Traffic simulation and digital twinborn fusion system and method based on Vissim

Abstract

The invention relates to the technical field of traffic, and discloses a Vissim-based traffic simulation and digital twin fusion system and method, and the system comprises a Vissim road network construction and programmed modeling module which is used for constructing a simulation road network based on high-precision map data and carrying out programmed modeling; the traffic flow data prediction module is used for predicting future short-term macroscopic traffic flow; the traffic flow simulation deduction module is used for performing traffic flow simulation deduction in combination with the prediction data; the coordinate conversion and alignment module is used for converting the simulation data coordinates into a geographic coordinate system and carrying out coordinate alignment; and the traffic flow twinning and correction module is used for realizing digital twinning and real-time correction of the traffic flow. According to the method, rapid construction and optimization of the Vissim road network based on the CAD picture format are realized, and real-time twinning and correction of traffic data are realized through traffic prediction, simulation deduction and coordinate conversion.

Description

Technical Field

[0001] The present invention relates to the technical field of transportation, and specifically relates to a traffic simulation and digital twin fusion system and method based on Vissim. Background Art

[0002] Common simulation software in the transportation field includes Vissim, SUMO, AIMSUN, etc. Among them, Vissim is widely used in traffic scenarios such as auxiliary decision-making and traffic flow prediction due to its rich built-in car-following models, high restoration degree, scientific and accurate evaluation models, and the ability to simulate complex traffic flow information. Digital twin can simulate, monitor, predict, and optimize various situations in the physical world based on captured and analyzed real-time data, and the key element of digital twin is real-world data.

[0003] Among them, the technical route for Vissim to implement traffic flow simulation and derivation and export data to the digital twin system includes: not using high-precision map data, only using planar CAD data as a reference for Vissim road network construction and twin modeling, and using the driving distance of vehicles on the road as the basis for coordinate conversion.

[0004] However, this technical method has the following problems:

[0005] Firstly, there may be a certain gap between CAD design drawings and actual construction organizations. Using CAD data as the sole standard for digital twin modeling will result in insufficient modeling accuracy, and subsequent collection of vehicle high-precision map data through radar, etc., will not be able to be unified with the modeling results;

[0006] Secondly, using the driving distance of vehicles in the Vissim road network as the basis for digital twin coordinate conversion has the problem of large cumulative errors. Summary of the Invention

[0007] Aiming at the deficiencies of the existing technology, the present invention provides a traffic simulation and digital twin fusion system and method based on Vissim to solve the above technical problems.

[0008] In the first aspect, a traffic simulation and digital twin fusion system based on Vissim is provided, including:

[0009] A Vissim road network construction and procedural modeling module for constructing a simulation road network based on high-precision map data and performing procedural modeling;

[0010] A traffic flow data prediction module for predicting future short-term macroscopic traffic flow;

[0011] A traffic flow simulation and derivation module for performing traffic flow simulation and derivation in combination with the prediction data;

[0012] A coordinate transformation and alignment module, which is used to transform the coordinates of simulation data into a geographic coordinate system and perform coordinate alignment;

[0013] A traffic flow digital twin and correction module, which is used to realize the digital twin and real-time correction of traffic flow.

[0014] Furthermore, the Vissim road network construction and procedural modeling module includes:

[0015] Project the shapefile data into the Mercator coordinate system and convert it into the CAD picture format, construct a road network in Vissim, extract traffic elements, and generate and optimize a road model through a procedural modeling tool.

[0016] Furthermore, the traffic flow data prediction module includes:

[0017] Adopt an improved GAN-LSTM-XGBoost short-term traffic flow prediction model to conduct macroscopic prediction based on historical traffic data and generate predicted traffic flow.

[0018] Furthermore, the traffic flow simulation and deduction module includes:

[0019] Combine traffic prediction data and the signal timing plan defined by the user to conduct traffic simulation in Vissim. After the traffic flow stabilizes, the future traffic flow changes can be deduced.

[0020] Furthermore, the coordinate transformation and alignment module includes:

[0021] Export the traffic flow data in Vissim through the python-com interface, including the two-dimensional plane coordinates of vehicles, convert the exported two-dimensional plane coordinates into a geographic coordinate system, and then convert them into a three-dimensional Cartesian coordinate system.

[0022] Furthermore, the traffic flow digital twin and correction module includes:

[0023] Connect the Vissim vehicle data to the digital twin engine, transmit the twin data through MQTT communication, and apply a correction algorithm for lateral correction and longitudinal correction.

[0024] Furthermore, the correction algorithm includes:

[0025] Lateral correction: Use the section information and lane information to correspond to the information in the Vissim road network, convert the longitude and latitude coordinates of the vehicle into world coordinates, and find the coordinates of the point closest to the road line for correction;

[0026] Longitudinal deviation correction: After lateral deviation correction, according to the vehicle size and road conditions, calculate the intersection coordinates of this straight line and the road surface model, and use this coordinate as the result of vehicle deviation correction, where this straight line refers to the straight line passing through the vehicle's centroid point and perpendicular to the road.

[0027] In a second aspect, a method for fusing traffic simulation and digital twin based on Vissim is provided. Based on the traffic simulation and digital twin fusion system based on Vissim described in any one of the foregoing, it includes the following steps:

[0028] Use the Vissim road network construction and procedural modeling module to construct a road network according to shp data and generate a road model;

[0029] Use the traffic flow data prediction module to conduct macroscopic traffic flow prediction on the historical real traffic flow of the research road object;

[0030] Use the traffic flow simulation and deduction module to conduct traffic flow simulation and deduction according to the predicted traffic flow and traffic light timing rules;

[0031] Use the coordinate conversion and alignment module to convert Vissim simulation data into longitude and latitude data, and then convert it into a three-dimensional Cartesian coordinate system to achieve data alignment;

[0032] Use the traffic flow twin and deviation correction module to connect Vissim vehicle data to the twin engine in real time, and after deviation correction processing, realize visual parallel twins.

[0033] Furthermore, process the shp file and generate a base map, extract traffic elements, conduct procedural modeling and generate a digital twin scene.

[0034] Furthermore, use an improved GAN-LSTM-XGBoost short-term traffic flow prediction model for prediction.

[0035] The invention adopting the above technical solutions has the following advantages:

[0036] 1. The present invention realizes the rapid construction and optimization of the Vissim road network based on CAD picture format, and through traffic prediction, simulation deduction and coordinate conversion, realizes the real-time twin and deviation correction of traffic data.

[0037] 2. The present invention provides a technical route reference for the simulation deduction of macroscopic traffic flow, the traffic operation simulation and situation deduction of the reconstructed and expanded sections, etc.

[0038] 3. The present invention can accurately and intuitively conduct traffic flow prediction, provides an important reference for traffic planning and decision-making, effectively improves the road safety level, and provides an important basis for the economic benefit calculation of road traffic flow. Description of the Drawings

[0039] To more clearly illustrate the specific embodiments of the present invention, the accompanying drawings required for the specific embodiments will be briefly introduced below. In all the drawings, the elements or parts are not necessarily drawn to scale.

[0040] Figure 1 It is a short-term traffic flow prediction model diagram in the traffic simulation and digital twin fusion system based on Vissim of the present invention;

[0041] Figure 2 It is a SectionIndex label definition diagram in the traffic simulation and digital twin fusion system based on Vissim of the present invention;

[0042] Figure 3 It is a coordinate diagram in the traffic simulation and digital twin fusion system based on Vissim of the present invention;

[0043] Figure 4 It is a result diagram of vehicle deviation correction in the traffic simulation and digital twin fusion system based on Vissim of the present invention;

[0044] Figure 5 It is the application demonstration effect in the traffic simulation and digital twin fusion system based on Vissim of the present invention Figure 1 ;

[0045] Figure 6 It is the application demonstration effect in the traffic simulation and digital twin fusion system based on Vissim of the present invention Figure 2 ;

[0046] Figure 7 It is a flowchart of the traffic simulation and digital twin fusion method based on Vissim of the present invention. Specific Embodiments

[0047] The embodiments of the technical solution of the present invention will be described in detail below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, so they are only examples and cannot be used to limit the protection scope of the present invention.

[0048] As Figures 1 to 7 shown, the traffic simulation and digital twin fusion system based on Vissim of the present invention includes:

[0049] A Vissim road network construction and procedural modeling module, used to construct a simulation road network based on high-precision map data and perform procedural modeling;

[0050] A traffic flow data prediction module, used to predict future short-term macroscopic traffic flow;

[0051] A traffic flow simulation deduction module, used to perform traffic flow simulation deduction in combination with prediction data;

[0052] A coordinate transformation and alignment module for transforming the coordinates of simulation data into a geographic coordinate system and performing coordinate alignment;

[0053] A traffic flow digital twin and correction module for realizing the digital twin and real-time correction of traffic flow.

[0054] In some embodiments, the Vissim road network construction and procedural modeling module includes:

[0055] Project the shapefile (hereinafter referred to as "shp" data) data onto the Mercator coordinate system and convert it into the CAD picture format, construct a road network in Vissim, extract traffic elements, and generate and optimize a road model through a procedural modeling tool.

[0056] Specifically, in order to make the vehicle data of the Vissim simulation have the attributes of a high-precision map while ensuring the stability and replicability of the technical route, the present invention uses the shp file of the road as the basis for Vissim road network construction and procedural modeling.

[0057] Vissim road network construction:

[0058] ① Shp file processing and base map generation

[0059] The shp data coordinates are in the 84 geographic coordinate system, using longitude and latitude to represent coordinate information, while the Vissim simulation software constructs a road network based on a planar coordinate system.

[0060] First, project the shp high-precision map information onto the corresponding Mercator projection coordinate system, and use CAD software to convert it into the.dwg format. The coordinate information of the converted.dwg format corresponds to the coordinate information of the projection coordinate system in the shp file.

[0061] ② Simulation road network construction

[0062] On this basis, import the.dwg format into Vissim as a reference for the base map of road network construction. Refer to the reference information such as road edges and centerlines in the reference base map to draw the Vissim simulation road network.

[0063] Traffic element extraction:

[0064] According to materials such as shp files, equipment lists, and design documents, use the Python language to import and parse them, obtain information such as the positions and types of road markings, mechanical and electrical facilities, signs and placards, and store them as a traffic element table for constructing a twin scene.

[0065] Procedural modeling:

[0066] The shp data of the road can be used for procedural modeling. Import the shp data into the procedural modeling software and apply rules for roads, buildings, intersections, etc., and the results of procedural modeling can be obtained.

[0067] Furthermore, import the results of procedural modeling into professional modeling software, and fine carving of models and terrain editing as well as deployment and application of traffic elements can be carried out to obtain a digital twin scenario.

[0068] In some embodiments, the traffic flow data prediction module includes:

[0069] An improved GAN-LSTM-XGBoost short-term traffic flow prediction model is adopted to conduct macroscopic prediction based on historical traffic data and generate predicted traffic flow.

[0070] Specifically, to realize the simulation and deduction of traffic flow, the present invention proposes an improved GAN-LSTM-XGBoost short-term traffic flow prediction model. The structure diagram of the model is as Figure 1 shown, and LSTM uses 10 consecutive data as neurons in the hidden layer.

[0071] Among them, GAN simulates the real traffic flow time series through the generator G to generate adversarial samples, enhancing the robustness and diversity of the data; the improved LSTM module is used to extract key dynamic features in the time series; the XGBoost module conducts final traffic flow prediction for the deep time series extracted by LSTM and external auxiliary features.

[0072] The inputs include:

[0073] ① Traffic flow characteristics, including the historical traffic flow X = [x 1 , x 2 ,... x T at each time step, where x T represents the traffic flow at time step t;

[0074] ② Auxiliary features, including weather factors, holidays, accident information, etc. A = [a 1 , a 2 ,..., a T ;

[0075] ③ For the GAN module, a random noise vector z ∼ N(0,1) should also be input.

[0076] GAN module

[0077] The generator generates enhanced traffic flow data with the random noise z and historical traffic flow data X

[0078]

[0079] The discriminator D takes the real historical traffic flow data X and the generated data as inputs and outputs the true and false probabilities p:

[0080]

[0081] The GAN loss function, which is the optimization objective, is the adversarial training of the generator G and the discriminator D:

[0082]

[0083] denotes the expected loss for the discriminator to correctly identify a real sample as real.

[0084] denotes the expected loss for the discriminator to correctly identify a generated sample as fake.

[0085] Improved LSTM module

[0086] LSTM module

[0087] LSTM is an improved recurrent neural network. Since there is only one state in the hidden layer of the original RNN, it is very sensitive to short-term inputs and there are also problems of gradient descent and gradient vanishing. LSTM effectively avoids the problems of gradient vanishing and long-term dependence in the RNN model by introducing three different function gate structures. The core of LSTM is the gating mechanism, which is used to control the memory and forgetting of information, and the generated data input to the GAN module The output calculation of the forget gate is as follows:

[0088] f t = σ(W f · [h t-1 , x t + b f )

[0089] where f t is the output of the forget gate, W f is the weight matrix, b f is the bias term, and h t-1 represents the hidden state at the previous moment.

[0090] The input gate calculation is as follows:

[0091] i t = σ(W i · [h t-1 , x t + b i )

[0092]

[0093] it is the output of the input gate, is the candidate memory cell state.

[0094] Update the memory cell state at time t according to the following formula:

[0095]

[0096] The output gate is calculated as follows:

[0097] o t = σ(W o · [h t-1 , x t + b o )

[0098] h t = o t · tanh(C t )

[0099] Introduce the attention mechanism

[0100] The attention mechanism can help the model automatically select the key parts in the sequence and improve the model's focusing ability on the important parts of the sequence.

[0101] The core idea of the attention mechanism is to assign a weight to each input, and the higher the weight, the more important the input at the current moment.

[0102] Calculate the correlation relaRef between the current hidden state and all historical hidden states:

[0103]

[0104] W k is the weight matrix for linear transformation of the hidden state.

[0105] Convert the correlation to a weight through the softmax function so that the sum of the weights is 1:

[0106]

[0107] The context vector output is calculated as follows:

[0108]

[0109] XGBoost module

[0110] There are too many parameters in the fully connected layer of LSTM, which is prone to overfitting. To solve this problem, the XGBoost model is introduced.

[0111] The input is the feature O extracted by LSTM, and the output is the predicted future traffic The loss function of the XGBoost module is as follows:

[0112]

[0113] In some embodiments, the traffic flow simulation and deduction module includes: combining traffic prediction data and a signal timing plan customized by the user to perform traffic simulation in Vissim. After the traffic flow stabilizes, the future traffic flow changes can be deduced.

[0114] Specifically, the output result of Module 2 is the initial state of the future macroscopic traffic flow at discrete time points. To achieve continuous deduction of the short-term traffic flow, this predicted value needs to be used as the input, and the red, green, and yellow signal timing plans of the signal machine need to be set in Vissim.

[0115] After the above settings are completed, traffic simulation is performed. After the traffic flow fills the entire road network and stabilizes, the traffic flow deduction results of different signal timing plans are obtained.

[0116] In some embodiments, the coordinate transformation and alignment module includes: exporting traffic flow data in Vissim through the python-com interface, including the two-dimensional plane coordinates of vehicles, converting the exported two-dimensional plane coordinates into a geographic coordinate system, and then converting them into a three-dimensional Cartesian coordinate system.

[0117] Specifically, traffic flow data export:

[0118] Vissim provides a python-com interface. By writing a python script program, traffic flow data can be exported. The exported fields mainly include vehicle type, color, speed, abscissa, ordinate, road number, and lane number, etc.

[0119] Coordinate transformation and alignment:

[0120] In the exported traffic flow data, the vehicle coordinates are two-dimensional plane coordinates that correspond one-to-one with the Mercator projection coordinate system of the shp data.

[0121] To achieve coordinate alignment of vehicles in the twin engine, this plane coordinate needs to be converted into a geographic coordinate system. The conversion method is as follows:

[0122] Calculate the latitude coordinate of the reference point map:

[0123]

[0124] EarthRadiu represents the earth radius of the Mercator projection, EarthRadius = 6378137 meters,

[0125] y RefMapRepresents the ordinate of the background map reference point, which can be obtained by querying in the Vissim road network settings.

[0126] Calculate the local scale factor:

[0127] LocalScale is the local scale factor for converting the vissim coordinate system to the Mercator coordinate system, and is calculated as follows:

[0128]

[0129] The Mercator coordinates are calculated as follows:

[0130] MercatorX = (x Vissim - x Re fNet ) · localScale + x Re fMap

[0131] MercatorY = (y Vissim - y Re fNet ) · localScale + y Re fMap

[0132] Where, x Re fNet , y Re fNet , x Re fMap , y Re fMap respectively represent the horizontal and vertical coordinates of the reference point in the road network and the horizontal and vertical coordinates of the background map reference point, and x Vissim , y Vissim are respectively the horizontal and vertical coordinates of the vehicle in the Vissim simulation.

[0133] Calculate the longitude Longitude and latitude Latitude coordinates as follows:

[0134]

[0135] In some embodiments, the traffic flow twin and correction module includes: connecting Vissim vehicle data to the digital twin engine, transmitting twin data through MQTT communication, and applying correction algorithms for lateral correction and longitudinal correction.

[0136] Specifically, the twin interface design:

[0137] The twin data is transmitted through MQTT communication. The Json fields and descriptions transmitted by the twin interface are as follows:

[0138] Table 1

[0139]

[0140] In some embodiments, the deviation correction algorithm includes:

[0141] Lateral deviation correction: Corresponding the road section information and lane information with the information in the Vissim road network, converting the longitude and latitude coordinates of the vehicle into world coordinates, and finding the coordinates of the point closest to the road line for deviation correction.

[0142] Longitudinal deviation correction: After lateral deviation correction, according to the vehicle size and road conditions, obtain the intersection coordinates of the straight line and the road surface model, and use these coordinates as the result of vehicle deviation correction, where the straight line refers to the straight line passing through the vehicle centroid point and perpendicular to the road.

[0143] Specifically, the application and real-time twin of the deviation correction algorithm:

[0144] Due to the adjustment of procedural modeling and model carving, there may be slight differences between the modeling results and the Vissim road network data. To prevent the vehicle from deviating from the lane line during the twin process, it is necessary to design and apply the deviation correction algorithm.

[0145] In the above interface design, the road section and lane line attributes are exported and can be used for vehicle position deviation correction applications. The deviation correction algorithm includes two aspects: lateral deviation correction and longitudinal deviation correction.

[0146] Lateral deviation correction:

[0147] According to the construction of the Vissim road network, preset the road line structure in the digital twin engine. The road line structure includes road section information and lane information, and use the above two pieces of information to correspond one by one with the information in the Vissim road network.

[0148] First, preset the road line structure in the digital twin scenario and add the SectionIndex and LaneIndex tags. These two tags correspond one by one with the Vissim road network. The definition of the SectionIndex tag is as Figure 2 shown. For each road section, the LaneIndex from the slow lane to the fast lane is defined as 1, 2,..., n in sequence, where n is the total number of lanes in this road section.

[0149] Second, the longitude and latitude information of the vehicle cannot be located and visualized in the twin engine, so it is necessary to perform coordinate conversion from the geographic coordinate system to the world coordinate system. Module four has obtained the longitude and latitude coordinates of the vehicle, and the world coordinates (x 1 , y 1 , z 1 ) of the vehicle in the twin engine will be obtained according to the following method.

[0150] Convert the longitude and latitude to radians:

[0151]

[0152] Calculate (x 1 , y 1 , z 1 ) as follows:

[0153]

[0154] On this basis, find the corresponding road line structure in the engine according to the road section and lane line number, and obtain the coordinates (x 1 , y 1 , z 1 ) of the point closest to the road line, as shown in 2 , y 2 , z 2 ). Figure 3 as shown

[0155] Longitudinal deviation correction:

[0156] To ensure that the vehicle always drives smoothly on the road surface, the vehicle should be corrected in the longitudinal position. First, collision bodies should be preset for the vehicle model and the road surface model respectively.

[0157] Secondly, after lateral deviation correction, taking (x 2 , y 2 , z 2 ) as the origin, draw a straight line in the Z-axis direction with a certain deviation range (take ±2m here), and obtain the intersection coordinates (x 3 , y 3 , z 3 ) of the straight line and the road surface model, and take this coordinate as the result of vehicle deviation correction, as shown in Figure 4 as shown

[0158] In some other embodiments, a method for fusing traffic simulation and digital twin based on Vissim is provided, and a traffic simulation and digital twin fusion system based on Vissim according to any one of the foregoing includes the following steps:

[0159] Step S01, use the Vissim road network construction and procedural modeling module to construct a road network according to the shp data and generate a road model;

[0160] Step S02, use the traffic flow data prediction module to perform macroscopic traffic flow prediction on the historical real traffic flow of the research road object;

[0161] Step S03, use the traffic flow simulation and deduction module to perform traffic flow simulation and deduction according to the predicted traffic flow and traffic light timing rules;

[0162] Step S04: Use the coordinate transformation and alignment module to convert the Vissim simulation data into latitude and longitude data, and then convert it into a three-dimensional Cartesian coordinate system to achieve data alignment;

[0163] Step S05: Use the traffic flow twin and correction module to connect the Vissim vehicle data to the twin engine in real time. After correction processing, realize visual parallel twinning.

[0164] In some embodiments, process the shp file and generate a base map, extract traffic elements, perform procedural modeling, and generate a digital twin scene.

[0165] In some embodiments, an improved GAN-LSTM-XGBoost short-term traffic flow prediction model is used for prediction.

[0166] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered by the scope of the claims and the description of the present invention.

Claims

1. Traffic simulation and digital twin fusion system based on Vissim, characterized by: include: Vissim road network construction and programmatic modeling module, used to build simulated road networks based on high-precision map data and perform programmatic modeling; Traffic flow data prediction module, used to predict future short-term macro traffic flow; Traffic flow simulation module, used to simulate traffic flow in combination with forecast data; The coordinate conversion and alignment module is used to convert the simulation data coordinates into the geographic coordinate system and perform coordinate alignment; The traffic flow twin and correction module is used to realize the digital twin and real-time correction of traffic flow.

2. The system according to claim 1, characterized in that The Vissim road network construction and programmatic modeling module includes: projecting shapefile data into the Mercator coordinate system and converting it into a CAD image format, building a road network in Vissim, extracting traffic elements, and generating and optimizing a road model through a programmatic modeling tool.

3. The system according to claim 2, characterized in that The traffic flow data prediction module includes: using an improved GAN-LSTM-XGBoost short-term traffic flow prediction model, performing macro prediction based on historical traffic data, and generating predicted traffic flow.

4. The system according to claim 3, characterized in that The traffic flow simulation module includes: combining traffic forecast data and user-defined signal light timing schemes to perform traffic simulation in Vissim, and after the traffic flow stabilizes, the future traffic flow changes can be deduced.

5. The system according to claim 4, characterized in that The coordinate conversion and alignment module includes: exporting traffic flow data in Vissim through a python-com interface, including two-dimensional plane coordinates of vehicles, converting the exported two-dimensional plane coordinates into a geographic coordinate system, and then converting them into a three-dimensional Cartesian coordinate system.

6. The system according to claim 5, characterized in that The traffic flow twin and correction module includes: connecting Vissim vehicle data to the digital twin engine, transmitting twin data through MQTT communication, and applying correction algorithms to perform lateral and longitudinal corrections.

7. The system according to claim 6, characterized in that The correction algorithm includes: Lateral deviation correction: Use the road section information and lane information to match the information in the Vissim road network, convert the latitude and longitude coordinates of the vehicle into world coordinates, and find the coordinates of the point closest to the road line for deviation correction; Longitudinal correction: After the lateral correction, the coordinates of the intersection of the straight line and the road model are calculated based on the vehicle size and road conditions, and the coordinates are used as the result of the vehicle correction, where the straight line refers to the straight line passing through the vehicle's center of mass and perpendicular to the road.

8. The traffic simulation and digital twin fusion method based on Vissim is characterized by: The traffic simulation and digital twin fusion system based on Vissim according to any one of claims 1 to 7 comprises the following steps: Use Vissim road network construction and programmatic modeling modules to build road networks and generate road models based on shp data; Using the traffic flow data prediction module, macro traffic flow prediction is performed on the historical real traffic flow of the research road object; Use the traffic flow simulation module to simulate traffic flow according to the predicted traffic flow and traffic light timing rules; Use the coordinate conversion and alignment module to convert Vissim simulation data into longitude and latitude data, and then convert it into a three-dimensional Cartesian coordinate system to achieve data alignment; By using the traffic flow twin and correction module, the Vissim vehicle data is connected to the twin engine in real time, and after correction processing, visual parallel twinning is achieved.

9. The method according to claim 8, characterized in that Process shp files and generate base maps, extract traffic elements, perform procedural modeling and generate digital twin scenes.

10. The method according to claim 8, characterized in that The improved GAN-LSTM-XGBoost short-term traffic flow prediction model is used for prediction.

Images