limited to the following steps:

[0067] S1. Acquire historical traffic flow data, road network data around the construction area, and multiple real-time sensor data deployed in the construction area. The historical traffic flow data includes historical traffic flow data and historical vehicle data. The real-time sensor data includes real-time traffic data, real-time vehicle data, and real-time environmental data.

[0068] S2. Extract the time dimension from the historical traffic data to obtain a time dimension sequence. Based on the historical vehicle data, obtain a vehicle trip chain. The vehicle trip chain represents the location and timestamp set of sensors that a vehicle passes through during travel. The vehicle trip chains are sorted chronologically to obtain a trip chain sequence.

[0069] S3. Input the time dimension sequence, trip chain sequence, and road network data around the construction area into the traffic flow prediction model for training, obtaining a trained traffic flow prediction model. Based on the real-time vehicle data, obtain a real-time trip chain sequence. Input the real-time traffic flow data and the real-time trip chain sequence into the trained traffic flow prediction model for prediction, and output the predicted traffic flow.

[0070] S4. Obtain real-time road data, update the pre-built digital twin model based on the real-time road data, real-time environmental data, and predicted traffic flow, and formulate a traffic diversion strategy based on the real-time traffic data and predicted traffic flow.

[0071] This embodiment discloses a construction scene traffic diversion prediction method based on digital twin technology, including obtaining historical traffic flow data, road network data around the construction area, and multiple real-time sensor data deployed in the construction area; extracting the time dimension of the historical flow data to obtain a time dimension sequence, obtaining a vehicle trip chain based on the historical vehicle data, wherein the vehicle trip chain is used to represent the sensor location and time stamp set of the vehicle passing through, and sorting the vehicle trip chains according to the chronological order to obtain a trip chain sequence; inputting the time dimension sequence, the trip chain sequence, and the road network data around the construction area into a traffic flow prediction model for training to obtain a trained traffic flow prediction model, obtaining a real-time trip chain sequence based on real-time vehicle data, inputting the real-time flow data and the real-time trip chain sequence into the trained traffic flow prediction model for prediction, and outputting predicted flow; obtaining real-time road data, updating a pre-built digital twin model based on the real-time road data, real-time environmental data, and predicted flow, and formulating a traffic diversion strategy based on the real-time flow data and predicted flow. The present invention obtains historical time dimension data and spatial dimension data by performing feature extraction operations of different dimensions on historical data, and trains a model based on the historical time dimension data and spatial dimension data, thereby improving the accuracy and practicality of the model.

[0072] A third aspect of this embodiment provides a construction scene traffic diversion prediction device based on digital twin technology, wherein the computer-readable storage medium stores instructions, and when the instructions are executed on a computer, they are used to execute the construction scene traffic diversion prediction method based on digital twin technology as described in the second aspect of the embodiment. The computer-readable storage medium refers to a carrier for storing data, and may include, but is not limited to, computer-readable storage media such as a floppy disk, an optical disk, a hard disk, a flash memory, a USB flash drive, and / or a memory stick. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device.

[0073] The working process, working details and technical effects of the aforementioned computer-readable storage medium provided in the third aspect of this embodiment can be found in the construction scene traffic diversion prediction method based on digital twin technology as described in the second aspect, and will not be repeated here.

[0074] The fourth aspect of this embodiment provides a computer program product, including a computer program or instructions. When the computer program or instructions are executed by a computer, they are used to implement the construction scene traffic diversion prediction method based on digital twin technology as described in the second aspect of the embodiment.

[0075] The working process, working details and technical effects of the aforementioned computer program product provided in the fourth aspect of this embodiment can be found in the construction scene traffic diversion prediction method based on digital twin technology as described in the second aspect, and will not be repeated here.

[0076] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included in the scope of protection of the present invention.

Claims

1. A construction scene traffic diversion prediction system based on digital twin technology, characterized by: It includes data acquisition module, feature extraction module, traffic prediction module and model update module; The data acquisition module is used to obtain historical traffic flow data, road network data around the construction area, and multiple real-time sensor data deployed in the construction area, the historical traffic flow data including historical traffic flow data and historical vehicle data, the real-time sensor data including real-time traffic flow data, real-time vehicle data, and real-time environmental data, upload the historical traffic flow data and historical vehicle data to the feature extraction module, upload the road network data around the construction area, real-time traffic flow data, and real-time vehicle data to the traffic prediction module, and upload the real-time traffic flow data and real-time environmental data to the model update module; The feature extraction module is used to extract the time dimension of historical traffic data to obtain a time dimension sequence, obtain a vehicle trip chain based on the historical vehicle data, and use the vehicle trip chain to represent the location and time stamp set of sensors that the vehicle passes through during driving. The vehicle trip chains are sorted according to chronological order to obtain a trip chain sequence, and the time dimension sequence and trip chain sequence are uploaded to the traffic prediction module; The traffic flow prediction module is used to input the time dimension sequence, the trip chain sequence and the road network data around the construction area into the traffic flow prediction model for training, obtain the trained traffic flow prediction model, obtain the real-time trip chain sequence based on the real-time vehicle data, input the real-time traffic flow data and the real-time trip chain sequence into the trained traffic flow prediction model for prediction, output the predicted traffic flow, and upload the predicted traffic flow to the model update module; The model update module is used to obtain real-time road data, update the pre-built digital twin model based on the real-time road data, real-time environmental data and predicted traffic, and formulate a traffic diversion strategy based on the real-time traffic data and predicted traffic.

2. The construction scene traffic diversion prediction system based on digital twin technology according to claim 1 is characterized in that: It also includes a model building module for obtaining aerial photography information and construction area drawings, and building a digital twin model based on the road network data around the construction area, aerial photography information and construction area drawings. The road network data around the construction area includes construction area road information.

3. The construction scene traffic diversion prediction system based on digital twin technology according to claim 1 is characterized in that: The data acquisition module is also used to perform data cleaning, missing value filling, standardization and normalization on the historical traffic flow data, the road network data around the construction area and the multiple real-time sensor data deployed in the construction area after acquiring the historical traffic flow data, the road network data around the construction area and the multiple real-time sensor data deployed in the construction area.

4. The construction scene traffic diversion prediction system based on digital twin technology according to claim 1 is characterized in that: Based on historical vehicle data, a vehicle travel chain is obtained, including: Obtain historical travel chains of different vehicles based on historical vehicle data; The SOM algorithm is used to cluster the historical travel chains of different vehicles to obtain multiple clustered historical travel chains; The clustered historical travel chains are classified to obtain vehicle travel chains.

5. The construction scene traffic diversion prediction system based on digital twin technology according to claim 1 is characterized in that: The traffic flow prediction model is constructed based on the spatiotemporal convolution algorithm. The traffic flow prediction model includes an attention module and a spatiotemporal convolution module. The attention module is used to extract time dimension features, and the spatiotemporal convolution module is used to extract space dimension features and combine the time dimension features and space dimension features.

6. The construction scene traffic diversion prediction system based on digital twin technology according to claim 5 is characterized in that: The time dimension sequence includes an adjacent time dimension sequence, a daily time dimension sequence, and a weekly time dimension sequence; inputting the time dimension sequence, the trip chain sequence, and the road network data around the construction area into the traffic flow prediction model for training includes: Feature extraction is performed on the adjacent time dimension series, daily time dimension series, weekly time dimension series, trip chain series, and road network data around the construction area to obtain time dimension features and spatial dimension features. The time dimension features include adjacent time features, daily time features, and weekly time features, and the spatial dimension features include trip chain features and road features. Based on the graph convolution operation, the time dimension features and the spatial dimension features are combined to obtain the combined features, and the convolution operation is performed on the combined features to output the feature map; The feature map is input into the traffic flow prediction model for training.

7. The construction scene traffic diversion prediction system based on digital twin technology according to claim 1 is characterized in that: Develop traffic diversion strategies based on real-time traffic data and forecasted traffic, including: Obtain the preset diversion category table, input the predicted traffic into the trained random forest model, and obtain the traffic flow label; Query the preset diversion category table based on the traffic flow label to obtain the corresponding diversion method; Develop traffic diversion strategies based on corresponding diversion measures and real-time traffic data.

8. A construction scene traffic diversion prediction method based on digital twin technology, characterized by: include: Acquire historical traffic flow data, road network data around the construction area, and multiple real-time sensor data deployed in the construction area, wherein the historical traffic flow data includes historical traffic flow data and historical vehicle data, and the real-time sensor data includes real-time traffic flow data, real-time vehicle data, and real-time environmental data; The time dimension of historical traffic data is extracted to obtain a time dimension sequence. Based on the historical vehicle data, a vehicle trip chain is obtained. The vehicle trip chain is used to represent the location and time stamp set of sensors that the vehicle passes while driving. The vehicle trip chains are sorted according to the chronological order to obtain a trip chain sequence. Input the time dimension sequence, trip chain sequence and road network data around the construction area into the traffic flow prediction model for training, obtain the trained traffic flow prediction model, obtain the real-time trip chain sequence based on the real-time vehicle data, input the real-time traffic flow data and the real-time trip chain sequence into the trained traffic flow prediction model for prediction, and output the predicted traffic flow; Obtain real-time road data, update the pre-built digital twin model based on real-time road data, real-time environmental data and predicted traffic, and formulate traffic diversion strategies based on real-time traffic data and predicted traffic.

9. A construction scene traffic diversion prediction device based on digital twin technology, characterized in that: It includes a memory, a processor and a transceiver that are communicatively connected in sequence, wherein the memory is used to store computer programs, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the construction scene traffic diversion prediction method based on digital twin technology as described in claim 8.

10. A computer program product comprising a computer program or instructions, characterized in that When executed by a computer, the computer program or the instruction implements the construction scene traffic diversion prediction method based on digital twin technology as described in claim 8.

Images