Traffic simulation analysis method for highway construction area based on VISSIM

Abstract

The invention provides a traffic simulation analysis method for an expressway construction area based on VISSIM. The method comprises the following steps: S1, collecting point cloud data, vehicle data and traffic video data of the landform of the construction area in real time; s2, preprocessing the collected multi-source data; s3, constructing a dynamic simulation model of the construction area; s4, simulating the traffic condition of the construction area; s5, dynamically planning an optimal driving path of the vehicle according to the simulated traffic condition and construction information; and analyzing the vehicle overlapping probability of the intersection in real time, and performing early warning according to an analysis result. Through the technical means of multi-source data fusion, high-precision dynamic modeling, real-time analogue simulation, intelligent path planning and the like, accurate simulation and optimal management of the traffic influence during the blasting demolition construction of the expressway overline overbridge group are realized, and remarkable technical advantages and application values are achieved; and a scientific and efficient solution is provided for urban traffic management.

Description

Technical Field

[0001] The present invention relates to the field of intelligent transportation technology, and in particular to a traffic simulation analysis method for a highway construction area based on VISSIM. Background Art

[0002] In today's rapidly urbanizing world, highways, as a key component of urban transportation networks, face increasing demand for their construction and maintenance. With the continuous expansion of urban areas and the continued rise in transportation demands, highway renewal and expansion projects have become increasingly important. However, these projects, particularly those involving the demolition and blasting of overpasses, often present a series of complex and challenging challenges.

[0003] Conventional traffic management methods struggle to cope with the complex traffic impacts associated with the demolition and blasting of highway overpasses. Due to the high volume and speed of traffic on highways, construction activities can cause traffic congestion, delays, and accidents. Existing technologies also suffer from significant deficiencies in data collection, model accuracy, and safety warnings. Furthermore, existing safety warning mechanisms are often incomplete, making it difficult to promptly detect and address potential traffic conflicts, increasing safety risks during construction.

[0004] Specifically, the demolition and blasting of overpasses requires not only precise blasting operations but also effective traffic control and diversion measures. However, existing technologies have significant shortcomings in this regard. For example, limited data collection methods fail to fully and accurately reflect real-time traffic conditions in the construction area; simulation models lack high accuracy, making it difficult to accurately predict the scope and extent of the construction's impact on surrounding traffic; and safety warning systems are slow to respond, failing to issue alerts and implement effective measures before traffic conflicts occur.

[0005] These issues not only affect construction progress and efficiency, but also cause significant inconvenience to citizens and may even endanger the safety of pedestrians and vehicles. Therefore, there is an urgent need for a dynamic simulation and visualization analysis system that can comprehensively, accurately, and in real time simulate traffic flow changes during construction and provide effective optimization management strategies. Summary of the Invention

[0006] The present invention proposes a traffic simulation analysis method for highway construction areas based on VISSIM, which solves the problem that the traffic organization method in the prior art cannot fully and accurately reflect the real-time traffic conditions in the construction area.

[0007] The technical solution of the present invention is achieved as follows: A first aspect of the present invention provides a traffic simulation analysis method for a highway construction area based on VISSIM, comprising the following steps: S1, multi-source data acquisition: Use LiDAR to collect real-time point cloud data of the construction area's topography, use ETC gantries to collect real-time vehicle data in the construction area, and use video surveillance systems to collect real-time traffic video data in the construction area; S2, preprocessing the collected multi-source data; S3, dynamic simulation model construction: Build a BIM model of the construction area, align the pre-processed data with the BIM model, and construct a high-precision dynamic road network model. Use VISSIM software to dynamically integrate construction machinery at different construction stages with the road network model to obtain a dynamic simulation model. S4, simulation: Based on the dynamic simulation model, traffic conditions in the construction area are simulated on VISSIM software; S5, path optimization and safety warning Based on simulated traffic conditions and construction information, the optimal vehicle driving path is dynamically planned; the matrix analysis method is used to analyze the probability of vehicle overlap at intersections in real time, and early warnings are issued based on the analysis results.

[0008] Specifically, in step S2, the method for preprocessing the collected multi-source data includes: Data cleaning: Use the Z-score method to detect and remove outliers and noise points in the test data; Filtering: Use Savitzky-Golay filter to smooth the data curve; Interpolation processing: Cubic spline interpolation method is used to fill data gaps.

[0009] Specifically, in step S3, the method of registering the preprocessed data with the BIM model includes: Initial alignment: Use CloudCompare to manually select corresponding feature points to achieve rough alignment between the construction area point cloud data and the BIM model; Fine registration: Using the iterative closest point algorithm, the rotation and translation between the construction area point cloud data and the BIM model are iteratively optimized; Registration verification: By calculating the residual distribution of the overlapping area, ensure that the overlap reaches the preset value.

[0010] Specifically, in step S3, the method for constructing a high-precision dynamic road network model includes: Import the registered point cloud and BIM model into the ArcGIS platform to extract the road centerline, cross section and intersection geometric elements; According to the construction schedule, the road network topology at different construction stages is timestamped and exported as a VISSIM INP format file. Write Python scripts to dynamically load the road network for the corresponding stage through the VISSIM COM interface, and combine it with the location data of construction machinery to achieve real-time switching of the road network model in the simulation environment; By comparing the actual collected traffic flow data with the simulation results in VISSIM, the time synchronization error is controlled to be less than the preset value.

[0011] Furthermore, step S3 further includes: performing multi-scale modeling of traffic flow in the construction area based on the improved ST-Transformer model, specifically including the following steps: The dynamic correlation of time series and the geometric constraints of spatial topology are captured synchronously through the multi-head self-attention mechanism. The expression is:

[0012] in, Q is the query vector, which is used to represent the current location or feature of interest; K is a key vector, used to indicate the position of the input sequence; V is a value vector containing the actual information of the input sequence; d k is the dimension of the key vector; T Represents the transpose of a matrix; is a spatiotemporal mask matrix used to suppress non-causal dependencies and invalid spatial associations; Introduce dynamic time warping loss function:

[0013] in, L DTW is the total value of the dynamic time warping loss function, which is used to measure the comprehensive error of two time series under the optimal alignment path. At the same time, spatial covariance constraints are introduced to prevent overfitting. p is the optimal alignment path, For time series points x i and y i Local distance measure between; l is the regularization coefficient; Σ is the spatial covariance matrix, which is calculated from the road network geometric parameters and dynamic construction data of the construction area; is the matrix trace operation; Multi-head attention through parallel computing h The attention head improves the ability to capture spatiotemporal features. The calculation formula is:

[0014] Each attention head is defined as:

[0015] in, , , is the learnable parameter matrix, is the output projection matrix.

[0016] Specifically, in step S5, the path optimization method is: Based on the Q-learning reinforcement learning algorithm, the optimal driving path of the vehicle is dynamically planned according to the simulated traffic conditions and construction information. The Q-learning update formula is:

[0017] in, Q ( s , a ) is in state s Next select action a The value function of α is the learning rate, r To perform an action a After receiving the instant reward, c is the discount factor, Indicates the next state Select an action The maximum expected value.

[0018] Specifically, in step S5, the safety warning method is: Building the Vehicle x The time-space trajectory matrix ; R means that the matrix belongs to the real field, T is the total number of simulation time steps, S is the number of spatial nodes; Calculate the number of any two cars k and m The probability of trajectory overlap:

[0019] If the vehicle k At the time node t Located at the spatial node s ,but ,otherwise ; If the vehicle m At the time node t Located at the spatial node s ,but ,otherwise ; Set the triggering conditions for the warning: like , a yellow warning is triggered; like , then an orange warning is triggered; like , a red alert is triggered.

[0020] Specifically, the simulation analysis method further includes: Step S6: 3D situation visualization display: A WebGL-based 3D rendering engine is used to build a traffic flow visualization scene in the construction area. This integrates real-time multi-source data, dynamic simulation models, and safety warning information with the visualization scene to display traffic flow changes and safety warning information in the construction area in real time.

[0021] A second aspect of the present invention provides an electronic device comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and the processor implements the steps of the traffic simulation analysis method when executing the computer program.

[0022] A third aspect of the present invention provides a computer-readable storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the traffic simulation analysis method are implemented.

[0023] Compared with the prior art, the present invention has the following beneficial effects: (1) The present invention improves the comprehensiveness of data collection by integrating multi-source data, achieves a comprehensive and accurate reflection of the traffic conditions in the construction area, and provides strong support for the subsequent formulation of scientific and reasonable traffic organization plans; (2) This invention constructs a high-precision dynamic road network model and combines it with an improved ST-Transformer model to perform multi-scale modeling of traffic flow in the construction area, thereby improving the accuracy and reliability of the simulation results, helping to predict the impact of construction on surrounding traffic, and providing a scientific basis for optimizing traffic management strategies; (3) By integrating vehicle conflict detection and real-time warning functions, the present invention enables the system to promptly detect and handle potential traffic conflicts, thereby greatly reducing safety risks during construction and ensuring the safety of pedestrians and vehicles; (4) Based on real-time monitoring data and simulation results, the present invention dynamically generates the optimal traffic management strategy, which helps to alleviate the traffic pressure caused by construction and improves construction efficiency and the overall operation efficiency of urban traffic. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0025] Figure 1 The figure is a flow chart of the traffic simulation analysis method for highway construction areas based on VISSIM of the present invention. DETAILED DESCRIPTION

[0026] The following will clearly and completely describe the technical solutions of the present invention in conjunction with the embodiments of the present invention. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0027] Reference Figure 1 The first aspect of the present invention provides a traffic simulation analysis method for a highway construction area based on VISSIM, comprising the following steps: S1, multi-source data acquisition: Use LiDAR to collect topographic information of the construction area in real time and generate high-precision point cloud data; The ETC gantry collects vehicle data in the construction area in real time, including vehicle flow, speed, and vehicle type, with a data collection frequency of once per minute; The video surveillance system is used to collect real-time traffic video data in the construction area, with a video frame rate of 25 frames per second.

[0028] S2, preprocessing the collected multi-source data: Data cleaning: Remove outliers and noise points from the test data, and use the Z-score method to detect and eliminate data points that exceed 3 times the standard deviation; Filtering: The Savitzky-Golay filter is used to smooth the data curve, with a filter window size of 5 and a polynomial order of 2; Interpolation processing: Cubic spline interpolation method is used to fill data gaps to ensure data continuity and integrity.

[0029] S3, dynamic simulation model construction: Build a BIM model of the construction area, align the pre-processed data with the BIM model, and construct a high-precision dynamic road network model. Use VISSIM software to dynamically integrate construction machinery at different construction stages with the road network model to obtain a dynamic simulation model. The specific steps include: S31, BIM model export and preprocessing: Use Revit software to export the construction area structure model into IFC or OBJ format; The model was simplified in Blender, a 3D modeling software, removing irrelevant components and retaining only the road surface, bridges, and key construction structures; Unify the model coordinate system to the global coordinate system of the project.

[0030] S32, LiDAR point cloud acquisition and preprocessing: Deploy bottom-surface LiDAR to scan the construction area along the designed path and collect high-density point clouds; Denoising and voxel filtering downsampling were performed in CloudCompare, with control point spacing of 5-10 cm; Based on the semantic segmentation algorithm Eigen, the point cloud is classified and the road surface, structure and environment point clouds are extracted.

[0031] S33, BIM model and point cloud data registration: Initial alignment: Use CloudCompare to manually select corresponding feature points to achieve rough alignment between the construction area point cloud data and the BIM model; Fine registration: Using the iterative closest point algorithm, the rotation and translation between the construction area point cloud data and the BIM model are iteratively optimized to ensure an average registration error of ≤0.15m; Registration verification: The overlap degree was ensured to be >95% by calculating the residual distribution in the overlapping area.

[0032] S34, build a high-precision dynamic road network model: Import the registered point cloud and BIM model into the ArcGIS platform to extract the road centerline, cross section and intersection geometric elements; According to the construction schedule, the road network topology at different construction stages is timestamped and exported as a VISSIM INP format file. Write Python scripts to dynamically load the road network for the corresponding stage through the VISSIM COM interface, and combine it with the location data of construction machinery to achieve real-time switching of the road network model in the simulation environment; In VISSIM, the actual collected traffic flow data is compared with the simulation results, and the time synchronization error is controlled to be <0.8s.

[0033] S35, ST-Transformer model integration and multi-scale traffic flow modeling: To achieve accurate simulation of traffic flow, this paper integrates an improved ST-Transformer (Spatiotemporal Transformer) model to capture the temporal dependency and spatial correlation of traffic data through a spatiotemporal attention mechanism. The specific steps are as follows: (1) Input data type: Time series data: vehicle flow and speed collected by ETC gantries (time resolution 1 minute); Spatial feature data: road section geometric parameters (such as number of lanes and curvature) extracted from lidar point clouds; Dynamic construction data: construction machinery location, blasting vibration intensity (updated in real time); (2) Data standardization: The Z-core method is used to normalize the input data. The normalization formula is:

[0034] in, m is the mean, s is the standard deviation, x For the original collected data, z The data are normalized.

[0035] (3) ST-Transformer core module design

[0036] The dynamic correlation of time series and the geometric constraints of spatial topology are captured synchronously through the multi-head self-attention mechanism. The expression is:

[0037] in, Q is the query vector, which is used to represent the current location or feature of interest; K is a key vector, used to indicate the position of the input sequence; V is a value vector containing the actual information of the input sequence; d k is the dimension of the key vector; T Represents the transpose of a matrix; is a spatiotemporal mask matrix used to suppress non-causal dependencies and invalid spatial associations; Introduce dynamic time warping loss function:

[0038] in, L DTW is the total value of the dynamic time warping (DTW) loss function, which is used to measure the comprehensive error between two time series (such as real-time traffic flow and simulated traffic flow in a construction area) under the optimal alignment path. At the same time, a spatial covariance constraint is introduced to prevent overfitting. p is the optimal alignment path, For time series points x i (such as actual traffic flow) and y i(e.g., traffic flow simulation prediction) local distance measurement; l is the regularization coefficient; Σ is the spatial covariance matrix, which is calculated from the road network geometry parameters (such as the number of lanes and curvature) and dynamic construction data (such as the position of machinery) in the construction area; It is a matrix trace operation, that is, the sum of the diagonal elements of the covariance matrix; the model achieved a breakthrough performance of time alignment error <0.8s and spatial registration accuracy of 0.15m in the construction area.

[0039] Multi-head attention through parallel computing h The attention head improves the ability to capture spatiotemporal features. The calculation formula is:

[0040] Each attention head is defined as:

[0041] in, , , is the learnable parameter matrix, is the output projection matrix.

[0042] The spatiotemporal dependency between the ETC gantry and the BIM model is captured through the multi-head attention mechanism, and the spatiotemporal mask matrix M Invalid associations (such as the impact of future moments on the current moment) are suppressed to improve alignment accuracy. Compared with traditional DTW, the computational efficiency of the method of the present invention is improved by 30%.

[0043] S4, simulation: Based on the dynamic simulation model, the traffic conditions in the construction area are simulated on the VISSIM software.

[0044] The present invention designed multiple sets of simulation experiments to simulate traffic flow conditions under different construction schemes. The key parameters of the present invention were calibrated based on actual construction data and dynamically configured through the VISSIM platform. The specific model parameter settings are shown in Table 1 below: Table 1 Model parameter settings

[0045] Taking the number of lanes in the road network parameters in Table 1 as an example, the number of lanes is dynamically adjusted according to the actual road network width extracted by the lidar. The formula is:

[0046] Among them, RoadWidth is the actual width of the road (meters), and 3.5 meters is the standard lane width.

[0047] This embodiment also verifies the simulation results, including: (1) The consistency between the simulation results and the measured data is verified by the GEH statistic, and the formula is as follows:

[0048] in, V s To simulate traffic, V m The actual flow rate.

[0049] (2) Evaluate the model accuracy by RMSPE, the formula is as follows:

[0050] in, V s,i For the i The measured traffic flow or speed value at a time point; V m,i For the i The simulation prediction value at each time point, n is the total number of samples (e.g. sampling by minute, n =1440 is the daily data volume).

[0051] Evaluation criteria: RMSPE < 15% indicates excellent model accuracy, and 15%-25% indicates acceptable model accuracy.

[0052] This embodiment also tests the stability of the simulation model, including: (1) Repeatability test Operation: Fixed input parameters, repeated simulation 10 times, and recorded output traffic flow; Indicator: Calculate the coefficient of variation of 10 results CV , the formula is as follows:

[0053] in, s is the standard deviation, m is the mean; standard: CV <5% indicates that the model stability is qualified.

[0054] (2) Time period sensitivity analysis

[0055] Operation: Select the data of morning peak (7:00-9:00), flat peak (12:00-14:00), and evening peak (17:00-19:00) to run the model; Indicator: Compare the RMSPE values ​​of each time period, and the fluctuation range is required to be <10%.

[0056] S5, path optimization and safety warning

[0057] Dynamically plan optimal vehicle routes based on simulated traffic conditions and construction information; use matrix analysis to analyze the probability of vehicle overlap at intersections in real time, and issue warnings based on the analysis results. This includes the following steps: S51, path optimization based on Q-learning algorithm: Based on the Q-learning reinforcement learning algorithm, the optimal driving path of the vehicle is dynamically planned according to the simulated traffic conditions and construction information. The Q-learning update formula is:

[0058] in, Q ( s , a ) is in state s Next select action a The value function of α is the learning rate, which controls the update amplitude (the default value is 0.2); r To perform an action a Immediate rewards obtained after the trip (such as saving time and avoiding congestion); c is a discount factor that measures the importance of future rewards (the default value is 0.9); Indicates the next state Select an action The maximum expected value.

[0059] S52, vehicle conflict detection and safety warning: Building the Vehicle x The time-space trajectory matrix ; R means that the matrix belongs to the real field, T is the total number of simulation time steps (e.g. T =86400, corresponding to 24 hours, with a step size of 1s), S is the number of spatial nodes; Calculate the number of any two cars k and m The probability of trajectory overlap:

[0060] If the vehicle k At the time node t Located at the spatial node s ,but ,otherwise ; If the vehicle m At the time node t Located at the spatial node s ,but ,otherwise ; Set the triggering conditions for the warning: like , a yellow warning is triggered; like , then an orange warning is triggered; like , a red alert is triggered.

[0061] Specifically, the simulation analysis method further includes: Step S6: 3D situation visualization display: A WebGL-based 3D rendering engine was used to construct a traffic flow visualization scene for the construction area. This integrated real-time multi-source data, dynamic simulation models, and safety warning information with the visualization scene to display traffic flow changes and safety warning information in the construction area in real time. The specific steps involved were as follows: S61, using WebGL technology to achieve 3D situation display of the construction area: (1) Scene initialization and renderer configuration Technical framework: The scene is constructed based on the WebGL 3D rendering engine. The perspective camera parameters are set (viewing angle 75°, near clipping plane 0.1m, far clipping plane 1000m), the rendering resolution is 1920×1080, and the anti-aliasing function is enabled to improve the smoothness of the picture.

[0062] Background and lighting: The scene background color is set to light gray, and parallel light sources are added to simulate natural lighting to enhance the three-dimensional effect of the road network and vehicles.

[0063] (2) Road network model loading and optimization

[0064] Model import: Export road network 3D geometry data (in GLTF 2.0 format) from the BIM construction model, including the detailed structure of roads, bridges, and construction machinery.

[0065] Material and detail optimization: Adjust the material properties (metallicity 0.2, roughness 0.8) to avoid reflections interfering with the visualization.

[0066] LOD (Level of Detail) technology is used to dynamically simplify the geometric complexity of distant models and improve rendering efficiency.

[0067] (3) Dynamic traffic flow rendering

[0068] Vehicle representation: Using instanced mesh technology, vehicles are uniformly modeled as cuboids (size 2×1×4 meters), and efficient rendering is achieved by batch updating instance positions.

[0069] Real-time data driven: The vehicle coordinates (longitude, latitude, elevation) are obtained from the backend via WebSocket every 0.1 seconds, converted into spatial positions in the 3D scene, and the instantiated grid matrix is ​​updated.

[0070] (4) Shader program optimization

[0071] Heat map generation: Write a custom shader to map traffic density data into the HSL color space (low-density green H=120°, high-density red H=0°), and calculate pixel colors in real time through the fragment shader.

[0072] Performance tuning: Use GPU accelerated computing to reduce data transmission between the CPU and GPU, ensuring a stable frame rate above 60FPS.

[0073] (5) Data rasterization and density calculation

[0074] Spatial division: Divide the road network into 10m x 10m grids and count the real-time number of vehicles in each grid.

[0075] Density formula:

[0076] Data update: The latest density matrix is ​​pushed through the backend service every 2 seconds, and the frontend updates the heat map texture after receiving it.

[0077] (6) Heat map visualization

[0078] Color mapping rules: Density <0.1 vehicles / ㎡: green.

[0079] Density ≥ 0.5 vehicles / ㎡: red.

[0080] Intermediate values ​​are linearly interpolated to calculate hue.

[0081] Transparency control: The transparency of low-density areas is set to 30%, and the transparency of high-density areas is set to 80% to highlight congested areas.

[0082] S62, visual interface design: (1) Main interface partition 3D viewing area (occupies 70% of the screen): displays dynamic traffic flow, heat maps and 3D models of construction areas, and supports perspective rotation, zooming and panning.

[0083] Data panel area (occupies 30% of the screen): Displays real-time statistics, warning lists, and parameter control panels in separate tabs.

[0084] (2) Multi-dimensional data display

[0085] Layer control: Users can check to show / hide elements such as road networks, vehicles, heat maps, construction machinery, etc.

[0086] Timeline playback: Provides a slider to control the simulation timeline, supporting accelerated (2x speed) and decelerated (0.5x speed) playback of historical traffic conditions.

[0087] (3) Real-time warning information display

[0088] Warning signs: Yellow warning (track overlap probability 0.1-0.3): The conflict location is marked with a flashing yellow circle in the 3D view, and the warning item is displayed synchronously on the data panel.

[0089] Red alarm (track overlap probability ≥ 0.5): Triggers a full-screen red border flashing and plays an alarm tone to alert the operator.

[0090] Detailed interaction: Clicking on a warning entry will automatically locate the corresponding intersection and pop up a detailed conflict analysis (time, vehicle ID, probability value).

[0091] S63, technical effects and verification: (1) Rendering performance indicators The frame rate is stable at ≥60 FPS and supports rendering 1,000 vehicle models simultaneously.

[0092] The heat map update delay is less than 200 milliseconds, meeting real-time decision-making needs.

[0093] (2) User interaction verification

[0094] A / B testing is used to verify the rationality of the interface layout and ensure that operators can locate the target warning area within 5 seconds.

[0095] Based on user feedback, the color gradient of the heat map was optimized to avoid recognition difficulties for color-blind users.

[0096] The second aspect of the present invention provides a traffic simulation analysis system for highway construction areas based on VISSIM, which includes four core parts: data layer, simulation layer, model layer and application layer; Data layer: By integrating various monitoring equipment such as lidar, ETC gantries, and video surveillance, real-time collection of multi-source data such as the construction area's topography, vehicle numbers, and speed is achieved. The collected data undergoes pre-processing operations such as cleaning, filtering, and interpolation to ensure data quality and provide a reliable foundation for subsequent analysis.

[0097] Model layer: Utilizes BIM construction models and LiDAR data to construct a high-precision dynamic road network model that truly reflects the geographical environment and traffic conditions of the construction area. Combined with the improved ST-Transformer model, multi-scale modeling of traffic flow is performed at both macro and micro levels to improve the accuracy and stability of the simulation model.

[0098] Simulation layer: Through the Python-COM interface, real-time traffic flow simulation and dynamic spatiotemporal alignment of construction machinery and traffic flow are achieved to ensure the accuracy and real-time nature of the simulation results. Through dynamic simulation verification and adaptive tuning, model parameters are continuously optimized to ensure that the model can accurately predict the impact of construction on surrounding traffic.

[0099] Application layer: Utilizing WebGL technology, the system enables 3D situation display and dynamic heat map generation. Based on real-time warning information, it dynamically generates optimal traffic management strategies and provides intuitive and scientific decision-making support for emergency response through a visual interface. The system also integrates vehicle conflict detection and warning capabilities, calculating the overlap probability of vehicles entering an intersection in real time and triggering a three-level warning mechanism (yellow, orange, and red), effectively mitigating safety risks.

[0100] A third aspect of the present invention provides an electronic device comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and the processor implements the steps of the traffic simulation analysis method when executing the computer program.

[0101] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the traffic simulation analysis method are implemented.

[0102] Through the above-described specific implementation, the present invention implements a complete process from data collection and preprocessing to dynamic simulation model construction, simulation experiment design, route optimization and safety warning, result verification and performance evaluation, and 3D situation visualization. The organic integration and synergy of these steps ensures accurate simulation and optimized management of traffic impacts during the demolition and blasting of a cluster of highway overpasses, providing a scientific and efficient solution for urban traffic management.

[0103] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A traffic simulation analysis method for highway construction areas based on VISSIM, characterized in that: The following steps are involved: S1, multi-source data acquisition: Use LiDAR to collect real-time point cloud data of the construction area's topography, use ETC gantries to collect real-time vehicle data in the construction area, and use video surveillance systems to collect real-time traffic video data in the construction area; S2, preprocessing the collected multi-source data; S3, dynamic simulation model construction: Build a BIM model of the construction area, align the pre-processed data with the BIM model, and construct a high-precision dynamic road network model. Use VISSIM software to dynamically integrate construction machinery at different construction stages with the road network model to obtain a dynamic simulation model. S4, simulation: Based on the dynamic simulation model, traffic conditions in the construction area are simulated on VISSIM software; S5, path optimization and safety warning Based on simulated traffic conditions and construction information, the optimal vehicle driving path is dynamically planned; the matrix analysis method is used to analyze the probability of vehicle overlap at intersections in real time, and early warnings are issued based on the analysis results.

2. A traffic simulation analysis method for a highway construction area based on VISSIM as claimed in claim 1, characterized in that: In step S2, the method for preprocessing the collected multi-source data includes: Data cleaning: Use the Z-score method to detect and remove outliers and noise points in the test data; Filtering: Use Savitzky-Golay filter to smooth the data curve; Interpolation processing: Cubic spline interpolation method is used to fill data gaps.

3. The traffic simulation analysis method for a highway construction area based on VISSIM as claimed in claim 1, characterized in that: In step S3, the method for registering the pre-processed data with the BIM model includes: Initial alignment: Use CloudCompare to manually select corresponding feature points to achieve rough alignment between the construction area point cloud data and the BIM model; Fine registration: Using the iterative closest point algorithm, the rotation and translation between the construction area point cloud data and the BIM model are iteratively optimized; Registration verification: By calculating the residual distribution of the overlapping area, ensure that the overlap reaches the preset value.

4. The traffic simulation analysis method for a highway construction area based on VISSIM as claimed in claim 1, characterized in that: In step S3, the method for constructing a high-precision dynamic road network model includes: Import the registered point cloud and BIM model into the ArcGIS platform to extract the road centerline, cross section and intersection geometric elements; According to the construction schedule, the road network topology at different construction stages is timestamped and exported as a VISSIM INP format file; Write Python scripts to dynamically load the road network for the corresponding stage through the VISSIM COM interface, and combine it with the location data of construction machinery to achieve real-time switching of the road network model in the simulation environment; By comparing the actual collected traffic flow data with the simulation results in VISSIM, the time synchronization error is controlled to be less than the preset value.

5. The traffic simulation analysis method for highway construction area based on VISSIM as claimed in claim 1, characterized in that: Step S3 also includes: performing multi-scale modeling of traffic flow in the construction area based on the improved ST-Transformer model, specifically including the following steps: The dynamic correlation of time series and the geometric constraints of spatial topology are captured synchronously through the multi-head self-attention mechanism. The expression is: in, Q is the query vector, which is used to represent the current location or feature of interest; K is a key vector, used to indicate the position of the input sequence; V is a value vector containing the actual information of the input sequence; d k is the dimension of the key vector; T Represents the transpose of a matrix; is a spatiotemporal mask matrix used to suppress non-causal dependencies and invalid spatial associations; Introduce dynamic time warping loss function: in, L DTW is the total value of the dynamic time warping loss function, which is used to measure the comprehensive error of two time series under the optimal alignment path. At the same time, spatial covariance constraints are introduced to prevent overfitting. p is the optimal alignment path, For time series points x i and y i Local distance measure between; λ is the regularization coefficient; Σ is the spatial covariance matrix, which is calculated from the road network geometric parameters and dynamic construction data of the construction area; is the matrix trace operation; Multi-head attention through parallel computing h The attention head improves the ability to capture spatiotemporal features. The calculation formula is: Each attention head is defined as: in, , , is the learnable parameter matrix, is the output projection matrix.

6. The traffic simulation analysis method for highway construction area based on VISSIM as claimed in claim 1, characterized in that: In step S5, the path optimization method is: Based on the Q-learning reinforcement learning algorithm, the optimal driving path of the vehicle is dynamically planned according to the simulated traffic conditions and construction information. The Q-learning update formula is: in, Q ( s , a ) is in state s Select Action a The value function of α is the learning rate, r To perform an action a After receiving the instant reward, γ is the discount factor, Indicates the next state Select an action The maximum expected value.

7. The traffic simulation analysis method for highway construction area based on VISSIM as claimed in claim 1, characterized in that: In step S5, the safety warning method is: Building the Vehicle x The time-space trajectory matrix ; R means that the matrix belongs to the real field, T is the total number of simulation time steps, S is the number of spatial nodes; Calculate the number of any two cars k and m The probability of trajectory overlap: If the vehicle k At the time node t Located at the spatial node s ,but ,otherwise ; If the vehicle m At the time node t Located at the spatial node s ,but ,otherwise ; Set the triggering conditions for the warning: like , a yellow warning is triggered; like , then an orange warning is triggered; like , a red alert is triggered.

8. The traffic simulation analysis method for highway construction area based on VISSIM as claimed in claim 1, characterized in that: The simulation analysis method further includes: Step S6: 3D situation visualization display: A WebGL-based 3D rendering engine is used to build a traffic flow visualization scene in the construction area. This integrates real-time multi-source data, dynamic simulation models, and safety warning information with the visualization scene to display traffic flow changes and safety warning information in the construction area in real time.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, wherein: When the processor executes the computer program, the steps of the traffic simulation analysis method according to any one of claims 1 to 8 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the traffic simulation analysis method according to any one of claims 1 to 8 are implemented.

Images