Bus travel carbon emission reduction accounting method fused with Beidou spatio-temporal data

Abstract

The invention relates to the technical field of urban traffic management and carbon emission reduction accounting, and provides a bus travel carbon emission reduction accounting method fused with Beidou spatio-temporal data, which comprises the following steps: acquiring spatio-temporal trajectory data based on a Beidou satellite navigation system, and constructing a passenger-trajectory-station three-dimensional correlation model for passenger flow analysis; constructing a passenger transfer analysis system based on the passenger flow analysis result; constructing a dynamic'perception-prediction-optimization 'dynamic scheduling optimization system for passenger flow detail analysis on the basis of space-time trajectory data and a reinforcement learning algorithm; and constructing a mileage-frequency distribution model and a station efficiency evaluation model based on the spatio-temporal trajectory data, respectively generating a trajectory matching optimization strategy and a differentiated station optimization scheme, and realizing adaptive optimization of the public transportation system under dynamic demand change. Compared with a traditional static data dependence method, the method achieves the dynamic and fine-grained monitoring of the operation process of the public transportation system, and effectively improves the timeliness and integrity of data.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of urban traffic management and carbon emission reduction accounting, in particular to a bus travel carbon emission reduction accounting method fusing Beidou space-time data. BACKGROUND

[0002] With the increasingly serious problems of global warming and carbon emission increase, the international community has paid high attention to the adverse effects brought by it. According to statistics, the carbon emission in the field of transportation accounts for 12% of the total carbon emission in China. As a green and low-carbon transportation mode, bus travel plays an important role in urban traffic. Accurate accounting of the carbon emission reduction of bus travel can not only provide a scientific basis for urban traffic planning and policy making, but also encourage the public to choose public transportation and promote the sustainable development of urban traffic. At present, the existing accounting methods for the carbon emission reduction of bus travel include: accounting method based on traffic carbon emission model, accounting method based on vehicle operation data and passenger flow data, accounting method based on carbon inclusive system, and accounting method based on traffic flow theory.

[0003] However, the above methods have some limitations in practical application: traditional models mostly rely on static data, which is difficult to accurately reflect the real-time operation characteristics of the bus system. In addition, the lack of passenger trip chain information collection leads to deviation in carbon emission reduction accounting. At the same time, the lack of fine-grained analysis of the actual trajectory of the bus, the boarding and alighting behavior of the passengers and the spatio-temporal distribution makes the accounting granularity coarse, which is difficult to support fine-grained management and personalized incentive mechanism construction. SUMMARY

[0004] In view of this, the application provides a bus travel carbon emission reduction accounting method fusing Beidou space-time data, which utilizes the advantages of high-precision positioning, high-frequency space-time data acquisition and continuous stable transmission of the Beidou satellite navigation system, and realizes real-time acquisition of dynamic information such as the running track, speed change and stop position of a bus, and simultaneously combines with space-time point data of passengers getting on and off the bus to accurately restore the actual running state of the bus and the travel behavior of passengers. In the process of carbon emission accounting, the energy consumption and carbon emission per unit distance of the bus are dynamically calculated through trajectory data and energy consumption model calculation, and further based on the travel mode substitution theory, the actual carbon emission reduction brought by bus travel is accurately quantified, and the accounting accuracy is effectively improved. As a specific type of big data, the Beidou space-time data not only fuses attribute information such as position and time, but also shows complex characteristics such as multi-scale, multi-source heterogeneity and multi-dimension. The rich space-time characteristics provide a solid data foundation for fine description of the running state of the bus system and carbon emission reduction accounting. Not only can the dynamic change trajectories of the bus and passengers in the spatial and temporal dimensions be described, but also the dynamic evaluation, statistics and trend prediction of carbon emission reduction can be realized under different time scales (such as hours, days and weeks), which can meet diversified application requirements such as real-time detection and prediction. Through fine-grained and dynamic carbon emission reduction analysis, scientific and reliable decision-making basis can be provided for decision-making departments.

[0005] To achieve the above object, the application provides a bus travel carbon emission reduction accounting method fusing Beidou space-time data, which comprises the following steps: S1, based on high-precision space-time trajectory data provided by the Beidou satellite navigation system, a three-dimensional correlation model of “passenger-trace-station” is constructed, and passenger flow analysis is performed; S2, based on deep fusion of Beidou space-time data and OD passenger flow model, a city-level transfer analysis system is constructed, and passenger transfer analysis is performed; S3, based on real-time positioning data of Beidou and reinforcement learning algorithm, a dynamic “perception-prediction-optimization” dynamic scheduling optimization system is constructed, and passenger flow detail analysis is performed; S4, deep fusion of Beidou space-time data and GIS geographic information technology is performed, a dynamic passenger flow-emission collaborative detection platform is constructed, and through generation of a city all-day passenger flow heat map, passenger flow response characteristics in different scenarios are analyzed; S5, submodules such as travel mileage distribution analysis and number of stations analysis are integrated, multi-dimensional travel characteristics are quantitatively presented, an analysis system of “demand perception-route optimization-emission reduction verification” is constructed, and passenger travel mileage analysis is performed.

[0006] Optionally, the "passenger-track-station" three-dimensional correlation model. By analyzing the space-time data matching relationship between vehicle track data (latitude, longitude, speed, direction) and passenger card time stamp, the original positioning data is mapped to the actual vehicle road network topology, realizing high-precision calculation of vehicle driving distance. In the data unified to space-time correlation, the boarding station is determined according to the correspondence between the card time and the nearest track point, and the passenger alighting station probability distribution is calculated according to the inference model. A dynamic OD matrix is constructed, which is further used for the calculation of the basic index of carbon emission accounting-person-kilometer (PKM). In addition, the module integrates dynamic network graph technology to generate multi-dimensional visual analysis graphs including time period passenger flow trend, monthly travel calendar heat map, line load distribution, etc., which intuitively shows the change rule of passenger flow at different time scales, and reveals the load distribution characteristics of line operation, providing data basis for subsequent data analysis and operation optimization.

[0007] Optionally, the city-level transfer analysis system. By analyzing the vehicle arrival time stamp in Beidou space-time data and passenger card records, a transfer relationship topology network is established by combining graph calculation: taking station as node and transfer relationship as edge to establish a weighted graph G=(V, E, A), wherein V represents a set of N bus stations, E represents the connection relationship between stations, and A is an adjacency matrix representing the transfer frequency between stations. The transfer network optimization algorithm is used to calculate and analyze the city bus transfer relationship graph, dynamically identify the key transfer area, and present the spatial distribution of transfer bottlenecks through the visual interface to assist in optimizing the transfer path design. The transfer time penalty function is introduced to further reduce redundant transfer and improve the proportion of direct travel, realizing the dual emission reduction benefits: on the one hand, reducing the idling emission of vehicles due to waiting for transfer passengers; on the other hand, reducing the probability of passengers choosing other high-carbon transportation modes due to inconvenience, thereby effectively promoting the construction of green and low-carbon travel system.

[0008] Optionally, the dynamic "perception-prediction-optimization" dynamic scheduling optimization system. According to the card data and track positioning, a space-time two-dimensional passenger flow matrix A(s, t) is formed, wherein s represents the station, t is the time period, and the matrix element represents the passenger flow difference of the station in a certain time period. Through the matrix, the system can detect the line passenger flow in real time and timely find the vehicle overload or empty vehicle problem.

[0009] In order to capture the space-time nonlinear characteristics of passenger flow, a space-time graph convolution network is used to model the space-time relationship between stations: taking station as node and passenger flow transfer probability between stations as edge weight to construct a dynamic graph G(V, E t), while integrating BeiDou trajectory features, station attributes, and other information to fine-tune the prediction of passenger flow in a certain period in the future. Based on passenger flow prediction, a multi-objective reinforcement learning framework is designed for intelligent scheduling optimization. The current period's station passenger flow vector is taken as the system state, and the vehicle schedule adjustment is taken as the action. The system considers both passenger flow stability and vehicle utilization rate improvement to construct a reward function and train an intelligent scheduling strategy, thereby achieving precise capacity allocation, reducing resource waste, and improving overall transportation efficiency and unit carbon emission reduction benefits.

[0010] Optionally, the dynamic passenger flow-emission collaborative detection platform. Based on BeiDou positioning data, the system captures vehicle trajectory and speed features in real time, visualizes passenger flow density through color gradients, forms a high-spatial-temporal resolution dynamic evolution map of urban public transportation passenger flow, presents the spatial distribution of global passenger flow within 24 hours in time sequence animation, and calculates the spatial distribution of passenger flow and carbon emissions using a heat map generation algorithm. In addition, this module integrates meteorological data and event information to construct a passenger flow anomaly regression model, quantifying the impact of external factors on passenger flow fluctuations, distinguishing between regular passenger flow and event-driven abnormal fluctuations, and generating targeted scheduling plans. Breaking through the limitations of traditional heat maps, which only show passenger flow, the system uses a "BeiDou trajectory-passenger capacity-emission factor" ternary coupling model to compare and analyze passenger flow response characteristics in different scenarios, providing multi-dimensional optimization solutions for traffic scheduling, and achieving an upgrade from "passenger flow diversion" to "carbon flow optimization" in traffic control, providing a spatio-temporal decision base for urban low-carbon transportation management.

[0011] Optionally, the "demand-aware-route optimization-emission verification" analysis system includes: Trip mileage analysis submodule: Based on high-precision BeiDou trajectory data, the system calibrates the actual passenger travel distance, establishes a mileage-frequency distribution model, identifies high-frequency travel intervals and long-distance travel demand, and generates trajectory matching optimization strategies, such as backtracking BeiDou trajectories for detours exceeding the designed mileage to reduce invalid mileage, increasing train frequencies in high-frequency intervals during peak hours, and running express trains through large stations during off-peak hours, and warning for empty driving.

[0012] Number of stops analysis submodule: Based on average stop time, number of stops throughout the day, and passenger flow at station stops, the system constructs a station efficiency evaluation model. Through this evaluation model, the system can accurately locate stations with low operational efficiency. For these stations, the system generates differentiated station optimization solutions, such as simplifying stops and setting up fast jump sections, to improve overall line average operating speed and travel efficiency.

[0013] Through continuous data feedback and model updates, the system can achieve rapid iteration of the supply side, supporting the adaptive optimization of the public transportation system under dynamic demand changes. Ultimately, it forms a virtuous cycle of "low-carbon service-demand growth-efficiency improvement," providing data support and decision-making basis for sustainable urban transportation development.

[0014] In addition, the application provides a bus travel carbon emission reduction amount accounting device fusing Beidou space-time data, comprising: a passenger flow statistical analysis module, which constructs a "passenger-track-station" three-dimensional correlation model based on high-precision space-time trajectory data provided by the Beidou satellite navigation system; a transfer analysis module, which constructs a city-level transfer analysis system based on deep fusion of Beidou space-time data and an OD passenger flow model; a line passenger flow detail module, which constructs a dynamic "perception-prediction-optimization" dynamic scheduling optimization system based on Beidou real-time positioning data and a reinforcement learning algorithm; a city all-day thermal map module, which constructs a dynamic passenger flow-emission collaborative detection platform by deeply fusing Beidou space-time data and GIS geographic information technology; and a travel mileage analysis module, which integrates submodules such as travel mileage distribution analysis and the number of stations analysis, quantitatively presents multi-dimensional travel characteristics, and constructs an analysis system of "demand perception-route optimization-emission reduction verification".

[0015] Compared with the prior art, the application has the following beneficial effects: Compared with a traditional static data dependent method, the application realizes dynamic and fine-grained monitoring of the operation process of a bus system, effectively improving the timeliness and integrity of data.

[0016] The application dynamically calculates the energy consumption and carbon emission per unit distance of a bus vehicle by trajectory data and energy consumption model calculation, further accurately quantifies the actual carbon emission reduction amount caused by bus travel based on a travel mode substitution theory, and effectively improves the accounting accuracy.

[0017] Based on Beidou space-time data, bus travel carbon emission reduction amount accounting can not only depict the dynamic change trajectories of bus vehicles and passengers in the spatial and temporal dimensions, but also support dynamic evaluation, statistics and trend prediction of carbon emission reduction amount at different time scales, and can meet diversified application requirements such as real-time detection and prediction. BRIEF DESCRIPTION OF DRAWINGS

[0018] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are merely illustrative and are not considered to be limiting in any way. In the drawings: Figure 1 The application provides a whole flow schematic diagram of a bus travel carbon emission reduction amount accounting method fusing Beidou space-time data; Figure 2 The application provides a specific embodiment flow schematic diagram of a bus travel carbon emission reduction amount accounting method fusing Beidou space-time data; Figure 3 The application provides a framework structure diagram of a bus travel carbon emission reduction amount accounting device fusing Beidou space-time data. DETAILED DESCRIPTION

[0019] Exemplary embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the drawings, it is understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided so that the present disclosure can be more thoroughly understood and so that the scope of the present disclosure can be conveyed to those skilled in the art. It should be noted that the embodiments in the present disclosure and the features in the embodiments can be combined with each other without conflict. The present disclosure will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

[0020] The present embodiment proposes a bus travel carbon emission reduction accounting method fusing Beidou space-time data, as shown in Figure 1 The method comprises the following steps: S1, based on the high-precision space-time trajectory data provided by the Beidou satellite navigation system, a "passenger-trace-station" three-dimensional correlation model is constructed, and passenger flow analysis is performed; Further, in the embodiment of the present disclosure, by analyzing the space-time data matching relationship between vehicle trajectory data (latitude, longitude, speed, direction) and passenger card time stamp, the original positioning data is mapped to the actual vehicle road network topology structure, and high-precision calculation of vehicle driving distance is realized. In the data unified to space-time correlation, the card time corresponds to the nearest trace point to determine the boarding station, and the passenger alighting station probability distribution is calculated according to the reasoning model. A dynamic OD matrix is constructed, which is further used for the calculation of the carbon emission accounting basic index-PKM (Passenger Kilometer). In addition, the module integrates dynamic network graph technology to generate multi-dimensional visual analysis graphs including time period passenger flow trend, monthly travel calendar heat map, line load distribution, etc. The change rule of passenger flow on different time scales is intuitively displayed, and the line operation load distribution characteristics are revealed, which provides data basis for subsequent data analysis and operation optimization.

[0021] S2, based on the deep fusion of Beidou space-time data and OD passenger flow model, a city-level transfer analysis system is constructed, and passenger transfer analysis is performed; Further, in the embodiments of the present application, by analyzing the vehicle arrival time stamp in Beidou space-time data and the passenger card swiping record, a transfer relationship topology network is established by combining graph calculation: taking the station as the node and the transfer relationship as the edge to establish a weighted graph G=(V, E, A), wherein V represents the set of N bus stations, E represents the connection relationship between stations, and A is an adjacency matrix representing the transfer frequency between stations. The transfer network optimization algorithm is used to calculate and analyze the city bus transfer relationship graph, dynamically identify the key transfer area, and present the spatial distribution of the transfer bottleneck through the visual interface to assist in optimizing the transfer path design. The transfer time penalty function is introduced to further reduce redundant transfer and improve the direct travel proportion, and the overall dual emission reduction benefit is achieved: on the one hand, the idling emission of vehicles due to waiting for transfer passengers is reduced; on the other hand, the probability of passengers choosing other high-carbon transportation modes due to inconvenience of transfer is reduced, thereby effectively promoting the construction of green and low-carbon travel system.

[0022] S3, based on Beidou real-time positioning data and reinforcement learning algorithm, a dynamic "perception-prediction-optimization" dynamic scheduling optimization system is constructed to analyze passenger flow details; Further, in the embodiments of the present application, according to the card swiping data and trajectory positioning, a space-time two-dimensional passenger flow matrix A(s, t) is formed, wherein s represents a station, t is a time period, and the matrix element represents the passenger flow difference of the station in a certain time period. Through the matrix, the system can detect the line passenger flow dynamics in real time and timely find the vehicle overload or empty vehicle problem. In order to capture the space-time nonlinear characteristics of passenger flow, a space-time graph convolution network is used to model the space-time relationship between stations: taking the station as the node and the passenger flow transfer probability between the stations as the edge weight to construct a dynamic graph G(V, E t ), and simultaneously fusing Beidou trajectory features, station attributes and other information to finely predict the passenger flow in a certain period in the future. On the basis of passenger flow prediction, a multi-objective reinforcement learning framework is designed for intelligent scheduling optimization. Taking the station passenger flow vector in the current period as the system state and the vehicle schedule adjustment as the action, two indexes of passenger flow stability and vehicle utilization rate improvement are considered to construct a reward function, and an intelligent scheduling strategy is trained, so as to realize precise capacity allocation, reduce resource waste, and improve overall transportation efficiency and unit carbon emission reduction benefit.

[0023] S4, deeply integrate Beidou space-time data and GIS geographic information technology to construct a dynamic passenger flow-emission collaborative detection platform, generate a city all-day passenger flow heat map, and analyze the passenger flow response characteristics in different scenarios; Further, in the embodiments of the present application, based on the real-time capture of vehicle trajectory and speed characteristics based on Beidou positioning data, the passenger flow density is visualized through color gradient to form a high spatiotemporal resolution dynamic evolution map of urban public transport passenger flow, and the spatial distribution law of global passenger flow within 24 hours is presented in time sequence animation, and the spatial distribution of passenger flow and carbon emission is calculated by using a heat map generation algorithm. In addition, this module fuses meteorological data (such as rainfall, air temperature change) and event information (such as large-scale activities), constructs a passenger flow abnormality regression model, quantifies the influence of external factors on passenger flow fluctuation, distinguishes between regular passenger flow and event-driven abnormal fluctuation, and generates a targeted scheduling plan. Breaking through the limitation of traditional heat map which only displays passenger flow, through the ternary coupling model of “Beidou trajectory-carrying capacity-emission factor”, the passenger flow response characteristics in different scenarios are compared and analyzed, a multi-dimensional optimization scheme is provided for traffic scheduling, the traffic control is upgraded from “passenger flow diversion” to “carbon flow optimization”, and a spatiotemporal decision base is provided for urban low-carbon traffic governance.

[0024] S5, an integrated travel mileage distribution analysis and a boarding station number analysis sub-module are integrated, multi-dimensional travel characteristics are quantitatively presented, and an analysis system of “demand perception-route optimization-emission reduction verification” is constructed, and the travel mileage of passengers is analyzed.

[0025] Further, in the embodiments of the present application, the module comprises: The travel mileage analysis sub-module: based on the Beidou high-precision trajectory data, the actual passenger travel distance is calibrated, a mileage-frequency distribution model is established, high-frequency travel intervals and long-distance travel demand are identified, and then a trajectory matching optimization strategy is generated, such as Beidou trajectory backtracking for detour sections exceeding the designed mileage to reduce invalid mileage, increasing train frequency in high-frequency intervals during peak hours, and opening express trains through large stations during off-peak hours, and empty driving early warning.

[0026] The boarding station number analysis sub-module: according to the average stop time, the number of train stops throughout the day, and the passenger flow of train stations, a station efficiency evaluation model is constructed. Through the evaluation model, the system can accurately locate the stations with low operating efficiency. Differentiated station optimization schemes are generated for these stations, such as simplifying stops and setting up fast jump sections, so as to improve the average running speed and travel efficiency of the line as a whole.

[0027] Through continuous data feedback and model updating, rapid iteration of the supply side can be realized, supporting the adaptive optimization of the public transport system under dynamic changes in demand. Ultimately, a virtuous cycle of “low-carbon service-demand growth-efficiency improvement” is formed, providing data support and decision basis for sustainable development of urban transportation.

[0028] The specific implementation of S1-S5 will be described below in conjunction with Figure 2

[0029] ​S201, data cleaning and verification: passenger flow analysis is based on two categories of information, static data and dynamic data. The static data includes: vehicle data, line data, line station relationship, station latitude and longitude positioning data and station track distance; dynamic data includes: passenger card data and code data. First, the static and dynamic data are tested in multiple dimensions to ensure that the key fields have no null values and no errors. For example, the line and station relationship must be complete and corresponding, and the station latitude and longitude must match the Beidou high-precision positioning information. Among them, the card or code data needs to be checked for uniqueness by a hash function to ensure the independence and consistency of each passenger record. If there are missing or abnormal core fields such as time, line, and station in the data, they will be supplemented according to the preset rules, and dirty data that cannot be repaired will be completely removed to ensure the overall data quality.

[0030] S202, matching route and station information: after the basic verification, the data enters the card line and station matching stage, accurately identifying the boarding station and route of each passenger. Based on the card time and vehicle trajectory data, the vehicle arrival time is calculated by the shortest path algorithm, and the nearest station is matched as the boarding point within the set time difference range. If there are multiple candidate stations, the nearest station is selected by the Beidou positioning assisted trajectory fitting algorithm.

[0031] For the above process that cannot complete the station line matching, the supplement rules are formulated according to the data integrity: for records with serious missing and cannot be corrected, directly discarded to avoid introducing bias; for passengers with historical travel data, based on their past travel characteristics, assign the commonly used station; if the passenger's historical travel records are insufficient to support the inference, and the card line exists but the station is missing, the reasonable boarding interval is calculated according to the average one-way travel distance of urban buses combined with the average speed of the line to assign the value; set the error threshold control, if the inference result is seriously inconsistent with the actual trajectory or time logic, it is considered as dirty data and discarded.

[0032] S203, identify transfer and return information: for the second card record of the same card / code in a short time, the spatial clustering algorithm of OD matrix estimation and Gaussian kernel density estimation is used to accurately judge the passenger transfer or return behavior. Taking the card number as the unique identifier, the continuous card record of the same passenger in a certain time window is tracked; combined with the card time interval and the spatial distance of the card location: if the two card time intervals are short and the spatial distance is close, combined with the historical transfer probability analysis of the OD matrix, when a certain probability threshold is met, it is inferred as a transfer behavior; on the contrary, if the two card time intervals are long and the spatial positions are dispersed, combined with the OD statistical distribution of the city bus network, when the distribution characteristics of the return behavior are met, it is inferred as a return behavior; for the transfer between different stations, the vehicle positioning data corresponding to the passenger card is used, and the nearest line or related line station is located in the public transportation network through spatial proximity search to determine the transfer location. Based on the trajectory position matching method, the actual travel path of the passenger is more truly restored.

[0033] S204, build travel habit model jinx: for high-frequency travel patterns with features such as transfer and return, a personalized travel chain inference based on a spatiotemporal probability model is constructed through spatiotemporal feature extraction. Adopting rule-based inference methods such as Naive Bayes classification, the highest probability travel route and station combination is matched for incomplete travel data. In model training, rule-based inference and machine learning algorithms are used to automatically identify travel chain time period rules, workday and rest day patterns, and set confidence threshold values to confirm high-confidence inference results and include them in individual travel habits to continuously optimize model inference accuracy and generalization ability.

[0034] S205, single ride record processing: for passengers with only single card record in a selected time and lack of complete travel chain, the single trip average travel distance is assigned according to the city bus system statistics to ensure full data coverage and avoid overall statistical bias.

[0035] S206, data verification and accuracy evaluation: to test the reasoning and matching accuracy, random sampling is performed from the travel chains that have completed reasoning, and compared with the measured GPS and questionnaire data to calculate matching accuracy, OD recognition accuracy and other indicators. If the prediction is lower than the preset, the matching parameters are adjusted or additional data sources are added.

[0036] S207, statistical summary: for individual data that cannot be completely inferred after the previous steps, the average travel distance of the passenger's personal history or the average distance of the city's overall public transport is reasonably assigned to ensure that the total data has a unified travel distance indicator. The energy consumption of the assigned data is calculated using the trajectory energy consumption model to estimate carbon emissions. Finally, classified statistics and analysis are performed as needed to extract core indicators such as per capita travel distance, single-trip carbon emission reduction contribution, and other core indicators in different time periods, lines, and regions. Ultimately, standardized and quantifiable low-carbon travel accounting data results are formed. This process ensures that even in the case of incomplete data, scientific and systematic statistical results of public transport carbon emission reduction can still be provided, providing strong support for subsequent urban transportation low-carbon development policy making.

[0037] A bus travel carbon emission reduction accounting method fusing Beidou space-time data, Figure 3 A structure block diagram of a bus travel carbon emission reduction accounting device fusing Beidou space-time data is shown, which is not convenient for illustration, and only parts related to the embodiments of the present application are shown.

[0038] Reference Figure 3 The device 300 comprises: A passenger flow statistical analysis module 301 constructs a "passenger-travel-station" three-dimensional correlation model based on high-precision space-time trajectory data provided by the Beidou satellite navigation system; A transfer analysis module 302 constructs a city-level transfer analysis system based on the deep fusion of Beidou space-time data and OD passenger flow models; A line passenger flow detail module 303 constructs a dynamic "perception-prediction-optimization" dynamic scheduling optimization system based on Beidou real-time positioning data and reinforcement learning algorithms; A city all-day thermal map module 304 constructs a dynamic passenger flow-emission collaborative detection platform based on the deep fusion of Beidou space-time data and GIS geographic information technology; A travel distance analysis module 305 integrates sub-modules such as travel distance distribution analysis and number of stations analysis to quantitatively present multi-dimensional travel characteristics and construct an "demand perception-route optimization-emission reduction verification" analysis system.

[0039] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application and not to limit them, although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that the specific embodiments of the present application can be modified or replaced, without departing from the spirit and scope of the present application. Any modification or equivalent replacement thereof should be covered within the protection scope of the claims of the present application.

Claims

1. A method for calculating carbon emission reductions from public transportation trips by integrating BeiDou spatiotemporal data, characterized in that, include: Based on spatiotemporal trajectory data obtained from the BeiDou Navigation Satellite System, a three-dimensional correlation model of "passenger-trajectory-station" is constructed for passenger flow analysis; A passenger transfer analysis system was constructed based on passenger flow analysis results; Based on the aforementioned spatiotemporal trajectory data and reinforcement learning algorithms, a dynamic scheduling optimization system of "perception-prediction-optimization" is constructed to analyze passenger flow details. Based on the spatiotemporal trajectory data, a mileage-frequency distribution model and a station efficiency evaluation model are constructed, generating trajectory matching optimization strategies and differentiated station optimization schemes respectively, so as to realize the adaptive optimization of the public transportation system under dynamic changes in demand.

2. The method according to claim 1, characterized in that, The process of constructing a three-dimensional correlation model of "passengers-trajectory-stations" for passenger flow analysis includes: Based on the spatiotemporal trajectory data analysis, the spatiotemporal data matching relationship between vehicle trajectory data and passenger card swipe timestamps is analyzed, and the original positioning data is mapped to the actual vehicle-road network topology to calculate the vehicle travel distance. The boarding station is determined by the nearest trajectory point corresponding to the card swipe time, and the probability distribution of passenger alighting stations is inferred according to the inference model. A dynamic OD matrix is ​​constructed to calculate the basic indicator for carbon emission accounting - passenger-kilometers. Dynamic network graph technology is used to generate a multi-dimensional visualization analysis map including time-period passenger flow trends, monthly travel calendar heat maps, and line load distribution to obtain the changing patterns of passenger flow at different time scales and reveal the characteristics of line operation load distribution.

3. The method according to claim 1, characterized in that, The passenger flow analysis is based on static and dynamic data. The static data includes vehicle data, route data, route-station relationships, station latitude and longitude positioning data, and trajectory distances between stations. The dynamic data includes passenger card swipe data and QR code swipe data. In the process of passenger flow analysis, firstly, multi-dimensional checks are performed on the static and dynamic data. Among them, card swipe or QR code swipe data needs to be uniquely verified by a hash function. If there are missing or abnormal core fields such as time, route, and station in the data, they are supplemented according to preset rules, and dirty data that cannot be repaired is removed.

4. The method according to claim 1, characterized in that, During the passenger flow analysis, for incomplete travel data, a travel habit model is constructed: for high-frequency travel patterns with features such as transfers and return trips, a personalized travel chain inference based on a spatiotemporal probability model is constructed through spatiotemporal feature extraction. Naive Bayes classification is used to match the incomplete travel data with the highest probability travel routes and station combinations. In model training, rule inference and machine learning algorithms are used to automatically identify the time period patterns and weekday and weekend patterns of travel chains, and a confidence threshold is set to confirm high-confidence inference results and incorporate them into individual travel habits, continuously optimizing the model's inference accuracy and generalization ability.

5. The method according to claim 2, characterized in that, The process of determining the boarding station based on the nearest trajectory point corresponding to the card swipe time includes: Based on card swiping time and vehicle trajectory data, the vehicle arrival time is estimated using the shortest path algorithm, and the nearest station is matched as the boarding point within a set time difference range; if there are multiple candidate stations, the station with the closest geographical location is selected by the trajectory fitting algorithm assisted by Beidou positioning. For cases where station and route matching cannot be completed, supplementary rules are formulated based on data integrity: records that are severely missing and cannot be corrected are directly discarded; for passengers with historical travel data, frequently used stations are assigned based on past travel characteristics; if the passenger's historical travel records are insufficient to support the inference, and the card-swiping route exists but stations are missing, a reasonable boarding interval is calculated based on the average one-way travel distance of urban public transport and the average speed of the route, and then assigned a station; an error threshold is set to control the data, and if the inference result does not match the actual trajectory or time logic, it is considered dirty data and discarded.

6. The method according to claim 1, characterized in that, The process of constructing a passenger transfer analysis system based on passenger flow analysis results includes: By analyzing vehicle arrival timestamps and passenger card swipe records from BeiDou spatiotemporal data, a transfer relationship topology network is established using graph computation. Specifically, this includes: constructing a weighted graph G=(V, E, A) with stations as nodes and transfer relationships as edges, where V represents the set of N bus stations, E represents the connection relationship between stations, and A is an adjacency matrix representing the transfer frequency between stations; using a transfer network optimization algorithm to calculate and analyze the urban public transport transfer relationship graph, identifying key transfer areas, and presenting the spatial distribution of transfer bottlenecks through a visualization interface to assist in optimizing transfer route design, while also introducing a transfer time penalty function.

7. The method according to claim 6, characterized in that, The process of calculating and analyzing urban public transport transfer relationship maps using transfer network optimization algorithms includes: For repeated card swipes of the same card / code within a short period, spatial clustering algorithms combining OD matrix estimation and Gaussian kernel density estimation are used to determine passenger transfer or return travel behavior. Card numbers are used as unique identifiers to track consecutive card swipes by the same passenger within a certain time window. If the time interval between two swipes is short and the spatial distance is close, historical transfer probability analysis of the OD matrix is ​​used; when a certain probability threshold is met, it is inferred to be a transfer. Conversely, if the time interval between two swipes is long and the spatial locations are dispersed, the statistical distribution of OD in the urban public transport network is used; when it matches the distribution characteristics of return travel behavior, it is inferred to be a return travel behavior. For transfers at different stations, the vehicle location data corresponding to the passenger's card swipe is used to locate the nearest station on the same or related route in the public transport network through spatial proximity search to determine the transfer location.

8. The method according to claim 1, characterized in that, The process of constructing a dynamic scheduling optimization system based on "perception-prediction-optimization" for passenger flow detail analysis includes: Based on the spatiotemporal trajectory data, passenger card swiping data and trajectory positioning are obtained to form a spatiotemporal two-dimensional passenger flow matrix A(s,t), where s represents the station, t is the time period, and the matrix elements represent the passenger flow difference of the station within a certain time period. The dynamics of passenger flow on the line are detected through the two-dimensional passenger flow matrix to discover vehicle overload or empty vehicle problems. Spatiotemporal graph convolutional networks are used to model the spatiotemporal relationships between stations. Specifically, this involves constructing a dynamic graph G(V, E) with stations as nodes and passenger flow transfer probabilities between stations as edge weights. t Simultaneously, it integrates BeiDou trajectory features and station attributes to predict passenger flow in future periods. Based on passenger flow prediction, a multi-objective reinforcement learning framework is designed to optimize intelligent scheduling. The current station passenger flow vector is used as the system state, and vehicle schedule adjustments are used as actions. Considering passenger flow stability and vehicle utilization improvement indicators, a reward function is constructed to train the intelligent scheduling strategy.

9. The method according to claim 1, characterized in that, The generation process of the trajectory matching optimization strategy includes: Based on the spatiotemporal trajectory data, the actual travel distance of passengers is calibrated, the mileage-frequency distribution model is constructed, and a trajectory matching optimization strategy is generated by identifying high-frequency travel intervals and long-distance travel needs.

10. The method according to claim 1, characterized in that, The process of generating the differentiated site optimization scheme includes: Based on average stop time, number of stops per day, and passenger flow data at each station, a station efficiency evaluation model is constructed. This model is used to identify stations with low operational efficiency, and differentiated station optimization plans are generated for these stations.

Images