A truck platoon carbon emission reduction evaluation method and system based on trajectory data

Abstract

The application relates to a truck platoon carbon emission reduction evaluation method and system based on trajectory data, which comprises the following steps: preprocessing trajectory data to obtain preprocessed trajectory data; performing spatial position matching on trajectory points in the preprocessed trajectory data and candidate road sections in a road network to construct a continuous driving path; performing trajectory similarity calculation on trajectory data between different trucks to identify cooperative driving behaviors between two trucks; constructing a multi-truck cooperative driving graph, connecting edges of vehicle pairs that meet trajectory similarity and cooperation duration requirements, extracting connected components, identifying a preliminary cooperative driving vehicle set that meets a set condition, and performing merging judgment on the preliminary cooperative driving vehicle set to obtain a final cooperative driving vehicle set; and based on a vehicle dynamics and aerodynamics model, modeling and analyzing the final cooperative driving vehicle set to obtain a carbon emission reduction evaluation result. The application has the effect of improving the precision of road freight energy saving and emission reduction evaluation.

Description

Technical Field

[0001] This application relates to the technical field of methods for calculating carbon emissions from vehicle operation, and in particular to a method for assessing carbon emission reduction in truck platooning based on trajectory data. Background Technology

[0002] Currently, the continuous efforts to reduce carbon emissions have made controlling carbon emissions from road freight transportation a crucial aspect requiring breakthroughs. Heavy-duty trucks, as the mainstay of freight transport, are a key focus of industry governance due to their high energy consumption and high emissions. Truck platooning, through multi-vehicle coordinated driving to reduce aerodynamic drag and achieve fuel savings and carbon emission reductions, is an important research and application direction for intelligent connected transportation and green freight transport.

[0003] However, most existing studies rely on small-scale simulations or experimental scenarios, failing to accurately reflect the spontaneous platooning characteristics and carbon reduction effects of trucks in real-world road networks and multi-enterprise mixed traffic. At the same time, the lack of a comprehensive technical system capable of efficiently and accurately mining spontaneous platooning and its energy-saving effects from large-scale trajectory data results in a lack of scientific basis for assessing carbon emission reduction effects, hindering intelligent scheduling, policy incentives, and industry promotion.

[0004] The existing technical solutions mentioned above have the following drawbacks: the assessment of existing carbon emission reduction often relies on static models or estimates, and fails to incorporate actual vehicle operating parameters, dynamic equations and aerodynamic characteristic factors for accurate calculation, thus there is room for improvement. Summary of the Invention

[0005] To improve the accuracy of energy conservation and emission reduction assessments for road freight transport, this application provides a method for assessing carbon emission reduction in truck platooning based on trajectory data.

[0006] The above-mentioned objective of this application is achieved through the following technical solution:

[0007] A method for assessing carbon emission reduction in truck platooning based on trajectory data, comprising:

[0008] Collect truck trajectory data within the target area, preprocess the trajectory data, and obtain preprocessed trajectory data;

[0009] Based on the road network map matching algorithm, the trajectory points in the preprocessed trajectory data are spatially matched with the candidate road segments in the road network, and the transition probability is introduced to construct a continuous driving path.

[0010] Based on the improved longest common subsequence algorithm, trajectory similarity is calculated for trajectory data between different trucks to identify the cooperative driving behavior between the two trucks.

[0011] A multi-vehicle cooperative driving graph is constructed based on graph theory modeling. Edge connections are constructed for vehicle pairs that meet the requirements of trajectory similarity and cooperative driving time. Connected components are extracted, and a preliminary cooperative driving vehicle set that meets the set conditions is identified. The preliminary cooperative driving vehicle set is then merged and judged to obtain the final cooperative driving vehicle set.

[0012] Based on vehicle dynamics and aerodynamics models, the final cooperative driving vehicle set is modeled and analyzed to obtain carbon emission reduction assessment results, which include fuel savings and carbon emission reduction.

[0013] The identification information of the final cooperative driving vehicle set and the carbon emission reduction assessment results are structured and organized for use by the multi-dimensional classification statistics and scheduling system.

[0014] By adopting the above technical solutions, and collecting trajectory data from multiple trucks and performing preprocessing operations on the trajectory data, invalid or abnormal data can be eliminated, improving the accuracy of subsequent data analysis and thus laying a high-quality data foundation for cooperative driving behavior recognition. Based on the preprocessed trajectory data, cooperative formations with spatiotemporal consistency can be identified, accurately uncovering naturally formed cooperative driving behaviors and achieving energy-saving behavior discovery without additional communication control. By merging and judging the initial cooperative driving vehicle set to obtain the final cooperative driving vehicle set, the problem of splitting overlapping or partially consistent formations can be solved, thereby improving the comprehensiveness and stability of the recognition results. Based on the final cooperative driving vehicle set, a vehicle driving dynamics model can be constructed, and an air resistance influencing factor can be introduced for carbon emission reduction assessment, enabling quantitative analysis of the fuel-saving benefits brought by cooperative driving, thus providing a scientific basis for energy-saving optimization and carbon accounting in road transportation.

[0015] In one example, this application can be further configured such that: the preprocessing of the trajectory data to obtain preprocessed trajectory data specifically includes:

[0016] A cleaning operation is performed on the trajectory data to obtain cleaned trajectory data, wherein the cleaning operation includes removing trajectory points with abnormal speed, missing timestamps, duplicate timestamps, or invalid coordinates;

[0017] In the cleaned trajectory data, if the time interval between adjacent trajectory points is greater than a preset time threshold, the cleaned trajectory data is divided into different travel segments to generate a structured trajectory sequence.

[0018] The structured trajectory sequence is numbered and the corresponding vehicle identifier and time information are retained to obtain the preprocessed trajectory data.

[0019] By adopting the above technical solutions, the trajectory data is cleaned to remove anomalies such as abnormal speed, duplicate or missing timestamps, and invalid coordinates. This significantly improves the spatiotemporal continuity and authenticity of the trajectory data, thereby reducing misjudgments of cooperative driving and improving the reliability of the algorithm. By dividing trajectory points with large time intervals into different travel segments and numbering and labeling each trajectory sequence with vehicle identity and time, the original trajectory data can be standardized into a controllable input format, which facilitates subsequent multi-vehicle comparison and similarity calculation.

[0020] In one example, this application can be further configured as follows: the road network map matching algorithm spatially matches the trajectory points in the preprocessed trajectory data with candidate road segments in the road network, and introduces transition probabilities to construct continuous driving paths, specifically including:

[0021] Using the geographic coordinates of each trajectory point in the preprocessed trajectory data as the center, a spatial buffer with a preset radius is constructed, and multiple road segments falling within the spatial buffer are retrieved as a candidate road segment set.

[0022] For each candidate road segment, the matching probability is calculated based on the vertical distance from the trajectory point to the candidate road segment, and the transition probability between trajectory points is calculated based on the geographical distance between trajectory points and the shortest network distance between candidate road segments. The formula for calculating the matching probability is as follows: , This represents the perpendicular distance from a trajectory point to a road segment. The formula for calculating the transition probability is: where is the GPS positioning standard deviation. , For the shortest distance on the road network, This is the actual geographical distance. The attenuation coefficient is used; the actual geographical distance formula is: , For the Earth's radius, , This represents the difference in latitude and longitude between two points;

[0023] The optimal path matching sequence is determined based on the matching probability and transition probability to generate a continuous driving trajectory aligned with the road network.

[0024] By adopting the above technical solutions, constructing a spatial buffer zone centered on trajectory points and extracting candidate road segments can limit the search range, improve matching efficiency and accuracy, and thus reduce computational costs and matching errors. By calculating the matching probability based on the spatial distance between trajectory points and road segments, and by combining the geographical distance between points and the road network distance to calculate the transition probability, the spatiotemporal rationality of the trajectory can be comprehensively evaluated, thereby improving the map matching algorithm's ability to reconstruct actual driving paths. By fusing the matching probability and the transition probability to generate the optimal matching sequence, the continuous path trajectory of vehicles can be better reconstructed, thus providing a realistic and reliable basic trajectory for similarity recognition and path modeling.

[0025] In one example, this application can be further configured as follows: the improved longest common subsequence algorithm is used to calculate the trajectory similarity of trajectory data between different trucks and identify the cooperative driving behavior of the two trucks, specifically including:

[0026] The trajectories of any two vehicles are uniformly sampled and time-aligned to construct the time interval of trajectory overlap;

[0027] Based on the improved longest common subsequence algorithm, matching point pairs that meet the spatial distance and temporal difference thresholds are identified, and the spatial matching rate is calculated based on the number of matching point pairs.

[0028] The directional difference between the matching point pairs is calculated to generate directional similarity. This directional similarity is then fused with the spatial matching rate to obtain an overall trajectory similarity score. The formula for calculating directional similarity is: , For the number of paired points, , The heading angle and the formula for calculating the overall trajectory similarity score are as follows: ,in, For spatial matching rate, These are weighting coefficients;

[0029] If the trajectory similarity score is higher than the preset similarity threshold, and the continuous time period covered by the matching point pair is not less than the set collaboration duration threshold, then it is determined that the two vehicles are cooperating.

[0030] By adopting the above technical solutions, and by uniformly sampling and aligning the trajectories of any two vehicles to construct the trajectory overlap time interval, it is possible to ensure that the trajectory comparison process has a unified time basis, thereby improving the fairness and consistency of the matching calculation. By identifying matching point pairs that meet the spatial and temporal thresholds and calculating the spatial matching rate, it is possible to filter out temporary overlapping non-cooperative behaviors, thereby improving the accuracy of cooperative driving behavior recognition. By calculating the directional differences of matching point pairs and combining them with the spatial matching rate to generate trajectory similarity scores, it is possible to introduce directional consistency judgment while considering path proximity, thereby more comprehensively reflecting whether the vehicles are cooperatively driving. When both the trajectory similarity score and the duration exceed the threshold, it is determined that the vehicles are cooperatively driving, which can effectively exclude occasional proximity events, thereby ensuring that the finally identified cooperative relationship has stability and statistical significance.

[0031] In one example, this application can be further configured as follows: the multi-vehicle cooperative driving graph constructed based on graph theory modeling, constructing edge connections for vehicle pairs that meet the requirements of trajectory similarity and cooperative driving duration, extracting connected components, and identifying a preliminary set of cooperative driving vehicles that meet the set conditions, specifically includes:

[0032] All vehicle pairs that meet the criteria of trajectory similarity score higher than a preset threshold and whose time period covered by the matching point pair is not less than the collaboration duration threshold are constructed as edge connections in a graph structure, and an undirected graph is constructed with the corresponding vehicle as a node.

[0033] In the undirected graph, all connected components are extracted using a depth-first traversal, and the connectivity density of each connected component is calculated. The formula for calculating the connectivity density is: , The actual number of edges within the component, The number of vehicles included in the component, Density represents the degree of vehicle cooperation;

[0034] Components with a connectivity density greater than a set density threshold are identified as valid cooperative formations, and the obtained valid cooperative formations are statistically analyzed to obtain the preliminary cooperative driving vehicle set.

[0035] By adopting the above technical solution, and by constructing vehicle pairs that meet the requirements of similarity and collaboration duration as edge connections to generate an undirected graph, it is possible to formally model collaborative driving relationships in a graph structure, thereby facilitating the overall analysis of multi-vehicle relationships. By extracting connected components and calculating their connectivity density, it is possible to quantitatively determine the internal connection strength of collaborative subgraphs, thereby effectively filtering out vehicle sets with high collaboration levels. By identifying connected components with connectivity density exceeding a threshold as valid collaborative formations, edge connections and low-density pseudo-collaborations can be excluded, thereby enhancing the accuracy and reliability of collaboration identification.

[0036] In one example, this application can be further configured such that: the step of merging and judging the preliminary set of cooperative vehicles to obtain the final set of cooperative vehicles specifically includes:

[0037] Based on each cooperative formation in the initial cooperative driving vehicle set, the vehicle member overlap ratio and time interval overlap ratio of any two components are calculated, where the vehicle member overlap ratio is defined as: , The number of intersections of the members. For a smaller group of vehicles;

[0038] When the overlap ratio of vehicle members and the overlap ratio of time intervals both exceed the preset merging threshold, the corresponding cooperative formations are merged into a set of cooperative driving vehicles. All cooperative formation sets that meet the merging conditions are numbered to determine and output the final set of cooperative driving vehicles.

[0039] By adopting the above technical solution, the structural similarity and temporal consistency between collaborative segments can be measured in two ways by calculating the overlap ratio of vehicle members and the overlap ratio of time intervals between collaborative formations, thus providing an accurate and controllable basis for subsequent merging judgment. By merging and numbering collaborative formations with an overlap exceeding the merging threshold, the problems of duplicate counting and fragmented identification caused by different division granularities can be effectively avoided, thereby improving the overall consistency and analysis stability of the collaborative driving vehicle set.

[0040] In one example, this application can be further configured as follows: Based on vehicle dynamics and aerodynamics models, the final cooperative driving vehicle set is modeled and analyzed to obtain carbon emission reduction assessment results, specifically including:

[0041] For each vehicle in the final cooperative driving vehicle set, the required traction force of the vehicle is calculated based on the trajectory data. The formula for calculating the required traction force of the vehicle is as follows: , For vehicle quality, For acceleration, For rolling resistance, For aerodynamic drag;

[0042] Using the vehicle positional relationships and speed characteristics of each formation in the final cooperative driving vehicle set as aerodynamic correction factors, air resistance reduction coefficients for the lead vehicle and following vehicles are generated respectively.

[0043] Based on the required traction force and engine efficiency, the fuel consumption per unit time is estimated, and the corresponding carbon emissions are calculated in combination with the carbon emission factor.

[0044] By statistically analyzing the fuel savings and carbon emission reductions of each vehicle in the final cooperative driving vehicle set within the cooperative driving range, the carbon emission reduction assessment results of each cooperative vehicle in the final cooperative driving vehicle set are obtained.

[0045] By adopting the above technical solutions, and calculating the required traction force based on vehicle trajectory data, and incorporating multiple factors such as vehicle mass, acceleration, rolling resistance, and aerodynamic drag, a dynamic model that truly reflects the driving load can be constructed, thus providing a physical basis for fuel consumption estimation. By extracting the speed and position characteristics of the lead vehicle and following vehicles in the formation, air resistance reduction coefficients can be generated, enabling quantitative modeling of the aerodynamic energy-saving effect brought about by coordinated formation, thereby improving the accuracy of carbon emission reduction calculations. By combining traction force and engine efficiency to estimate fuel consumption per unit time, and then extrapolating carbon emissions through carbon emission factors, quantitative analysis of energy consumption and emissions throughout the entire process can be achieved, thus providing scientific data support for carbon accounting, energy-saving assessment, and scheduling strategy formulation. Finally, by statistically analyzing the fuel savings and carbon emission reductions of each vehicle within the coordinated driving range, the overall energy-saving and emission-reduction benefits of the formation can be comprehensively evaluated.

[0046] The second objective of this invention is achieved through the following technical solution:

[0047] A truck platooning carbon emission reduction assessment system based on trajectory data, the system comprising:

[0048] The trajectory acquisition and preprocessing module is used to acquire truck running trajectory data within the target area, preprocess the trajectory data, and obtain preprocessed trajectory data.

[0049] The trajectory matching construction module is used to perform spatial location matching between trajectory points in the preprocessed trajectory data and candidate road segments in the road network based on the road network map matching algorithm, and to introduce transition probability to construct a continuous driving path.

[0050] The trajectory similarity recognition module is used to calculate the trajectory similarity between different trucks based on the improved longest common subsequence algorithm, and to identify the cooperative driving behavior between the two trucks.

[0051] The collaborative graph modeling and merging module is used to construct a multi-vehicle collaborative driving graph based on graph theory modeling. It constructs edge connections for vehicle pairs that meet the requirements of trajectory similarity and collaborative duration, extracts connected components, identifies a preliminary collaborative driving vehicle set that meets the set conditions, and merges the preliminary collaborative driving vehicle set to obtain the final collaborative driving vehicle set.

[0052] The carbon emission reduction modeling module is used to model and analyze the final cooperative driving vehicle set based on vehicle dynamics and aerodynamics models to obtain carbon emission reduction assessment results, wherein the carbon emission reduction assessment results include fuel savings and carbon emission reductions.

[0053] The result structuring module is used to structure and organize the identification information of the final cooperative driving vehicle set with the carbon emission reduction assessment results for use by the multi-dimensional classification statistics and scheduling system.

[0054] By adopting the above technical solutions, and collecting trajectory data from multiple trucks and performing preprocessing operations on the trajectory data, invalid or abnormal data can be eliminated, improving the accuracy of subsequent data analysis and thus laying a high-quality data foundation for cooperative driving behavior recognition. Based on the preprocessed trajectory data, cooperative formations with spatiotemporal consistency can be identified, accurately uncovering naturally formed cooperative driving behaviors and achieving energy-saving behavior discovery without additional communication control. By merging and judging the initial cooperative driving vehicle set to obtain the final cooperative driving vehicle set, the problem of splitting overlapping or partially consistent formations can be solved, thereby improving the comprehensiveness and stability of the recognition results. Based on the final cooperative driving vehicle set, a vehicle driving dynamics model can be constructed, and an air resistance influencing factor can be introduced for carbon emission reduction assessment, enabling quantitative analysis of the fuel-saving benefits brought by cooperative driving, thus providing a scientific basis for energy-saving optimization and carbon accounting in road transportation.

[0055] In summary, this application includes the following beneficial technical effects:

[0056] 1. By collecting trajectory data from multiple trucks and performing preprocessing operations on the trajectory data, invalid or abnormal data can be removed, improving the accuracy of subsequent data analysis and thus laying a high-quality data foundation for cooperative driving behavior recognition. Based on the preprocessed trajectory data, cooperative formations with spatiotemporal consistency can be identified, and naturally formed cooperative driving behaviors can be accurately discovered, thereby achieving energy-saving behavior discovery without additional communication control.

[0057] 2. By merging and judging the initial set of vehicles traveling in coordination, the final set of vehicles traveling in coordination can be obtained, which can solve the problem of overlapping or partially consistent formations being split up, thereby improving the comprehensiveness and stability of the identification results. By constructing a vehicle driving dynamics model based on the final set of vehicles traveling in coordination and introducing the air resistance influence factor for carbon emission reduction assessment, it is possible to achieve quantitative analysis of the fuel-saving benefits brought by coordinated driving, thereby providing a scientific basis for energy-saving optimization and carbon accounting in road transportation. Attached Figure Description

[0058] Figure 1 This is a flowchart of a truck platooning carbon emission reduction assessment method based on trajectory data in one embodiment of this application;

[0059] Figure 2 This is a flowchart illustrating the implementation of step S10 in a truck platooning carbon emission reduction assessment method based on trajectory data in one embodiment of this application.

[0060] Figure 3 This is a flowchart illustrating the implementation of step S20 in a truck platooning carbon emission reduction assessment method based on trajectory data in one embodiment of this application.

[0061] Figure 4 This is a flowchart illustrating the implementation of step S30 in a truck platooning carbon emission reduction assessment method based on trajectory data in one embodiment of this application.

[0062] Figure 5 This is a flowchart illustrating the implementation of step S40 in a truck platooning carbon emission reduction assessment method based on trajectory data in one embodiment of this application.

[0063] Figure 6 This is another implementation flowchart of step S40 in a truck platooning carbon emission reduction assessment method based on trajectory data in one embodiment of this application;

[0064] Figure 7 This is a flowchart illustrating the implementation of step S50 in a truck platooning carbon emission reduction assessment method based on trajectory data in one embodiment of this application.

[0065] Figure 8 This is a schematic diagram of a truck platooning carbon emission reduction assessment system based on trajectory data, according to one embodiment of this application. Detailed Implementation

[0066] The present application will be further described in detail below with reference to the accompanying drawings.

[0067] In one embodiment, such as Figure 1 As shown, this application discloses a method for assessing carbon emission reduction in truck platooning based on trajectory data, which specifically includes the following steps:

[0068] S10: Collect truck trajectory data within the target area, preprocess the trajectory data, and obtain preprocessed trajectory data.

[0069] Specifically, when collecting truck trajectory data within the target area, raw trajectory data containing fields such as vehicle ID, timestamp, latitude and longitude, speed, and heading is obtained through a data access interface with the vehicle-mounted GPS terminal or roadside monitoring equipment. For noise points, drift points, and timestamp anomalies in the raw trajectory data, time interval standardization, trajectory smoothing filtering, and missing point interpolation are performed. The final output is preprocessed trajectory data with a unified format, enhanced coordinate accuracy, and continuous timestamps, serving as the basis for subsequent analysis. For example, for trajectory breaks caused by signal obstruction in some mountainous or tunnel scenarios, missing segments can be filled in using linear interpolation and Bézier curve fitting.

[0070] S20: Based on the road network map matching algorithm, the trajectory points in the preprocessed trajectory data are spatially matched with the candidate road segments in the road network, and the transition probability is introduced to construct a continuous driving path.

[0071] Specifically, when spatially matching trajectory points in the preprocessed trajectory data with candidate road segments in the road network, a road network map matching algorithm is used. First, a set of candidate road segments is constructed for each trajectory point. Then, the observation probability is calculated by combining the spatial distance and direction angle of the trajectory points with the consistency of the road segment orientation. A transition probability model is used, incorporating changes in vehicle speed and distance / direction between adjacent trajectory points as the basis for transition probabilities. Viterbi path decoding is then performed to reconstruct the optimal continuous driving path sequence, achieving accurate fitting between the vehicle's actual driving trajectory and the road network. For example, when a truck is traveling in the intersection of a highway and a ramp, the speed change trend between trajectory points can effectively determine whether it has entered the main road.

[0072] S30: Based on the improved longest common subsequence algorithm, the trajectory similarity of trajectory data between different trucks is calculated to identify the cooperative driving behavior of the two trucks.

[0073] Specifically, when calculating trajectory similarity between different truck trajectory data to identify collaborative driving behavior between two workshops, an improved Time Normalized Longest Common Subsequence (DTW-LCSS) algorithm is introduced. This algorithm performs time alignment and normalization on trajectories from different time periods and sampling frequencies. Then, matching point pairs are selected based on spatial distance and time difference thresholds between trajectory points. The trajectory similarity score is calculated using the ratio of the maximum matching length to the total trajectory length. When the score is higher than a set threshold and the duration exceeds a set collaborative time threshold, collaborative driving behavior between the two workshops is determined. For example, if two vehicles continuously travel more than 20 kilometers on a logistics trunk road and the point-to-point distance remains within 20 meters, a collaborative relationship can be considered to exist.

[0074] S40: Construct a multi-vehicle cooperative driving graph based on graph theory modeling, construct edge connections for vehicle pairs that meet the requirements of trajectory similarity and cooperative driving duration, extract connected components, identify the preliminary cooperative driving vehicle set that meets the set conditions, and merge and judge the preliminary cooperative driving vehicle set to obtain the final cooperative driving vehicle set.

[0075] Specifically, when constructing a multi-vehicle cooperative driving graph, each vehicle is first treated as a node in the graph. Undirected edges are established between vehicle pairs that meet the trajectory similarity requirement and whose cooperative duration exceeds a set threshold, resulting in a preliminary cooperative driving graph structure. All connected components are extracted using breadth-first search, with each connected component representing a possible cooperative driving vehicle group. The number of vehicles and cooperative strength indicators within each connected component are then analyzed. Subgraphs with insufficient members or substandard strength are eliminated, and subgraphs with highly overlapping node intersections are merged to form the final cooperative driving vehicle set that meets the cooperative scale requirements. For example, if three vehicle subsets have highly overlapping cooperative edges and share the same main road path, they will be merged into an effective cooperative vehicle convoy of five or more vehicles.

[0076] S50: Based on vehicle dynamics and aerodynamics models, the final cooperative driving vehicle set is modeled and analyzed to obtain carbon emission reduction assessment results, which include fuel savings and carbon emission reduction.

[0077] Specifically, when modeling and analyzing the final coordinated vehicle ensemble to obtain carbon emission reduction assessment results, the traction force required by each vehicle per unit time is calculated based on the vehicle dynamics model. The traction force is determined by the vehicle's mass and acceleration, and rolling resistance and air resistance are superimposed. Air resistance is further affected by vehicle speed and formation structure. A correction coefficient curve for the air resistance of the front and following vehicles is constructed. Then, the fuel consumption per unit time is estimated by combining the engine efficiency model. The carbon emission per unit time is obtained by multiplying it with the fuel carbon emission factor. Finally, the difference between the total fuel consumption and total carbon emissions under coordinated formation driving and independent driving is compared, and the fuel savings and carbon emission reductions are output as assessment results. For example, in a typical six-vehicle formation, the last five vehicles can reduce fuel consumption by 8%-15% respectively due to the drag reduction effect, and the overall carbon emission reduction can reach 12 kg per 100 kilometers.

[0078] S60: The identification information of the final cooperative vehicle set and the carbon emission reduction assessment results are structured and organized for use by the multi-dimensional classification statistics and scheduling system.

[0079] Specifically, when structuring the identification information of the final collaborative vehicle set and the carbon emission reduction assessment results, a collaborative information data table is constructed based on fields such as vehicle ID, collaborative time segment, and collaborative vehicle group number. The carbon emission reduction assessment results are mapped to vehicle information using fuel savings and carbon emission reduction as core fields. All data is uniformly organized and stored in structured JSON or CSV format, and supports multi-condition classification and filtering according to dimensions such as time range, vehicle group size, and regional range. Interfaces are also reserved for subsequent calls and display by traffic dispatching platforms, carbon trading management platforms, or fleet operation management systems. For example, the dispatching platform can prioritize recommending low-carbon and efficient fleets to participate in long-distance transportation tasks based on collaborative vehicle group number and fuel savings ranking.

[0080] By adopting the above technical solutions, and collecting trajectory data from multiple trucks and performing preprocessing operations on the trajectory data, invalid or abnormal data can be eliminated, improving the accuracy of subsequent data analysis and thus laying a high-quality data foundation for cooperative driving behavior recognition. Based on the preprocessed trajectory data, cooperative formations with spatiotemporal consistency can be identified, accurately uncovering naturally formed cooperative driving behaviors and achieving energy-saving behavior discovery without additional communication control. By merging and judging the initial cooperative driving vehicle set to obtain the final cooperative driving vehicle set, the problem of splitting overlapping or partially consistent formations can be solved, thereby improving the comprehensiveness and stability of the recognition results. Based on the final cooperative driving vehicle set, a vehicle driving dynamics model can be constructed, and an air resistance influencing factor can be introduced for carbon emission reduction assessment, enabling quantitative analysis of the fuel-saving benefits brought by cooperative driving, thus providing a scientific basis for energy-saving optimization and carbon accounting in road transportation.

[0081] In one embodiment, such as Figure 2 As shown, in step S10, the trajectory data is preprocessed to obtain preprocessed trajectory data, specifically including:

[0082] S11: Perform a cleaning operation on the trajectory data to obtain cleaned trajectory data. The cleaning operation includes removing trajectory points with abnormal speed, missing timestamps, duplicate timestamps, or invalid coordinates.

[0083] Specifically, perform data cleaning operations on the collected original trajectory data, including judging the speed value of each trajectory point. If the speed exceeds the set upper threshold (such as 120 km / h) or is negative, it is determined that the trajectory point is abnormal and excluded. At the same time, exclude trajectory points with missing or duplicate timestamps to ensure the time continuity of the trajectory. When there are null values, zero values, or non-standard formats (such as longitude and latitude exceeding the effective geographical range) in the coordinate field, the trajectory point needs to be excluded. For example, during the actual collection process, due to GPS loss, coordinate mutations or duplicate timestamps may occur in certain time periods. At this time, it is necessary to perform smoothing and clearing processing in combination with the position information of the previous and subsequent points to effectively avoid the influence of trajectory errors on the subsequent analysis results.

[0084] S12: In the cleaned trajectory data, if the time interval between adjacent trajectory points is greater than the preset time threshold, then divide the cleaned trajectory data into different trip segments to generate a structured trajectory sequence.

[0085] Specifically, traverse the set of cleaned trajectory points. For any two adjacent trajectory points, if their time interval is greater than the set threshold (such as 5 minutes), then divide the trajectory into different trip segments based on this time interval. For example, a truck collected continuous trajectory data from 10:00 to 10:10 in the morning, and there was a 10-minute data interruption after 10:10. The next trajectory record time was 10:20. Then it is considered that the vehicle was in an inactive state or data loss state from 10:10 to 10:20. Thus, the period from 10:00 to 10:10 is defined as a complete trip segment, and the points after 10:20 are assigned to the next new trip segment. This division strategy can effectively identify non-continuous running states such as vehicle stops and engine shutdowns, ensuring the continuity and analyzability of the trajectory sequence.

[0086] S13: Number the structured trajectory sequence and retain the corresponding vehicle identification and time information to obtain the preprocessed trajectory data.

[0087] Specifically, assign a unique serial number code to each segment of the structured trajectory sequence, and retain the vehicle identification field (such as license plate number or device ID) contained in the original data and the start and end timestamp information of this segment in each segment of the trajectory sequence. For example, the numbering method can be generated in the format of "license plate number_date_segment number", such as "粤A12345_20250919_03" indicating the third segment of the trajectory sequence of the vehicle on the specified date. In practical applications, this structured identification can support the efficient retrieval, archiving, and subsequent analysis of multi-vehicle and multi-time period trajectory data. At the same time, combined with vehicle and time information, multi-dimensional analysis of specific transportation tasks, operation efficiency, or path optimization can be realized, providing an accurate data basis for carbon emission reduction assessment.

[0088] In one embodiment, such as Figure 3 As shown, in step S20, based on the road network map matching algorithm, the trajectory points in the preprocessed trajectory data are spatially matched with candidate road segments in the road network, and transition probabilities are introduced to construct a continuous driving path, specifically including:

[0089] S21: Using the geographic coordinates of each trajectory point in the preprocessed trajectory data as the center, construct a spatial buffer with a preset radius, and retrieve multiple road segments falling within the spatial buffer as a candidate road segment set.

[0090] Specifically, for each preprocessed trajectory point, a spatial buffer with radius r and center is constructed based on its latitude and longitude coordinates. This radius value can be set according to the actual application scenario. For example, it is usually set to 20 to 50 meters in urban roads and within 100 meters in highways to adapt to different road network densities. Then, spatial indexing methods such as R-trees or quadtrees are used to quickly retrieve all road segments falling within the buffer range and these road segments are used as the candidate matching set for the trajectory point. For example, if a trajectory point is located at the exit of a highway toll station, its spatial buffer may contain three road segments: the exit ramp, the service area connecting road, and the main road. These will all be used as candidate objects to be evaluated.

[0091] S22: For each candidate road segment, calculate the matching probability based on the vertical distance from the trajectory point to the candidate road segment, and calculate the transition probability between trajectory points based on the geographical distance between trajectory points and the shortest network distance between candidate road segments. The formula for calculating the matching probability is: , This represents the perpendicular distance from a trajectory point to a road segment. The formula for calculating the transition probability is: where is the GPS positioning standard deviation. , For the shortest distance on the road network, This is the actual geographical distance. The attenuation coefficient is used; the actual geographical distance formula is: , For the Earth's radius, , This represents the difference in latitude and longitude between two points.

[0092] Specifically, for each trajectory point and each road segment in its candidate road segment set, the shortest perpendicular distance from the trajectory point to the road segment is first calculated. Then, based on this distance value and a Gaussian distribution or exponential decay function, the matching probability is calculated. The formula for calculating the matching probability is as follows: , This represents the perpendicular distance from a trajectory point to a road segment. The standard deviation of GPS positioning is given. For adjacent track point pairs, the geographical distance between them is calculated using the formula: , For the Earth's radius, , The difference between the latitude and longitude of the two points is compared with the shortest distance of the network path between candidate road segments. If the difference is small, it indicates that the transfer path is reasonable and can be assigned a higher transfer probability. The formula for calculating the transfer probability is: , For the shortest distance on the road network, This is the actual geographical distance. As a decay coefficient, for example, if the straight-line distance between two trajectory points is 80 meters and the shortest network distance between their corresponding candidate road segments in the road map is 85 meters, then the transfer probability is high; if the straight-line distance between two points is 100 meters and the shortest network distance between road segments is 300 meters, then it indicates that there is a path detour or matching error, and the transfer probability of this pair of road segments is low, thereby achieving a joint evaluation of spatial and logical coherence.

[0093] S23: Determine the optimal path matching sequence based on the matching probability and transition probability to generate a continuous driving trajectory aligned with the road network.

[0094] Specifically, by combining the matching probabilities of all trajectory points with the transition probabilities between trajectory point pairs, a complete trajectory matching graph model is constructed. Then, dynamic programming or the Viterbi algorithm is used to search for the sequence of paths with the highest joint probability in this graph, which is taken as the optimal trajectory matching result that is finally aligned with the road network. For example, in a trajectory sequence consisting of five trajectory points, if each point has three candidate road segments, the above algorithm can select a path sequence with the highest overall probability from 3-5 combinations. This makes the matching result not only reasonable in local location but also coherent in overall driving logic, thus forming a continuous, smooth driving trajectory that is consistent with the road topology, laying an accurate foundation for subsequent driving behavior analysis and carbon emission estimation.

[0095] In one embodiment, such as Figure 4 As shown, in step S30, based on the improved longest common subsequence algorithm, trajectory similarity is calculated for trajectory data between different trucks to identify the cooperative driving behavior of the two trucks. Specifically, this includes:

[0096] S31: Perform unified sampling and time alignment on the trajectories of any two vehicles to construct the time interval of trajectory overlap.

[0097] Specifically, for the preprocessed trajectory data of any two vehicles, the start and end time ranges of their trajectories are first determined. Based on a preset sampling period, a unified time axis sampling operation is performed on both trajectories. On this basis, non-aligned trajectory points are mapped to a unified time axis through linear interpolation or a nearest neighbor strategy, thereby ensuring that the two trajectory sequences have corresponding geographical location data at the same time point. Then, the segments where the two trajectories intersect in the time dimension are identified as the trajectory overlap time interval. For example, if vehicle A and vehicle B have driving trajectory data from 08:00:00 to 08:15:00 and 08:05:00 to 08:20:00 respectively, then their trajectory overlap interval is from 08:05:00 to 08:15:00. Within this interval, the trajectory points of the two vehicles are aligned and sampled every 5 seconds to build a basis for comparison.

[0098] S32: Based on the improved longest common subsequence algorithm, identify matching point pairs that meet the spatial distance and temporal difference thresholds, and calculate the spatial matching rate based on the number of matching point pairs.

[0099] Specifically, within the time interval of trajectory overlap, the improved longest common subsequence algorithm is used to determine whether the spatial distance between the two vehicles at the corresponding time point in the aligned trajectory sequence is less than a preset distance threshold, such as 30 meters, and at the same time, whether the time difference between the trajectory points of the two vehicles is within the tolerance range, such as ±2 seconds. If both conditions are met, they are marked as a matching point pair. Then, the total number of matching point pairs is counted and divided by the total number of sampling points in the overlapping interval to calculate the spatial matching rate of the two trajectories. For example, if 85 matching point pairs are identified in 120 sampling points, the spatial matching rate is 70.8%, which is used to measure the spatial proximity of the two vehicle trajectories.

[0100] S33: Calculate the directional difference between matching point pairs, generate directional similarity, and fuse the directional similarity with the spatial matching rate to obtain the overall trajectory similarity score. The formula for calculating directional similarity is: , For the number of paired points, , The heading angle and the formula for calculating the overall trajectory similarity score are as follows: ,in, For spatial matching rate, These are the weighting coefficients.

[0101] Specifically, for each pair of identified matching points, the directional difference value is obtained by calculating the angle between the motion vectors formed by adjacent trajectory points, and a directional similarity score is generated based on the average directional difference of all matching point pairs. The formula for calculating directional similarity is as follows: , For the number of paired points, , The heading angle is used as an example, with the proportion of matching points with an angle difference of less than 15° as the direction similarity index. Then, the spatial matching rate and direction similarity are fused according to a set weighting ratio. The formula for calculating the overall trajectory similarity score is as follows: ,in, For spatial matching rate, The weighting coefficients are used, for example, a weighting ratio of 0.6:0.4. The final trajectory similarity score is used as an indicator of the degree of coordination between the two vehicles in a specific time interval. For example, if the spatial matching rate is 72% and the directional similarity is 80%, then the trajectory similarity score after fusion is 0.6×72%+0.4×80%=75.2%.

[0102] S34: When the trajectory similarity score is higher than the preset similarity threshold, and the continuous time period covered by the matching point pair is not less than the set coordination duration threshold, it is determined that the two vehicles have coordinated driving behavior.

[0103] Specifically, the fused trajectory similarity score is compared with a set similarity threshold, for example, 70%. If the score is higher than this threshold, it is considered to have cooperative characteristics. At the same time, the continuous distribution of matching point pairs on the time axis is further checked to ensure that the continuous time length of the matching segment is not less than the cooperative duration threshold, for example, 180 seconds. If both conditions are met, it is determined that the two vehicles have cooperative driving behavior within this time interval. For example, if vehicles A and B continuously generate matching point pairs from 08:07:00 to 08:10:30 and the trajectory similarity score is 78.6%, then the judgment condition is met, and the cooperative driving event record can be output as a key basis for subsequent analysis.

[0104] In one embodiment, such as Figure 5 As shown, in step S40, a multi-vehicle cooperative driving graph is constructed based on graph theory modeling. Edge connections are established between vehicle pairs that meet the requirements of trajectory similarity and cooperative driving duration. Connected components are extracted, and a preliminary set of cooperative driving vehicles that meets the set conditions is identified. Specifically, this includes:

[0105] S41: Construct all vehicle pairs that meet the criteria of trajectory similarity score higher than the preset threshold and the time period covered by the matching point pair is not less than the collaboration duration threshold, and build an edge connection in the graph structure with the corresponding vehicle as the node to construct an undirected graph.

[0106] Specifically, the similarity calculation results of all pairwise vehicle trajectories are iterated. Vehicle pairs with trajectory similarity scores higher than a preset threshold (e.g., 70%) and whose continuous time period covered by the matching point pair is not less than a coordination duration threshold (e.g., 180 seconds) are marked as having a cooperative driving relationship. An undirected graph structure describing the cooperative driving relationship is constructed using these vehicle pairs as undirected edges and each vehicle as a graph node. Within each time window t, an undirected graph is constructed. ,in This represents the set of currently active vehicles. A set of vehicle pairs that satisfy the following conditions:

[0107]

[0108]

[0109]

[0110] Let be the trajectory similarity between vehicles i and j. This is the similarity threshold. The duration of coordinated driving between the two vehicles.

[0111] This is the threshold for the duration of collaboration. The time interval for the two vehicles to coordinate. This represents the current time window for analysis. For example, if vehicles A and B satisfy the conditions, B and C satisfy the conditions, but A and C do not satisfy the conditions, then the graph contains edges AB and BC, but no edge AC, forming a partially connected structure with three nodes. This graph structure provides the basis for subsequent extraction of cooperative groups.

[0112] S42: In an undirected graph, extract all connected components using depth-first traversal and calculate the connectivity density of each connected component. The formula for calculating the connectivity density is: , The actual number of edges within the component, The number of vehicles included in the component, Density represents the degree of vehicle coordination.

[0113] Specifically, a depth-first traversal operation is performed on the aforementioned constructed undirected graph to identify all connected components as potential cooperative driving subgroups. Then, for each connected component, the number of vehicles C and the number of edges EC are counted, and the results are calculated according to the formula... Calculate the connectivity density index ρ of the component, where the value of ρ ranges from [0,1], representing the degree of cooperation between vehicles in the component. For example, if a component contains 4 vehicles (C=4) and forms 5 edges that meet the conditions (EC=5), then the density is ρ=2×5 / 4×3=10 / 12≈0.83, indicating that the cooperation between vehicles in this subgroup is relatively close.

[0114] S43: Identify components with connectivity density greater than a set density threshold as valid cooperative formations, and statistically analyze the obtained valid cooperative formations to obtain a preliminary set of cooperative vehicles.

[0115] Specifically, the density ρ calculated for each connected component is compared with a preset density threshold, for example, a density threshold of 0.75. If the ρ of a connected component exceeds this threshold, it is considered a valid cooperative formation. All vehicles contained within it are marked as cooperative vehicles, and they are counted into the preliminary identified cooperative vehicle set on a component basis. For example, if three valid components are extracted, containing 5, 3, and 6 vehicles respectively, the preliminary cooperative set contains 14 vehicles, which can provide input for subsequent trajectory fusion and multi-vehicle behavior modeling.

[0116] In one embodiment, such as Figure 6 As shown, in step S40, the initial set of vehicles traveling in cooperative mode is merged and judged to obtain the final set of vehicles traveling in cooperative mode, which specifically includes:

[0117] S44: Based on each cooperative formation in the initial cooperative driving vehicle set, calculate the vehicle member overlap ratio and time interval overlap ratio for any two components, where the vehicle member overlap ratio is defined as: , The number of intersections of the members. For a smaller group of vehicles.

[0118] Specifically, iterate through any two collaborative components in the initial set of collaborative vehicles, first extract their corresponding vehicle member sets C1 and C2, and calculate the number of elements in their intersection C1∩C2. Simultaneously, take the smaller of |C1| and |C2| as the denominator to calculate the vehicle member overlap ratio Ω. For example, if two components contain 6 and 4 vehicles respectively, and 3 of them are the same vehicle, then the overlap ratio Ω is 3 / 4 = 0.75. Also, based on the time intervals covered by the two components, take the ratio of the length of the overlapping portion to the length of the smaller time interval as the time interval overlap ratio. For example, if one component is [08:00, 08:20] and the other is [08:10, 08:25], with an overlap interval of 10 minutes and a smaller interval of 20 minutes, then the time overlap ratio is 0.5. These two indicators will be used together to determine whether collaborative components need to be merged.

[0119] S45: When the overlap ratio of vehicle members and the overlap ratio of time intervals both exceed the preset merging threshold, the corresponding cooperative formations will be merged into a set of cooperative driving vehicles. All cooperative formation sets that meet the merging conditions will be numbered to determine and output the final set of cooperative driving vehicles.

[0120] Specifically, all cooperative components whose overlap ratio Ω exceeds a set threshold (e.g., 0.6) and whose time interval overlap ratio also exceeds the corresponding threshold (e.g., 0.4) are merged into the same cooperative driving vehicle set. This merging logic is repeated until there are no new components to merge. Each cooperative driving vehicle set is assigned a unique number. For example, if 6 cooperative components are initially identified, after merging, 3 final sets are formed, numbered Group_1, Group_2, and Group_3, respectively. The vehicle numbers, cooperative driving time ranges, and other cooperative behavior characteristics contained in each set are output, providing basic data support for subsequent energy consumption comparison and carbon emission estimation.

[0121] In one embodiment, such as Figure 7 As shown, in step S50, based on vehicle dynamics and aerodynamics models, the final cooperative driving vehicle set is modeled and analyzed to obtain carbon emission reduction assessment results, specifically including:

[0122] S51: For each vehicle in the final cooperative driving vehicle set, calculate the required traction force of the vehicle based on the trajectory data. The formula for calculating the required traction force of the vehicle is as follows: , For vehicle quality, For acceleration, For rolling resistance, This represents aerodynamic drag.

[0123] Specifically, for each vehicle in the final coordinated driving vehicle set, the instantaneous acceleration 'a' is calculated based on the velocity change between adjacent trajectory points in its trajectory data. Then, the rolling resistance 'F' is determined by combining the vehicle mass 'm', the road rolling coefficient, and the vehicle speed. roll With aerodynamic drag F aero and substitute The formula completes the traction force calculation, and the rolling resistance and aerodynamic resistance are respectively:

[0124]

[0125]

[0126] g is the acceleration due to gravity, c r Where ρ is the rolling resistance coefficient, ρ is the air density, A is the frontal area, and C is the air density. d Where v is the drag coefficient and v is the velocity.

[0127] For example, for a vehicle with a mass of 1600 kg, an acceleration of 0.5 m / s² at a certain time period, an estimated rolling resistance of 120 N, and an aerodynamic drag of 180 N, then the traction force F trac =1600×0.5+120+180=1100N.

[0128] S52: Using the positional relationships and speed characteristics of vehicles in each formation within the final coordinated driving vehicle set as aerodynamic correction factors, air resistance reduction coefficients are generated for the lead vehicle and following vehicles, respectively.

[0129] Specifically, the relative positions and speeds of vehicles in each coordinated formation are traversed. Based on whether they are in a typical car-following configuration, aerodynamic drag correction factors and speed adjustment factors are set according to an empirical model. v is the vehicle's current speed (unit: km / h), and the scale adjustment factor is: N is the current number of vehicles in the platoon, and the energy efficiency of the lead vehicle is: Fuel efficiency of following vehicle: For example, when the distance between two vehicles is less than 20 meters and the speed difference is less than 2 km / h, the following vehicle can obtain an air resistance reduction coefficient of 0.3, while the lead vehicle may obtain a reduction effect of 0.05. This allows for dynamic correction of the aerodynamic drag Faero term in S51 to more accurately estimate the traction force. For example, the original Faero is 180 N, which is corrected to 126 N after applying the reduction coefficient of 0.3.

[0130] S53: Estimate fuel consumption per unit time based on required traction and engine efficiency, and calculate the corresponding carbon emissions by combining carbon emission factors.

[0131] Specifically, for each cooperative vehicle at each trajectory point, its traction power is estimated using its traction force and real-time speed, and fuel power is calculated using an engine thermal efficiency model. Then, fuel consumption per unit time is calculated based on the unit fuel calorific value, resulting in instantaneous power. Fuel consumption rate: , Positive power indication. Fuel conversion coefficient, For engine efficiency, For diesel fuel with a lower heating value, the carbon emissions are: , To accumulate fuel consumption, The carbon emission factor is used as an example. For instance, if the traction power is 15kW and the engine efficiency is 0.3 at a certain moment, the fuel input power is 50kW, which translates to a fuel consumption of approximately 6.25L per hour. Combined with the standard carbon emission factor of 2.63kg / L, the carbon emission is calculated to be 16.44kg / h. By accumulating the values ​​hourly, a complete carbon emission curve can be obtained.

[0132] S54: Calculate the fuel savings and carbon emission reductions of each vehicle in the final cooperative driving vehicle set within the cooperative driving range, and obtain the carbon emission reduction assessment results of each cooperative vehicle in the final cooperative driving vehicle set.

[0133] Specifically, the corrected fuel consumption of each vehicle within the cooperative driving range is compared with the fuel consumption benchmark under the corresponding non-platoon driving condition. The fuel savings are calculated, and the corresponding carbon emission reduction is calculated simultaneously. Finally, the fuel savings and CO2 emission reduction generated by each vehicle in the cooperative driving behavior are output. For example, if a vehicle consumes 0.6L less fuel during a 15-minute cooperative driving period than when driving alone, the carbon emission reduction is 0.6 × 2.63 = 1.578 kg. The carbon emission reduction contribution of each vehicle is statistically analyzed by cooperative group for subsequent environmental benefit analysis and optimization suggestion generation.

[0134] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0135] In one embodiment, a truck platooning carbon emission reduction assessment system based on trajectory data is provided, which corresponds one-to-one with the truck platooning carbon emission reduction assessment method based on trajectory data in the above embodiments. For example... Figure 8 As shown, this truck platooning carbon emission reduction assessment system based on trajectory data includes a trajectory acquisition and preprocessing module, a trajectory matching and construction module, a trajectory similarity recognition module, a platooning graph modeling and merging module, a carbon emission reduction modeling module, and a result structuring module. Detailed descriptions of each functional module are as follows:

[0136] The trajectory acquisition and preprocessing module is used to acquire truck running trajectory data within the target area, preprocess the trajectory data, and obtain preprocessed trajectory data.

[0137] The trajectory matching construction module is used to perform spatial location matching between trajectory points in preprocessed trajectory data and candidate road segments in the road network based on the road network map matching algorithm, and to introduce transition probability to construct continuous driving paths.

[0138] The trajectory similarity recognition module is used to calculate the trajectory similarity between different trucks based on the improved longest common subsequence algorithm, and to identify the cooperative driving behavior between the two trucks.

[0139] The collaborative graph modeling and merging module is used to construct a multi-vehicle collaborative driving graph based on graph theory modeling. It constructs edge connections for vehicle pairs that meet the requirements of trajectory similarity and collaborative duration, extracts connected components, identifies a preliminary collaborative driving vehicle set that meets the set conditions, and merges the preliminary collaborative driving vehicle set to obtain the final collaborative driving vehicle set.

[0140] The carbon emission reduction modeling module is used to model and analyze the final cooperative vehicle set based on vehicle dynamics and aerodynamics models to obtain carbon emission reduction assessment results, which include fuel savings and carbon emission reduction.

[0141] The results structuring module is used to organize the identification information of the final cooperative vehicle set and the carbon emission reduction assessment results in a structured manner for use by the multi-dimensional classification statistics and scheduling system.

[0142] Optionally, the trajectory acquisition and preprocessing module includes:

[0143] The trajectory cleaning submodule is used to perform cleaning operations on trajectory data to obtain cleaned trajectory data. The cleaning operation includes removing trajectory points with abnormal speed, missing timestamps, duplicate timestamps, or invalid coordinates.

[0144] The trajectory segmentation module is used to divide the cleaned trajectory data into different travel segments if the time interval between adjacent trajectory points is greater than a preset time threshold, so as to generate a structured trajectory sequence.

[0145] The trajectory annotation submodule is used to number the structured trajectory sequences and retain the corresponding vehicle identifiers and time information to obtain preprocessed trajectory data.

[0146] Optionally, the trajectory matching building blocks include:

[0147] The buffer retrieval submodule is used to construct a spatial buffer with a preset radius, centered on the geographic coordinates of each trajectory point in the preprocessed trajectory data, and to retrieve multiple road segments falling within the spatial buffer as a candidate road segment set.

[0148] The probability calculation submodule is used to calculate the matching probability for each candidate road segment based on the vertical distance from the trajectory point to the candidate road segment, and to calculate the transition probability between trajectory points based on the geographical distance between trajectory points and the shortest network distance between candidate road segments. The formula for calculating the matching probability is as follows: , This represents the perpendicular distance from a trajectory point to a road segment. The formula for calculating the transition probability is: where is the GPS positioning standard deviation. , For the shortest distance on the road network, This is the actual geographical distance. The attenuation coefficient is used; the actual geographical distance formula is: , For the Earth's radius, , This represents the difference in latitude and longitude between two points;

[0149] The path generation submodule is used to determine the optimal path matching sequence based on the matching probability and transition probability, so as to generate a continuous driving trajectory aligned with the road network.

[0150] Optionally, the trajectory similarity recognition module includes:

[0151] The trajectory alignment submodule is used to uniformly sample and align the trajectories of any two vehicles in time, and construct the trajectory overlap time interval;

[0152] The matching point identification submodule is used to identify matching point pairs that meet the spatial distance and temporal difference thresholds based on the improved longest common subsequence algorithm, and to calculate the spatial matching rate based on the number of matching point pairs.

[0153] The similarity calculation submodule is used to calculate the directional difference between matching point pairs, generate directional similarity, and fuse the directional similarity with the spatial matching rate to obtain the overall trajectory similarity score. The formula for calculating directional similarity is: , For the number of paired points, , The heading angle and the formula for calculating the overall trajectory similarity score are as follows: ,in, For spatial matching rate, These are weighting coefficients;

[0154] The behavior determination submodule is used to determine that the two vehicles have cooperative driving behavior when the trajectory similarity score is higher than the preset similarity threshold and the continuous time period covered by the matching point pair is not less than the set cooperative duration threshold.

[0155] Optionally, the collaboration graph modeling and merging module includes:

[0156] The graph construction submodule is used to construct all vehicle pairs that meet the criteria of trajectory similarity score higher than a preset threshold and the time period covered by the matching point pair is not less than the collaboration duration threshold as edge connections in the graph structure, and construct an undirected graph with the corresponding vehicle as the node.

[0157] The connectivity extraction submodule is used to extract all connected components in an undirected graph using a depth-first traversal and to calculate the connectivity density of each connected component. The formula for calculating the connectivity density is: , The actual number of edges within the component, The number of vehicles included in the component, Density represents the degree of vehicle cooperation;

[0158] The preliminary set identification submodule is used to identify components with a connectivity density greater than a set density threshold as valid cooperative formations, and to statistically analyze the obtained valid cooperative formations to obtain a preliminary set of cooperative vehicles.

[0159] The overlap calculation submodule is used to calculate the vehicle member overlap ratio and time interval overlap ratio of any two components based on each cooperative formation in the initial cooperative driving vehicle set. The vehicle member overlap ratio is defined as follows: , The number of intersections of the members. For a smaller group of vehicles;

[0160] The set merging submodule is used to merge the corresponding collaborative formations into a set of collaborative driving vehicles when the overlap ratio of vehicle members and the overlap ratio of time intervals both exceed the preset merging threshold. It also performs numbering processing on all collaborative formation sets that meet the merging conditions, determines and outputs the final set of collaborative driving vehicles.

[0161] Optionally, the carbon reduction modeling module includes:

[0162] The traction calculation submodule is used to calculate the required traction force for each vehicle in the final cooperative driving vehicle set based on trajectory data. The formula for calculating the required traction force for a vehicle is as follows: , For vehicle quality, For acceleration, For rolling resistance, For aerodynamic drag;

[0163] The aerodynamic correction submodule is used to take the positional relationship and speed characteristics of each formation in the final coordinated driving vehicle set as aerodynamic correction factors to generate the air resistance reduction coefficients for the lead vehicle and the following vehicles, respectively.

[0164] The fuel consumption and emissions calculation submodule is used to estimate fuel consumption per unit time based on the required traction force and engine efficiency, and to calculate the corresponding carbon emissions by combining the carbon emission factor.

[0165] The evaluation results generation submodule is used to calculate the fuel savings and carbon emission reductions of each vehicle in the final cooperative driving vehicle set within the cooperative driving range, and obtain the carbon emission reduction evaluation results of each cooperative vehicle in the final cooperative driving vehicle set.

[0166] Specific limitations regarding a truck platooning carbon emission reduction assessment system based on trajectory data can be found in the above description of a truck platooning carbon emission reduction assessment method based on trajectory data, and will not be repeated here. Each module in the aforementioned truck platooning carbon emission reduction assessment system based on trajectory data can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0167] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above.

[0168] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for assessing carbon emission reduction in truck platooning based on trajectory data, characterized in that, The truck platooning carbon emission reduction assessment method based on trajectory data includes: Collect truck trajectory data within the target area, preprocess the trajectory data, and obtain preprocessed trajectory data; Based on the road network map matching algorithm, the trajectory points in the preprocessed trajectory data are spatially matched with the candidate road segments in the road network, and the transition probability is introduced to construct a continuous driving path. Based on the improved longest common subsequence algorithm, trajectory similarity is calculated for trajectory data between different trucks to identify the cooperative driving behavior between the two trucks. A multi-vehicle cooperative driving graph is constructed based on graph theory modeling. Edge connections are constructed for vehicle pairs that meet the requirements of trajectory similarity and cooperative driving time. Connected components are extracted, and a preliminary cooperative driving vehicle set that meets the set conditions is identified. The preliminary cooperative driving vehicle set is then merged and judged to obtain the final cooperative driving vehicle set. Based on vehicle dynamics and aerodynamics models, the final cooperative driving vehicle set is modeled and analyzed to obtain carbon emission reduction assessment results, which include fuel savings and carbon emission reductions. The carbon emission reduction assessment results are obtained by modeling and analyzing the final cooperative driving vehicle set based on vehicle dynamics and aerodynamics models, specifically including: For each vehicle in the final cooperative driving vehicle set, the required traction force of the vehicle is calculated based on the trajectory data. The formula for calculating the required traction force of the vehicle is as follows: , For vehicle quality, For acceleration, For rolling resistance, For aerodynamic drag; The vehicle positional relationships and speed characteristics of each formation in the final coordinated driving vehicle set are used as aerodynamic correction factors to generate air resistance reduction coefficients for the lead vehicle and following vehicles, and the required traction force is dynamically corrected based on the air resistance reduction coefficients. Based on the required traction force and engine efficiency, the fuel consumption per unit time is estimated, and the corresponding carbon emissions are calculated in combination with the carbon emission factor. The fuel savings and carbon emission reductions of each vehicle in the final cooperative driving vehicle set within the cooperative driving range are statistically analyzed to obtain the carbon emission reduction assessment results of each cooperative vehicle in the final cooperative driving vehicle set. The identification information of the final cooperative driving vehicle set and the carbon emission reduction assessment results are structured and organized for use by the multi-dimensional classification statistics and scheduling system.

2. The truck platooning carbon emission reduction assessment method based on trajectory data according to claim 1, characterized in that, The preprocessing of the trajectory data to obtain preprocessed trajectory data specifically includes: The trajectory data is cleaned to obtain cleaned trajectory data, wherein the cleaning operation includes removing trajectory points with abnormal speed, missing timestamps, duplicate timestamps, or invalid coordinates; In the cleaned trajectory data, if the time interval between adjacent trajectory points is greater than a preset time threshold, the cleaned trajectory data is divided into different travel segments to generate a structured trajectory sequence. The structured trajectory sequence is numbered and the corresponding vehicle identifier and time information are retained to obtain the preprocessed trajectory data.

3. The truck platooning carbon emission reduction assessment method based on trajectory data according to claim 1, characterized in that, The road network map matching algorithm matches the trajectory points in the preprocessed trajectory data with candidate road segments in the road network, and introduces transition probabilities to construct continuous driving paths, specifically including: Using the geographic coordinates of each trajectory point in the preprocessed trajectory data as the center, a spatial buffer with a preset radius is constructed, and multiple road segments falling within the spatial buffer are retrieved as a candidate road segment set. For each candidate road segment, the matching probability is calculated based on the vertical distance from the trajectory point to the candidate road segment, and the transition probability between trajectory points is calculated based on the geographical distance between trajectory points and the shortest network distance between candidate road segments. The formula for calculating the matching probability is as follows: , This represents the perpendicular distance from a trajectory point to a road segment. The formula for calculating the transition probability is: where is the GPS positioning standard deviation. , For the shortest distance on the road network, This is the actual geographical distance. The attenuation coefficient is used; the actual geographical distance formula is: , For the Earth's radius, , Φ1 represents the difference in latitude and longitude between two points, where Φ1 is the latitude value of the first trajectory point in the adjacent trajectory point pair and Φ2 is the latitude value of the second trajectory point in the adjacent trajectory point pair. The optimal path matching sequence is determined based on the matching probability and transition probability to generate a continuous driving trajectory aligned with the road network.

4. The truck platooning carbon emission reduction assessment method based on trajectory data according to claim 1, characterized in that, The improved longest common subsequence algorithm calculates trajectory similarity between trajectory data of different trucks to identify cooperative driving behavior between the two trucks, specifically including: The trajectories of any two vehicles are uniformly sampled and time-aligned to construct the time interval of trajectory overlap; Based on the improved longest common subsequence algorithm, matching point pairs that meet the spatial distance and temporal difference thresholds are identified, and the spatial matching rate is calculated based on the number of matching point pairs. The directional difference between the matching point pairs is calculated to generate directional similarity. This directional similarity is then fused with the spatial matching rate to obtain an overall trajectory similarity score. The formula for calculating directional similarity is: , For the number of paired points, , The heading angle and the formula for calculating the overall trajectory similarity score are as follows: ,in, For spatial matching rate, is the weighting coefficient, where i is vehicle i, j is vehicle j, and |M| is the total number of paired points; If the trajectory similarity score is higher than the preset similarity threshold, and the continuous time period covered by the matching point pair is not less than the set collaboration duration threshold, then it is determined that the two vehicles are cooperating.

5. The truck platooning carbon emission reduction assessment method based on trajectory data according to claim 1, characterized in that, The method of constructing a multi-vehicle cooperative driving graph based on graph theory modeling involves connecting vehicle pairs that meet the requirements of trajectory similarity and cooperative driving duration with edges, extracting connected components, and identifying a preliminary set of cooperative driving vehicles that meet the set conditions. Specifically, this includes: All vehicle pairs that meet the criteria of trajectory similarity score higher than a preset threshold and whose time period covered by the matching point pair is not less than the collaboration duration threshold are constructed as edge connections in a graph structure, and an undirected graph is constructed with the corresponding vehicle as a node. In the undirected graph, all connected components are extracted using a depth-first traversal, and the connectivity density of each connected component is calculated. The formula for calculating the connectivity density is: , The actual number of edges within the component, The number of vehicles included in the component, Let be the density, representing the degree of vehicle cooperation, where |E c | represents the number of actual relationships between vehicles; Components with a connectivity density greater than a set density threshold are identified as valid cooperative formations, and the obtained valid cooperative formations are statistically analyzed to obtain the preliminary cooperative driving vehicle set.

6. The truck platooning carbon emission reduction assessment method based on trajectory data according to claim 1, characterized in that, The step of merging and judging the preliminary set of cooperative vehicles to obtain the final set of cooperative vehicles specifically includes: Based on each cooperative formation in the initial cooperative driving vehicle set, the vehicle member overlap ratio and time interval overlap ratio of any two components are calculated, where the vehicle member overlap ratio is defined as: , The number of intersections of the members. For a smaller group of vehicles; When the overlap ratio of vehicle members and the overlap ratio of time intervals both exceed the preset merging threshold, the corresponding cooperative formations are merged into a set of cooperative driving vehicles. All cooperative formation sets that meet the merging conditions are numbered to determine and output the final set of cooperative driving vehicles.

7. A truck platooning carbon emission reduction assessment system based on trajectory data, characterized in that, The truck platooning carbon emission reduction assessment system based on trajectory data includes: The trajectory acquisition and preprocessing module is used to acquire truck running trajectory data within the target area, preprocess the trajectory data, and obtain preprocessed trajectory data. The trajectory matching construction module is used to perform spatial location matching between trajectory points in the preprocessed trajectory data and candidate road segments in the road network based on the road network map matching algorithm, and to introduce transition probability to construct a continuous driving path. The trajectory similarity recognition module is used to calculate the trajectory similarity between different trucks based on the improved longest common subsequence algorithm, and to identify the cooperative driving behavior between the two trucks. The collaborative graph modeling and merging module is used to construct a multi-vehicle collaborative driving graph based on graph theory modeling. It constructs edge connections for vehicle pairs that meet the requirements of trajectory similarity and collaborative duration, extracts connected components, identifies a preliminary collaborative driving vehicle set that meets the set conditions, and merges the preliminary collaborative driving vehicle set to obtain the final collaborative driving vehicle set. The carbon emission reduction modeling module is used to model and analyze the final cooperative driving vehicle set based on vehicle dynamics and aerodynamics models to obtain carbon emission reduction assessment results, wherein the carbon emission reduction assessment results include fuel savings and carbon emission reductions. The result structuring module is used to structure and organize the identification information of the final cooperative driving vehicle set and the carbon emission reduction assessment results for use by the multi-dimensional classification statistics and scheduling system. The carbon emission reduction modeling module includes: The traction calculation submodule is used to calculate the required traction force for each vehicle in the final cooperative driving vehicle set based on trajectory data. The formula for calculating the required traction force for a vehicle is as follows: , For vehicle quality, For acceleration, For rolling resistance, For aerodynamic drag; The aerodynamic correction submodule is used to take the vehicle position relationship and speed characteristics of each formation in the final cooperative driving vehicle set as aerodynamic correction factors, generate the air resistance reduction coefficients of the lead vehicle and the following vehicles respectively, and dynamically correct the required traction force according to the air resistance reduction coefficients. The fuel consumption and emissions calculation submodule is used to estimate the fuel consumption per unit time based on the required traction force and engine efficiency, and to calculate the corresponding carbon emissions in combination with the carbon emission factor. The evaluation result generation submodule is used to calculate the fuel savings and carbon emission reductions of each vehicle in the final cooperative driving vehicle set within the cooperative driving range, and to obtain the carbon emission reduction evaluation results of each cooperative vehicle in the final cooperative driving vehicle set.

Images