Highway carbon emission simulation deduction system based on digital twinning

Abstract

The invention relates to the technical field of highway emission reduction, and discloses a highway carbon emission simulation deduction system based on digital twinning. The system comprises a carbon emission data acquisition layer, a digital twin modeling layer, a multi-dimensional carbon emission calculation layer, a simulation deduction optimization layer and an execution feedback adjustment layer. The carbon emission data acquisition layer acquires multi-source traffic parameters through a distributed sensor network, generates a dynamic carbon emission factor matrix and performs sensitivity grading; the digital twinborn modeling layer constructs a road network twinborn body, generates a road three-dimensional topological structure, and superposes a vehicle energy consumption model to form a dynamic twinborn scene; the multi-dimensional carbon emission calculation layer establishes a space-time mapping relation, and integrates emission data to generate a road section-level carbon emission intensity map; the simulation deduction optimization layer converts the atlas into a management and control strategy set, predicts a carbon emission change trend and outputs a Pareto optimal strategy combination; and executing feedback adjustment layer monitoring data, calculating a deviation rate, generating an adaptation degree index, and dynamically correcting model parameters until the index is stable.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of highway emission reduction technology, in particular to a highway carbon emission simulation and deduction system based on digital twinning. BACKGROUND

[0002] Current highway carbon emission control faces many challenges. The dynamic change characteristics of traffic flow make it difficult to accurately capture real-time carbon emission data. Traditional data collection methods rely on fixed monitoring points, which have limited coverage and are difficult to integrate vehicle driving state, road environment and other multi-source information, resulting in large errors in the calculation of carbon emission factors.

[0003] In terms of modeling, existing road network models are mostly static or semi-static, and cannot real-time integrate high-precision maps and real-time traffic data, making it difficult to truly reflect the dynamic changes of road three-dimensional topological structure and vehicle energy consumption, resulting in a large deviation between carbon emission simulation scenarios and actual road conditions.

[0004] In terms of carbon emission calculation, the emission data of mobile sources and fixed sources are often separated from each other, lacking effective spatio-temporal mapping relationship, and unable to form an accurate road section-level carbon emission intensity map, making it difficult to support fine-grained carbon emission control.

[0005] In the simulation and deduction process, traditional methods mostly develop control strategies based on experience, lack data-driven scientific analysis, and are difficult to predict the carbon emission trend under different strategies, resulting in insufficient pertinence and effectiveness of control measures. At the same time, the feedback mechanism after the implementation of the strategy is not perfect, and the model parameters cannot be corrected in time, so that the deviation between the simulation results and the actual situation gradually increases, affecting the continuous optimization of the control effect.

[0006] Different factors have different degrees of influence on carbon emissions, and existing technologies do not classify carbon emission factors by sensitivity, making it difficult to highlight key influencing factors in the modeling and calculation process, resulting in resource waste and low efficiency. These problems jointly restrict the accuracy and efficiency of highway carbon emission control, and a new system is needed to break through the limitations of existing technologies. SUMMARY

[0007] The purpose of the present application is to provide a highway carbon emission simulation and deduction system based on digital twinning to solve the problems raised in the background art.

[0008] To achieve the above purpose, the present application provides a highway carbon emission simulation and deduction system based on digital twinning, which comprises: A carbon emission data acquisition layer acquires multi-source traffic parameters through a distributed sensor network, processes vehicle driving state data, road environment data and meteorological data using a spatio-temporal fusion algorithm, generates a dynamic carbon emission factor matrix, and classifies the carbon emission factor matrix by sensitivity; The digital twin modeling layer constructs a road network twin according to the sensitivity ranking results, fuses high-precision maps and real-time traffic flow data, generates a three-dimensional topological structure of the road through multi-scale grid division, and superimposes a vehicle energy consumption model to form a dynamic twin scene. The multi-dimensional carbon emission calculation layer establishes a spatiotemporal mapping relationship between the twin scene and carbon emission factors, synchronizes the traffic flow state through an event triggering mechanism, integrates mobile source and fixed source emission data through a carbon emission accounting model, and generates a road section-level carbon emission intensity atlas. The simulation deduction optimization layer converts the carbon emission intensity atlas into a set of control strategies based on a reinforcement learning algorithm, predicts the carbon emission trend under different strategies through Monte Carlo simulation, and outputs a Pareto optimal strategy combination. The execution feedback adjustment layer monitors the carbon emission data after the implementation of the control strategy in real time, calculates the deviation rate between the actual carbon emission and the simulation prediction value, generates a strategy adaptation index, dynamically corrects the parameters of the carbon emission accounting model, and stops until the strategy adaptation index reaches a stable threshold.

[0009] Preferably, the method for obtaining multi-source traffic parameters through a distributed sensing network comprises: Millimeter wave radars, video monitoring devices, and environmental sensors are deployed along the highway to collect vehicle speed, traffic volume, vehicle type classification data, road surface temperature, and real-time wind direction and speed data. Multi-source traffic parameters include dynamic traffic flow data, static road attribute data, and meteorological interference data. An adaptive sampling mechanism is used to dynamically adjust the data collection frequency according to the traffic flow density. Based on historical congestion data, the entrance and exit of toll stations, bridge and tunnel areas, and interchanges are marked as high-frequency collection areas. Straight road segments are marked as low-frequency collection areas. A sampling frequency function is established to divide high-frequency collection areas and low-frequency collection areas through traffic flow density thresholds. High-frequency collection areas use millisecond-level sampling intervals, and low-frequency collection areas use second-level sampling intervals. Noise interference in multi-source traffic parameters is corrected in real time through a Kalman filter algorithm. The input of the Kalman filter algorithm is the original collected data, and the output is the denoised traffic parameters. The deployment position of the collection terminal is dynamically adjusted based on the denoised data. If a new congestion-prone area is detected, a new collection terminal is added in that area. If the traffic flow in the coverage area of the original collection terminal continuously falls below the threshold, the number of collection terminals in that area is reduced. The collection data from different terminals is integrated through a data synchronization protocol to obtain complete multi-source traffic parameters.

[0010] Preferably, the method for generating a dynamic carbon emission factor matrix by processing vehicle driving state data, road environment data, and meteorological data using a spatiotemporal fusion algorithm comprises: The multi-source traffic parameters are normalized, and the normalized parameters are arranged into a three-dimensional array according to time sequences and spatial positions, the dimensions of the three-dimensional array including road section numbers, collection time points and parameter types; a space-time fusion network structure is constructed, the structure including a time convolution module, a spatial attention module and a feature splicing layer; The three-dimensional array is input into the space-time fusion network structure, in the time convolution module, a convolutional neural network is used to extract time dimension features of the input data, generating time feature sequences of different lengths, the time feature sequences of different lengths including short-term feature sequences, medium-term feature sequences and long-term feature sequences; in the spatial attention module, the features of the spatial dimension are weighted and fused through a self-attention mechanism; a feature interaction channel is arranged between the time convolution module and the spatial attention module, and the time features and the spatial features of the same road section are spliced. A full connection layer is applied on an output layer of the space-time fusion network structure to compress the dimensions, converting the high-dimensional feature vector output by the network into a two-dimensional matrix to generate a dynamic carbon emission factor matrix.

[0011] Preferably, the method for grading the sensitivity of the carbon emission factor matrix comprises: A sensitivity coefficient of each element in the carbon emission factor matrix is calculated, the sensitivity coefficient is obtained by performing partial derivative calculation on vehicle speed data, road surface temperature data and wind direction and speed data in the multi-source traffic parameters; A first threshold value of the sensitivity coefficient and a second threshold value of the sensitivity coefficient are preset, the sensitivity coefficient is compared with the first threshold value of the sensitivity coefficient and the second threshold value of the sensitivity coefficient respectively, if the sensitivity coefficient is less than the first threshold value of the sensitivity coefficient, the parameter corresponding to the element is marked as a low sensitive factor; if the sensitivity coefficient is greater than the first threshold value of the sensitivity coefficient and less than the second threshold value of the sensitivity coefficient, the parameter corresponding to the element is marked as a medium sensitive factor; if the sensitivity coefficient is greater than the second threshold value of the sensitivity coefficient, the parameter corresponding to the element is marked as a high sensitive factor; The carbon emission factor matrix is graded in sensitivity using different identifiers, different identifiers representing different sensitivity levels, the low sensitive factor is defined as a regular monitoring item, the medium sensitive factor is defined as a key monitoring item, and the high sensitive factor is defined as a core monitoring item.

[0012] Preferably, the method for generating a road three-dimensional topology structure comprises: According to the sensitivity classification result, load high-precision road basic data in the digital twin platform; the high-precision road basic data includes road design parameters, road surface material properties and along-line facility distribution data; different modeling accuracies are adopted for parameters of different sensitivity levels, centimeter-level modeling accuracy is adopted for road sections corresponding to high sensitivity factors, decimeter-level modeling accuracy is adopted for road sections corresponding to medium sensitivity factors, and meter-level modeling accuracy is adopted for road sections corresponding to low sensitivity factors; a grid subdivision algorithm is applied to finely process the road model, terrain elevation data is introduced to correct the road three-dimensional coordinates, and the grid density is continuously adjusted through an iterative optimization algorithm, and the processing is stopped when the model error is less than a preset threshold; a three-dimensional modeling software is used to construct a road three-dimensional topology structure including lane lines, isolation belts and traffic signs, and finally a road three-dimensional topology structure with a sensitivity level mark is generated.

[0013] Preferably, the method for establishing the spatio-temporal mapping relationship between the twin scene and the carbon emission factor comprises: A unified scene coordinate system is defined with the starting point of the expressway as the origin, the X-axis parallel to the road center line, the Y-axis perpendicular to the road cross section, and the Z-axis in the elevation direction; a GPS positioning device is used to calibrate the collection terminal and obtain the position coordinates of the collection terminal; a coordinate conversion algorithm is used to convert the parameter coordinates of the dynamic carbon emission factor matrix to the scene coordinate system, and then the carbon emission factor distribution in the scene coordinate system is obtained; the origin of the scene coordinate system is taken as a reference datum, the coordinates of the digital twin scene are aligned, and the absolute coordinates of the twin scene are obtained; The road three-dimensional topology structure is converted to the scene coordinate system through a coordinate mapping algorithm, and then the twin scene model in the scene coordinate system is obtained; in the scene coordinate system, the carbon emission factor distribution and the twin scene model are spatio-temporally registered to establish a mapping relationship in the spatial dimension; the timestamp of the collection terminal is taken as a reference clock, the time axis of the twin scene is unified through a time synchronization protocol, and a mapping relationship in the time dimension is established.

[0014] Preferably, the method for integrating the emission data of mobile sources and fixed sources through the carbon emission accounting model comprises: According to the carbon emission factor distribution in the scene coordinate system and the twin scene model, an emission source network is constructed, and a carbon emission accounting model based on a graph neural network is used to integrate the emission data of mobile sources and fixed sources; the carbon emission accounting model comprises an input layer, a feature extraction layer, a source item fusion layer, a spatio-temporal correction layer and an output layer; the emission source network is input into the input layer of the carbon emission accounting model, and a road section-level carbon emission intensity atlas is generated through the output layer.

[0015] Preferably, the method for constructing the emission source network comprises: Each vehicle in the mobile source emission data in the scene coordinate system is taken as a mobile node, and the instantaneous emission rate of each vehicle is extracted as the feature of the mobile node; each facility in the fixed source emission data is taken as a fixed node, and the hourly total emission of each facility is extracted as the feature of the fixed node; all mobile nodes and fixed nodes are collected to obtain a node set; All mobile nodes are traversed, and the relative distance between any two mobile nodes in the scene coordinate system is calculated; a preset mobile node distance threshold is set, if the relative distance between any two mobile nodes in the scene coordinate system is less than the preset mobile node distance threshold, a directed edge is added between the corresponding two mobile nodes; if the relative distance between any two mobile nodes in the scene coordinate system is greater than or equal to the preset mobile node distance threshold, no edge is added; All fixed nodes are traversed, and the straight line distance between any two fixed nodes in the scene coordinate system is calculated; a preset fixed node distance threshold is set, if the straight line distance between any two fixed nodes in the scene coordinate system is less than the preset fixed node distance threshold, an undirected edge is added between the corresponding two fixed nodes; if the straight line distance between any two fixed nodes in the scene coordinate system is greater than or equal to the preset fixed node distance threshold, no edge is added; through the space index technology, the nearest fixed node to each mobile node is matched, and a directed edge between the mobile node and the corresponding nearest fixed node is added; all directed edges and undirected edges are collected to obtain an edge set; and the emission source network is constructed based on the obtained node set and edge set.

[0016] Preferably, the method for converting the carbon emission intensity map into a set of control strategies based on the reinforcement learning algorithm comprises: The expressway is divided into a plurality of control units, and each unit records the current carbon emission intensity value and traffic flow data; meanwhile, all adjustable control parameters are listed, including speed limit value, lane number, signal timing scheme and new energy vehicle incentive measures; Three optimization objectives are established, including carbon emission reduction target, traffic efficiency target and implementation cost target; two types of constraint conditions are set, including safety constraint and feasibility constraint; The reinforcement learning algorithm is used to initialize the strategy network parameters, generate a plurality of sets of control strategy samples by interacting with the environment, evaluate the achievement degree of the three optimization objectives for each sample, obtain the achievement degree scores of the three optimization objectives, and perform weighted calculation on the achievement degree scores of the three optimization objectives to obtain a comprehensive performance score, and select a set with the highest comprehensive performance score from the plurality of samples as the optimal control strategy; the selected optimal control strategy is converted into specific executable control instructions; the executable control instructions include speed limit instructions, lane control instructions and signal timing instructions; The method for obtaining the achievement degree scores of the three optimization objectives comprises: The reduction ratio of the controlled carbon emission intensity is calculated as a score of the achievement of the carbon emission reduction target; the road section traffic volume in a unit time is counted as a score of the achievement of the traffic efficiency target; and the total manpower and material resources required by the control measures are calculated as a score of the achievement of the implementation cost target.

[0017] Preferably, the method for generating the strategy adaptation index comprises: Actual carbon emission data after the execution of the control instruction is collected, the traffic operation state change is recorded synchronously, the deviation rate of the actual carbon emission amount and the simulation prediction value is calculated, the standardized strategy adaptation index with a value range of 0 to 1 is generated, the first threshold value of the standardized strategy adaptation index and the second threshold value of the standardized strategy adaptation index are preset, if the strategy adaptation index is greater than the second threshold value, it is judged that the control strategy is adapted, if the strategy adaptation index is between the first threshold value and the second threshold value, it is judged that the control strategy needs to be fine-tuned, and if the strategy adaptation index is less than the first threshold value, it is judged that the control strategy is not adapted, the carbon emission accounting model parameters are re-optimized, and the strategy updating process is triggered.

[0018] Compared with the prior art, the present application has the following beneficial effects: By means of the distributed sensing network, multi-source traffic parameters are acquired, the related data is processed by means of the space-time fusion algorithm, the dynamic carbon emission factor matrix is generated, and the sensitivity grading is simultaneously performed, so that the key factors influencing the carbon emission can be more accurately captured, and the determination of the carbon emission factor is more in line with the actual situation.

[0019] The digital twin modeling layer constructs a road network twin according to the sensitivity grading result, fuses the high-precision map and the real-time traffic flow data, generates the road three-dimensional topological structure through multi-scale grid division, superimposes the vehicle energy consumption model to form a dynamic twin scene, so that the constructed scene can reflect the dynamic changes of the road and the vehicle in real time, and the fitting degree with the actual road condition is higher, thereby providing a more real basic scene for the subsequent carbon emission calculation and simulation deduction.

[0020] The multi-dimensional carbon emission calculation layer establishes the space-time mapping relationship between the twin scene and the carbon emission factor, synchronizes the traffic flow state by means of the event triggering mechanism, integrates the mobile source and the fixed source emission data to generate the road section level carbon emission intensity atlas, realizes the effective fusion of different emission source data, makes the presentation of the carbon emission intensity more refined and targeted, and can clearly display the carbon emission conditions of different road sections.

[0021] The simulation deduction optimization layer converts the carbon emission intensity atlas into a control strategy set based on the reinforcement learning algorithm, predicts the carbon emission change trend under different strategies by means of the Monte Carlo simulation, and outputs the Pareto optimal strategy combination, so that the formulation of the control strategy is more scientific and systematic, and diversified strategy selection can be provided according to different carbon emission conditions.

[0022] The feedback adjustment layer monitors the carbon emission data after the implementation of the control strategy in real time, calculates the deviation rate of the actual carbon emission and the simulation prediction value, generates a strategy adaptation index and dynamically corrects the carbon emission accounting model parameters until the strategy adaptation index reaches a stable threshold, forming a closed-loop optimization process, so that the model can continuously adapt to changes in the actual situation and improve the fit of the strategy and the actual demand.

[0023] Each layer works cooperatively from data acquisition, modeling, calculation, deduction to feedback adjustment, forming a complete system covering all aspects of highway carbon emission control, which helps to realize the whole process and fine management of carbon emission, making the carbon emission control more dynamic and adaptive, and better responding to complex and variable highway traffic conditions. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 a working principle diagram of the highway carbon emission simulation deduction system based on digital twinning according to the present application; Figure 2 a flowchart of distributed sensor network data acquisition; Figure 3 a flowchart of road three-dimensional topological structure generation; Figure 4 a flowchart of carbon emission data integration by the carbon emission accounting model; Figure 5 a flowchart of generating control strategies by reinforcement learning. DETAILED DESCRIPTION

[0025] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0026] Please refer to Figure 1 The present application provides a highway carbon emission simulation deduction system based on digital twinning, which comprises: The carbon emission data acquisition layer obtains multi-source traffic parameters through a distributed sensor network, processes vehicle driving state data, road environment data and meteorological data by applying a space-time fusion algorithm, generates a dynamic carbon emission factor matrix, and classifies the sensitivity of the carbon emission factor matrix; The digital twinning modeling layer constructs a road network twin body according to the sensitivity classification results, fuses high-precision maps and real-time traffic flow data, generates a road three-dimensional topological structure through multi-scale grid division, and superimposes a vehicle energy consumption model to form a dynamic twin scene; The multi-dimensional carbon emission calculation layer establishes the spatial and temporal mapping relationship between the twin scene and the carbon emission factor, synchronizes the traffic flow state using an event-triggering mechanism, integrates the mobile source and fixed source emission data through a carbon emission accounting model, and generates a road section-level carbon emission intensity atlas; The simulation deduction optimization layer converts the carbon emission intensity atlas into a set of control strategies based on a reinforcement learning algorithm, predicts the carbon emission trend under different strategies through Monte Carlo simulation, and outputs a Pareto optimal strategy combination; The execution feedback adjustment layer monitors the carbon emission data after the implementation of the control strategy in real time, calculates the deviation rate between the actual carbon emission and the simulation prediction value, generates a strategy adaptation index, dynamically corrects the parameters of the carbon emission accounting model, and continues until the strategy adaptation index reaches a stable threshold.

[0027] Embodiment 1: see Figure 2 In the carbon emission data collection layer, the process of obtaining multi-source traffic parameters through a distributed sensing network needs to be implemented in combination with hardware deployment and algorithm optimization. Specifically, three types of collection terminals, including millimeter wave radars, video monitoring devices, and environmental sensors, are deployed along the highway according to functional requirements. The millimeter wave radars are installed on gantries or side poles above the road. The electromagnetic waves emitted by the radars can penetrate meteorological disturbances such as rain and fog, continuously capturing real-time speed and traffic volume data of passing vehicles, and distinguishing different vehicle types such as large trucks and small passenger cars through the characteristics of radar echoes. The video monitoring devices use high-definition cameras and are installed in areas with blocked vision such as curves and slopes. Through image recognition technology, they further refine vehicle type classification and supplement detailed information on vehicle type judgment for millimeter wave radars. The environmental sensors are distributed in the median strips on both sides of the road and are set at intervals in environmentally sensitive areas such as bridges and tunnel entrances and exits, collecting real-time data such as road surface temperature, atmospheric humidity, and wind direction and speed. Multi-source traffic parameters thus cover dynamic traffic flow data (vehicle speed, traffic volume, vehicle type classification), static road attribute data (road grade, number of lanes, road surface material), and meteorological disturbance data (wind direction and speed, road surface temperature).

[0028] An adaptive sampling mechanism is used to dynamically adjust the data collection frequency. The core is to divide the collection area based on traffic flow density and set the sampling interval. The system first calls the historical traffic database, filters the congestion occurrence frequency and duration of the past year in toll station entrances and exits, bridge and tunnel areas, and interchanges, and marks these areas as high-frequency collection areas. Straight road segments are marked as low-frequency collection areas because the traffic flow is relatively stable. A sampling frequency function is established, with traffic flow density threshold as the dividing basis. When the number of vehicles in a certain area per unit time exceeds the threshold, it automatically switches to high-frequency collection mode with millisecond-level sampling interval to ensure that instantaneous state changes such as vehicle sudden acceleration and sudden deceleration can be captured. When the number of vehicles is below the threshold, it switches to a second-level sampling interval to reduce unnecessary data redundancy.

[0029] The noise of the multi-source traffic parameters is corrected by a Kalman filtering algorithm, which takes the original data output by each collection terminal as input and processes through two steps of prediction and update. In the prediction stage, the traffic parameters at the next time are estimated according to the physical model of vehicle motion; in the update stage, the estimated values are corrected combined with the newly collected actual data, so as to filter out abnormal values caused by device vibration and electromagnetic interference, and output the denoised traffic parameters. Based on the denoised data, the system regularly analyzes the traffic flow trend of each region, if a road segment has congestion for a week and the original collection terminal coverage is insufficient, the deployment instruction of adding a new collection terminal is automatically triggered, and the device is temporarily installed through a mobile support; if the traffic flow of a certain area is continuously lower than the set threshold for a month, and the historical data shows that this state is stable, the instruction to reduce the collection terminal is issued, and part of the device is removed to reduce energy consumption. Finally, the collection data of different terminals is integrated through a data synchronization protocol to form a multi-source traffic parameter set with unified time stamp and standardized format.

[0030] When applying the spatio-temporal fusion algorithm to process vehicle driving state data, road environment data and meteorological data to generate a dynamic carbon emission factor matrix, first, the multi-source traffic parameters are normalized. The vehicle speed, traffic flow and other data are mapped to the interval of 0 to 1 according to their value range, eliminating the dimensional differences between different parameters. The processed parameters are arranged into a three-dimensional array according to the road section number, collection time and parameter type, for example, the coordinate (3, 15:30, 2) of an element in the array can represent the 2nd parameter (such as road surface temperature) collected at 15:30 on the 3rd road section.

[0031] The constructed spatio-temporal fusion network structure includes a time convolution module, a spatial attention module and a feature concatenation layer. After inputting the three-dimensional array into the network, the convolutional neural network in the time convolution module performs sliding calculation on the data through convolution kernels of different sizes to generate short-term feature sequences (such as traffic flow fluctuations within 5 minutes), medium-term feature sequences (such as traffic flow trends within 1 hour) and long-term feature sequences (such as 24-hour traffic cycle rules). The spatial attention module weights the importance of parameters for different road sections, for example, the weight of road surface temperature is higher than that of flat road sections in bridge areas, and the proportion of each road section parameter in the overall calculation is automatically adjusted through the self-attention mechanism. Feature interaction channels are set between the time convolution module and the spatial attention module to concatenate the time features (such as the speed change of a road section during the morning peak) and the spatial features (such as the slope and curvature of the road section) of the same road section, forming a comprehensive feature vector with spatio-temporal properties.

[0032] At the output layer of the spatiotemporal fusion network structure, the fully connected layer performs dimension compression on the high-dimensional feature vector, and converts the multi-dimensional data into a two-dimensional matrix through matrix operation. The rows of the matrix represent different road sections, the columns represent different time points, and the numerical values in the matrix reflect the carbon emission influence factor of the road section at the corresponding time point, thereby generating a dynamic carbon emission factor matrix. This matrix is dynamically updated over time, and when significant changes in traffic flow or meteorological conditions are detected, the numerical values in the matrix will be adjusted in real time to reflect the distribution of carbon emission influence factors under the current state.

[0033] Embodiment 2: refer to Figure 3 When classifying the sensitivity of the carbon emission factor matrix, it needs to be carried out from three links of sensitivity coefficient calculation, threshold division and grade identification. First, the sensitivity coefficient of each element in the carbon emission factor matrix is calculated, and the specific way is to perform partial derivative calculation on the vehicle speed data, road temperature data and wind direction and speed data in the multi-source traffic parameters. Through the analysis of the influence degree of the slight change of these parameters on the numerical value of the carbon emission factor, the sensitivity coefficient corresponding to each element is obtained.

[0034] The first threshold value of the sensitivity coefficient and the second threshold value of the sensitivity coefficient are preset, and the two threshold values are determined based on a large amount of historical carbon emission data and traffic parameter correlation analysis. The calculated sensitivity coefficient is compared with the two threshold values to divide the sensitivity level. If the sensitivity coefficient is less than the first threshold value, it means that the parameter corresponding to the element has little influence on the carbon emission factor, and it is marked as a low-sensitive factor; if the sensitivity coefficient is greater than the first threshold value and less than the second threshold value, it means that the parameter has a certain influence on the carbon emission factor, and it is marked as a medium-sensitive factor; if the sensitivity coefficient is greater than the second threshold value, it means that the parameter has a significant influence on the carbon emission factor, and it is marked as a high-sensitive factor.

[0035] The carbon emission factor matrix is classified by sensitivity using different identifiers, such as different colors or numerical codes as identifiers, and different identifiers correspond to different sensitivity levels. The low-sensitive factor is defined as a regular monitoring item, that is, data collection and monitoring are performed at a regular frequency; the medium-sensitive factor is a key monitoring item, which requires an increased monitoring frequency to more accurately capture its changes; the high-sensitive factor is a core monitoring item, which uses the highest monitoring frequency and the strictest monitoring standard to ensure that its dynamics can be mastered in real time.

[0036] When generating the three-dimensional topological structure of the road, the sensitivity classification results are used as the basis to load high-precision road basic data in the digital twin platform. These high-precision road basic data include road design parameters such as road width, design speed, and curve radius; road surface material properties such as asphalt or cement pavement and road surface roughness; and along-line facility distribution data such as the location of traffic signs, the type and length of the median strip, and the distribution of streetlights.

[0037] Different modeling accuracies are adopted for parameters of different sensitivity levels. For road segments corresponding to high-sensitivity factors, since the parameter changes have a greater impact on the carbon emission simulation results, centimeter-level modeling accuracy is adopted to present the small undulations of the road surface and the accurate positions of the lane lines in detail. For road segments corresponding to medium-sensitivity factors, decimeter-level modeling accuracy is adopted to clearly show the overall structure and main facilities of the road. For road segments corresponding to low-sensitivity factors, meter-level modeling accuracy is adopted to mainly present the general trend and basic layout of the road.

[0038] The road model is finely processed by using a mesh subdivision algorithm. According to the sensitivity levels and terrain characteristics of different road segments, the road model is divided into meshes of different sizes. The terrain elevation data is introduced to correct the three-dimensional coordinates of the road, so as to ensure that the road model matches the actual terrain. The mesh density is continuously adjusted by using an iterative optimization algorithm. In the iteration process, the model is compared with the actual measurement data, the model error is calculated, and the iteration is stopped when the model error is less than a preset threshold.

[0039] The three-dimensional topological structure of the road is constructed based on a three-dimensional modeling software. The number and position of the lane lines, the shape and range of the median strips, and the style and setting position of various traffic signs are accurately restored in the model. The sensitivity level identifier is integrated into the three-dimensional topological structure, for example, specific textures or marks are added to road segments corresponding to high-sensitivity factors, and finally the three-dimensional topological structure of the road with the sensitivity level identifier is generated, which provides an accurate road model basis for subsequent carbon emission simulation deduction.

[0040] Embodiment 3: refer to Figure 4 When establishing the spatiotemporal mapping relationship between the twin scene and the carbon emission factor, a unified scene coordinate system needs to be defined first. The origin is set at the starting point of the expressway, the X-axis is parallel to the road centerline and is set along the extension direction of the road, the Y-axis is perpendicular to the road cross section and points to the outside of the road, and the Z-axis is the elevation direction and is perpendicular to the horizontal plane and points upward. Through such coordinate system setting, the positions of all elements along the expressway can be quantitatively described. The GPS positioning device is used to calibrate the collection terminals deployed along the line, and the accurate position coordinates of each collection terminal in the coordinate system are obtained, with the coordinate values being accurate to the centimeter level, so as to ensure the spatial accuracy of subsequent data mapping.

[0041] The parameter coordinates of the dynamic carbon emission factor matrix are converted to the scene coordinate system through a coordinate conversion algorithm. Each parameter in the dynamic carbon emission factor matrix is originally recorded based on the local coordinates of the collection terminal. It needs to be mapped to a unified scene coordinate system through rotation, translation and scaling operations, and then generate the carbon emission factor distribution in the scene coordinate system. The distribution result is presented in grid form, each grid corresponds to a specific spatial range, and records the carbon emission factor value in that range. Take the origin of the scene coordinate system as the reference benchmark, align the coordinates of the digital twin scene, adjust the spatial position parameters of the twin scene, so that the elements such as roads and vehicles in the twin scene correspond one-to-one with the position of the actual highway, and obtain the absolute coordinates of the twin scene.

[0042] The three-dimensional topological structure of the road is converted to the scene coordinate system through a coordinate mapping algorithm. The three-dimensional topological structure of the road is originally constructed based on the relative coordinates of the design drawing. It needs to calculate its absolute position in the scene coordinate system through the coordinate mapping algorithm, and then generate the twin scene model in the scene coordinate system. The model contains the three-dimensional geometric shape of the road, lane division, spatial structure of bridges and tunnels, etc. In the scene coordinate system, the carbon emission factor distribution and the twin scene model are spatio-temporally registered. Spatial registration adjusts the grid position of the carbon emission factor distribution by calculating the spatial overlap of the two, so that it accurately corresponds to the road section in the twin scene model; time registration synchronizes the time stamps of the two, ensuring that the value change of the carbon emission factor and the change of the driving state of the vehicle in the twin scene are time consistent, thereby establishing the mapping relationship of spatial and temporal dimensions. Take the time stamp of the collection terminal as the reference clock, unify the time axis of the twin scene through the network time protocol, so that the time record of all dynamic elements in the twin scene (such as vehicle movement, signal light change) is synchronized with the collection terminal, and the time error is controlled within milliseconds.

[0043] The carbon emission data of mobile sources and fixed sources are integrated through a carbon accounting model. According to the distribution of carbon emission factors in the scene coordinate system and the twin scene model, an emission source network is constructed. Mobile sources include all vehicles driving on the highway, and fixed sources include office buildings at toll stations, charging pile facilities at service areas, road lighting equipment, etc. The carbon accounting model based on graph neural network includes five levels: input layer, feature extraction layer, source item fusion layer, space-time correction layer and output layer. The input layer receives the data of the emission source network and converts the emission parameters of mobile sources and fixed sources into vectors recognizable by the neural network. The feature extraction layer extracts the instantaneous emission characteristics of mobile sources (such as the surge of emissions when the vehicle accelerates) and the stable emission characteristics of fixed sources (such as the continuous energy consumption emissions of lighting equipment) through convolution operation. The source item fusion layer calculates the mutual influence between mobile sources and fixed sources through attention mechanism, for example, the vehicle emissions of a certain section of road will be affected by the superposition of emissions from nearby fixed sources in service areas, and the fusion layer will quantify this superposition effect. The space-time correction layer corrects the fused emission data according to the spatial distance and time interval in the scene coordinate system, for example, the farther the emission sources are from each other, the less they affect each other, and the emission data at different time points are not directly superimposed. The output layer converts the processed feature vector into a road section-level carbon intensity map through a fully connected neural network. The map shows the differences in carbon intensity of different road sections in the form of color gradient, with darker colors indicating higher carbon intensity. The final generated map covers the entire highway network, and the carbon intensity data of each road section is updated every 5 minutes.

[0044] The core operation formula of the carbon accounting model is:

[0045] wherein, represents the total carbon emissions of a certain road section, represents the mobile source carbon emissions of the road section, represents the fixed source carbon emissions of the road section, represents the interactive emissions between mobile sources and fixed sources, , , are the weight coefficients of the three, which are dynamically adjusted according to the type of road section (such as urban road section or suburban road section).

[0046] In the implementation example 4, the mobile source and the fixed source are first subjected to node processing when constructing the emission source network. Each vehicle in the mobile source emission data in the scene coordinate system is taken as a mobile node, and the data such as the instantaneous fuel injection amount and the rotating speed of the engine are collected through the OBD interface of the vehicle, combined with the vehicle type (such as heavy truck and small car) and the real-time speed, to calculate the instantaneous emission rate of each vehicle as the mobile node feature. For example, a heavy diesel truck driving at 90 km / h on the highway will update the instantaneous emission rate in real time according to the engine load, and this data will be continuously recorded as the core feature of the corresponding mobile node. Each facility in the fixed source emission data is taken as a fixed node. For the air conditioning system of the toll station, the hourly energy consumption data are read through the electricity meter and the gas meter, and the hourly total emission is converted. For the charging pile in the service area, the indirect carbon emission is calculated according to the charging power and the use time, as the fixed node feature. The mobile nodes corresponding to all the vehicles driving and the fixed nodes corresponding to all the facilities along the line are summarized to form a node set, and each node in the set contains a unique identifier, spatial coordinates and emission feature data.

[0047] All mobile nodes are traversed to calculate the relative distance between any two mobile nodes in the scene coordinate system. The relative distance is calculated by the difference between the spatial coordinates of the two points. For example, in the scene coordinate system, the coordinates of mobile node A are (x1, y1, z1), and the coordinates of mobile node B are (x2, y2, z2). The relative distance between the two is the straight-line distance between the two points in the three-dimensional space. A preset mobile node distance threshold is set according to the width of the highway lane and the vehicle safety distance, for example, 50 meters. When the relative distance between two driving vehicles is less than 50 meters, a directed edge is added between the corresponding two mobile nodes, and the direction of the edge is from the front vehicle to the rear vehicle, because the exhaust emission of the front vehicle will directly affect the driving environment of the rear vehicle. If the relative distance is greater than or equal to 50 meters, no edge is added, indicating that the emission influence between the two vehicles can be ignored.

[0048] Traverse all fixed nodes, calculate the straight-line distance between any two fixed nodes in the scene coordinate system. For example, the charging pile group of the service area and the nearby street lamp group calculate the straight-line distance through their coordinates. A preset fixed node distance threshold is set according to the emission influence range of the facility, for example, set to 100 meters. When the straight-line distance between two fixed nodes is less than 100 meters, add an undirected edge between the corresponding two fixed nodes, because their emissions will form a superposition effect in the local space; if the straight-line distance is greater than or equal to 100 meters, no edge is added. Through spatial indexing technology such as R-tree indexing, all fixed nodes are spatially divided, and the nearest fixed node is quickly matched for each mobile node. For example, a vehicle driving near a service area finds the nearest charging pile facility through spatial indexing, and adds a directed edge between the mobile node and the fixed node, with the direction pointing from the fixed node to the mobile node, because the emissions of the fixed source will spread to the area where the vehicle is located.

[0049] Collect all directed edges and undirected edges to form an edge set. Each node in the node set contains a unique identifier, spatial coordinates, and emission characteristics, and each edge in the edge set contains a connected node pair and the type of edge (directed or undirected). Based on the node set and the edge set, build an emission source network, which is presented in a graphical form, with nodes representing mobile sources and fixed sources, and edges representing the emission influence relationship between them. The network can be dynamically updated, when the vehicle's driving position changes, the coordinates of the mobile node are updated in real time, and the corresponding edges will also be added or reduced; when the running state of the fixed source changes (such as a facility stopping running), its node characteristics will be updated, and the attributes of the related edges will also be adjusted. The emission source network completely presents the spatial distribution of all emission sources on the highway and their mutual influence, providing a structured data foundation for subsequent carbon emission accounting.

[0050] Example 5: see Figure 5 When converting the carbon intensity map into a set of control strategies based on reinforcement learning algorithm, the highway is first divided into several control units. The division criteria include natural segmentation of the road (such as road segments between different interchanges), traffic flow characteristics (such as congested road segments during peak hours and smooth road segments), and differences in carbon intensity distribution, and the length of each control unit can be adjusted according to actual conditions. Each unit records the current carbon intensity value, which comes from the corresponding regional data of the road segment-level carbon intensity map, as well as traffic flow data, including the number of vehicles passing through per unit time, average speed, etc. List all adjustable control parameters, including speed limit values (such as adjusting the speed limit of a road segment from 120 km / h to 100 km / h), lane number (such as temporarily opening the emergency lane as a traffic lane to increase the number of lanes), signal timing scheme (such as adjusting the green light duration of the entrance signal at the toll station), and incentives for new energy vehicles (such as implementing toll fee reduction for new energy vehicles).

[0051] Three optimization objectives are established, including carbon emission reduction, traffic efficiency, and implementation cost. Two types of constraints are set, including safety constraints such as speed limit values not being lower than the minimum safe speed and lane adjustment not affecting emergency traffic, and feasibility constraints such as signal timing scheme adjustment complying with traffic signal control specifications and new energy vehicle incentive measures within the existing policy framework.

[0052] When using reinforcement learning algorithm, the policy network parameters are initialized first. The input of the network is the carbon intensity value and traffic flow data of the control unit, and the output is the possible control policy. Through interaction with the environment, multiple sets of control policy samples are generated, each containing a specific set of control parameter adjustment scheme. The achievement of the three optimization objectives is evaluated for each sample, and the achievement score of the three optimization objectives is obtained. The decline rate of carbon intensity after control is calculated as the achievement score of the carbon emission reduction objective. The road traffic volume per unit time is counted as the achievement score of the traffic efficiency objective. The total investment of manpower and material resources required for control measures (such as the number of traffic guidance personnel required for temporary opening of emergency lanes, equipment deployment cost, etc.) is calculated as the achievement score of the implementation cost objective. The achievement scores of the three optimization objectives are weighted and calculated, and the weights are determined according to the priority of the current traffic management. The comprehensive performance score is obtained, and the group with the highest comprehensive performance score is selected from the multiple samples as the optimal control policy. The selected optimal control policy is converted into specific executable control instructions, such as speed limit instructions through variable speed limit signs, lane control instructions through traffic cones and guide signs, and signal timing instructions through traffic signal control systems.

[0053] When the strategy adaptation index is generated, the actual carbon emission data after the execution of the control and management instruction can be obtained through the carbon emission monitoring equipment along the line. The real-time changes of traffic running state, including vehicle speed and traffic flow, are recorded synchronously. The deviation rate of the actual carbon emission and the simulation prediction value is calculated. The calculation method of the deviation rate is that the difference between the actual value and the prediction value is divided by the absolute value of the prediction value, and then the standardized strategy adaptation index with a value range of 0 to 1 is generated through standardization processing. The first threshold value of the standardized strategy adaptation index and the second threshold value of the standardized strategy adaptation index are preset, for example, the first threshold value is 0.6 and the second threshold value is 0.8. If the strategy adaptation index is greater than 0.8, it is judged that the control strategy is adaptive, and the execution can continue; if the strategy adaptation index is between 0.6 and 0.8, it is judged that the control strategy needs to be fine-tuned, such as appropriately adjusting the speed limit value or the signal lamp timing; if the strategy adaptation index is less than 0.6, it is judged that the control strategy is not adaptive, the carbon emission accounting model parameters are re-optimized, such as correcting the parameter coefficients in the vehicle energy consumption model, and the strategy updating process is triggered to generate a new control strategy. This process is repeated until the strategy adaptation index reaches a stable threshold, that is, the fluctuation range of the strategy adaptation index calculated for several times in succession is less than a set value.

[0054] It should be noted that, in this text, relational terms such as first and second are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "contain" or any other variant thereof are intended to cover non-exclusive inclusion, so that the process, method, article or equipment including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or equipment.

[0055] Although the embodiments of the present application have been shown and described, it can be understood by those skilled in the art that various changes, modifications, replacements and variations can be made to the embodiments without departing from the principles and spirits of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A highway carbon emission simulation system based on digital twins, characterized by: include: The carbon emission data collection layer acquires multi-source traffic parameters through a distributed sensor network, applies a spatiotemporal fusion algorithm to process vehicle driving status data, road environment data, and meteorological data, generates a dynamic carbon emission factor matrix, and performs sensitivity grading on the carbon emission factor matrix. The digital twin modeling layer constructs a road network twin based on the sensitivity classification results, integrates high-precision maps with real-time traffic data, generates a three-dimensional road topology through multi-scale grid division, and superimposes vehicle energy consumption models to form a dynamic twin scenario; The multi-dimensional carbon emission calculation layer establishes a spatiotemporal mapping relationship between twin scenarios and carbon emission factors, uses an event trigger mechanism to synchronize traffic flow status, integrates mobile and fixed source emission data through a carbon emission accounting model, and generates a road section-level carbon emission intensity map; The simulation and deduction optimization layer converts the carbon emission intensity map into a set of management and control strategies based on the reinforcement learning algorithm. It predicts the carbon emission trend under different strategies through Monte Carlo simulation and outputs the Pareto optimal strategy combination. The execution feedback adjustment layer monitors the carbon emission data after the execution of the control strategy in real time, calculates the deviation rate between the actual carbon emissions and the simulation prediction value, generates the strategy fitness index, and dynamically corrects the carbon emission accounting model parameters until the strategy fitness index reaches the stable threshold.

2. The highway carbon emission simulation and deduction system based on digital twin according to claim 1 is characterized in that: The method for obtaining multi-source traffic parameters through a distributed sensor network includes: Millimeter-wave radar, video surveillance equipment, and environmental sensors are deployed along highways to collect vehicle speed, traffic volume, vehicle type classification data, road surface temperature, and real-time wind direction and speed data. Multi-source traffic parameters include dynamic traffic flow data, static road attribute data, and meteorological interference data. Adopting an adaptive sampling mechanism, the data collection frequency is dynamically adjusted according to traffic flow density. Based on historical congestion data, toll station entrances and exits, bridge and tunnel areas, and interchange areas are marked as high-frequency collection areas, while straight road sections are marked as low-frequency collection areas. A sampling frequency function is established, and high-frequency collection areas are divided into low-frequency collection areas based on traffic flow density thresholds. High-frequency collection areas use millisecond sampling intervals, while low-frequency collection areas use second sampling intervals. The Kalman filter algorithm is used to correct noise interference in multi-source traffic parameters in real time. The input of the Kalman filter algorithm is the original collected data, and the output is the denoised traffic parameters. The deployment position of the collection terminal is dynamically adjusted according to the denoised data. If a new high-congestion area is detected, a new collection terminal is added to the area. If the traffic flow in the area covered by the original collection terminal is detected to be continuously below the threshold, the number of collection terminals in the area is reduced. The collection data of different terminals are integrated through the data synchronization protocol to obtain complete multi-source traffic parameters.

3. The highway carbon emission simulation and deduction system based on digital twin according to claim 2 is characterized in that: The method of applying a spatiotemporal fusion algorithm to process vehicle driving status data, road environment data, and meteorological data to generate a dynamic carbon emission factor matrix includes: Normalize the multi-source traffic parameters and arrange them into a three-dimensional array based on time series and spatial position. The dimensions of the three-dimensional array include the road section number, collection time, and parameter type. Build a spatiotemporal fusion network structure, which includes a temporal convolution module, a spatial attention module, and a feature splicing layer. The three-dimensional array is input into the spatiotemporal fusion network structure. In the temporal convolution module, a convolutional neural network is used to extract temporal features from the input data, generating temporal feature sequences of different lengths, including short-term feature sequences, medium-term feature sequences, and long-term feature sequences. In the spatial attention module, the spatial dimension features are weighted and fused through the self-attention mechanism. A feature interaction channel is set up between the temporal convolution module and the spatial attention module to splice the temporal and spatial features of the same road section. A fully connected layer is applied to the output layer of the spatiotemporal fusion network structure to perform dimensionality compression, converting the high-dimensional feature vector output by the network into a two-dimensional matrix to generate a dynamic carbon emission factor matrix.

4. The highway carbon emission simulation and deduction system based on digital twin according to claim 3 is characterized in that: The method for performing sensitivity grading on the carbon emission factor matrix includes: Calculate the sensitivity coefficient of each element in the carbon emission factor matrix by performing partial derivative calculations on vehicle speed data, road surface temperature data, and wind direction and speed data from multiple traffic sources to obtain the sensitivity coefficient; A first sensitivity coefficient threshold and a second sensitivity coefficient threshold are preset, and the sensitivity coefficient is compared with the preset first sensitivity coefficient threshold and the preset second sensitivity coefficient threshold respectively. If the sensitivity coefficient is less than the first sensitivity coefficient threshold, the parameter corresponding to the element is marked as a low sensitivity factor; if the sensitivity coefficient is greater than the first sensitivity coefficient threshold and less than the second sensitivity coefficient threshold, the parameter corresponding to the element is marked as a medium sensitivity factor; if the sensitivity coefficient is greater than the second sensitivity coefficient threshold, the parameter corresponding to the element is marked as a high sensitivity factor; Different labels are used to grade the sensitivity of the carbon emission factor matrix. Different labels represent different sensitivity levels. Low-sensitivity factors are defined as routine monitoring items, medium-sensitivity factors as key monitoring items, and high-sensitivity factors as core monitoring items.

5. The highway carbon emission simulation and deduction system based on digital twin according to claim 4 is characterized in that: The method for generating a three-dimensional road topological structure comprises: Based on the sensitivity grading results, high-precision road basic data are loaded into the digital twin platform; high-precision road basic data include road design parameters, pavement material properties and distribution data of facilities along the road; differentiated modeling accuracy is adopted for parameters with different sensitivity levels, with centimeter-level modeling accuracy adopted for road sections corresponding to high sensitivity factors, decimeter-level modeling accuracy adopted for road sections corresponding to medium sensitivity factors, and meter-level modeling accuracy adopted for road sections corresponding to low sensitivity factors; the road model is refined using a grid subdivision algorithm, and terrain elevation data is introduced to correct the three-dimensional coordinates of the road. The grid density is continuously adjusted through an iterative optimization algorithm, and the model stops when the error is less than the preset threshold; a three-dimensional road topology structure including lane lines, isolation belts and traffic signs is constructed based on the three-dimensional modeling software, and finally a three-dimensional road topology structure with sensitivity level marks is generated.

6. The highway carbon emission simulation and deduction system based on digital twin according to claim 5 is characterized in that: The method for establishing the spatiotemporal mapping relationship between the twin scenario and the carbon emission factor includes: Define a unified scene coordinate system with the starting point of the highway as the origin, the X-axis parallel to the road centerline, the Y-axis perpendicular to the road cross-section, and the Z-axis as the elevation direction. Use a GPS positioning device to calibrate the acquisition terminal to obtain its location coordinates. Use a coordinate conversion algorithm to convert the parameter coordinates of the dynamic carbon emission factor matrix to the scene coordinate system to obtain the carbon emission factor distribution in the scene coordinate system. Use the origin of the scene coordinate system as a reference to align the digital twin scene to obtain the absolute coordinates of the twin scene. The three-dimensional topological structure of the road is converted into the scene coordinate system through the coordinate mapping algorithm, and then the twin scene model in the scene coordinate system is obtained; in the scene coordinate system, the carbon emission factor distribution and the twin scene model are temporally and spatially aligned to establish a mapping relationship in the spatial dimension; the timestamp of the acquisition terminal is used as the reference clock, and the time axis of the twin scene is unified through the time synchronization protocol to establish a mapping relationship in the time dimension.

7. The highway carbon emission simulation and deduction system based on digital twin according to claim 6 is characterized in that: The method of integrating mobile source and stationary source emission data through the carbon emission accounting model includes: According to the carbon emission factor distribution and twin scenario model in the scene coordinate system, an emission source network is constructed, and the carbon emission accounting model based on graph neural network is used to integrate the emission data of mobile sources and fixed sources; the carbon emission accounting model includes input layer, feature extraction layer, source term fusion layer, spatiotemporal correction layer and output layer; the emission source network is used as the input layer of the carbon emission accounting model, and the section-level carbon emission intensity map is generated through the output layer.

8. The highway carbon emission simulation and deduction system based on digital twin according to claim 7 is characterized in that: The method for constructing an emission source network includes: Each vehicle in the mobile source emission data in the scene coordinate system is treated as a mobile node, and the instantaneous emission rate of each vehicle is extracted as the mobile node feature; each facility in the fixed source emission data is treated as a fixed node, and the total hourly emission of each facility is extracted as the fixed node feature; all mobile nodes and fixed nodes are collected to obtain a node set; Traverse all mobile nodes and calculate the relative distance between any two mobile nodes in the scene coordinate system; preset a mobile node distance threshold. If the relative distance between any two mobile nodes in the scene coordinate system is less than the preset mobile node distance threshold, add a directed edge between the corresponding two mobile nodes; if the relative distance between any two mobile nodes in the scene coordinate system is greater than or equal to the preset mobile node distance threshold, do not add it; Traverse all fixed nodes and calculate the straight-line distance between any two fixed nodes in the scene coordinate system; preset a fixed node distance threshold. If the straight-line distance between any two fixed nodes in the scene coordinate system is less than the preset fixed node distance threshold, add an undirected edge between the corresponding two fixed nodes; if the straight-line distance between any two fixed nodes in the scene coordinate system is greater than or equal to the preset fixed node distance threshold, do not add it; use spatial indexing technology to match the nearest fixed node for each mobile node, and add a directed edge between the mobile node and the corresponding nearest fixed node; collect all directed edges and undirected edges to obtain an edge set; construct an emission source network based on the obtained node set and edge set.

9. The highway carbon emission simulation and deduction system based on digital twin according to claim 8 is characterized in that: The method for converting the carbon emission intensity map into a set of management and control strategies based on a reinforcement learning algorithm includes: Divide the highway into several control units, each of which records current carbon emission intensity values ​​and traffic flow data. It also lists all adjustable control parameters, including speed limits, number of lanes, signal timing schemes, and incentives for new energy vehicles. Establish three optimization goals, including carbon emission reduction, traffic efficiency, and implementation cost; set two types of constraints, including safety constraints and feasibility constraints; Using a reinforcement learning algorithm, the policy network parameters are initialized. Multiple control policy samples are generated through interaction with the environment. The achievement of the three optimization objectives is evaluated for each sample, resulting in a score for the achievement of the three optimization objectives. These scores are then weighted to obtain a comprehensive effectiveness score. The optimal control policy with the highest comprehensive effectiveness score is selected from the multiple sample groups. The selected optimal control policy is then converted into specific executable control instructions. Executable control instructions include speed limit instructions, lane control instructions, and signal timing instructions. The methods for obtaining the achievement scores of the three optimization goals include: The achievement score of the carbon emission reduction target is calculated by calculating the reduction ratio of carbon emission intensity after control; the achievement score of the traffic efficiency target is calculated by counting the traffic volume of the road section per unit time; and the achievement score of the implementation cost target is calculated by calculating the total human and material resources required for the control measures.

10. The highway carbon emission simulation and deduction system based on digital twin according to claim 9 is characterized in that: The method for generating the strategy fitness index includes: Collect actual carbon emission data after the execution of the control command, synchronously record changes in traffic operation status, calculate the deviation rate between actual carbon emissions and simulation prediction values, and generate a standardized strategy fitness index with a value range of 0 to 1; preset the first threshold of the standardized strategy fitness index and the second threshold of the standardized strategy fitness index; if the strategy fitness index is greater than the second threshold, the control strategy is judged to be adapted; if the strategy fitness index is between the first threshold and the second threshold, the control strategy is judged to need fine-tuning; if the strategy fitness index is less than the first threshold, the control strategy is judged to be unsuitable, the carbon emission accounting model parameters are re-optimized, and the strategy update process is triggered.

Images