Road carbon emission built environment influence assessment method and system based on multi-source trajectory

Abstract

The invention discloses a road carbon emission built environment influence assessment method and system based on a multi-source trajectory. The method comprises the steps of obtaining and preprocessing multi-source vehicle trajectory data, urban road network data, street view data and interest point data; traffic carbon emission of different pollutants on the road scale is calculated, multi-dimensional representation of the urban built environment is carried out, and built environment indexes on the road scale are calculated; an evaluation model of the influence of the urban built environment on the road carbon emission is constructed, training and prediction are carried out, an importance sequence of built environment indexes is obtained, and correlation analysis is carried out on the road carbon emission and the urban built environment; and evaluating the influence of the road carbon emission construction environment. According to the method, the urban road carbon emission is quantified and evaluated in combination with the multi-source vehicle trajectory data and various built-up environment indexes, the maximum built-up environment index for the road carbon emission is found, the influence degree and the action mechanism of the maximum built-up environment index are researched, and the method has good benefits in the fields of traffic carbon emission reduction, urban planning and the like.

Description

Technical Field

[0001] The present invention relates to the field of urban planning, and in particular to a method and system for evaluating the environmental impact of road carbon emissions based on multi-source trajectories. Background Art

[0002] Global warming caused by greenhouse gas emissions is seriously threatening human survival and sustainable development, and is one of the major global challenges currently facing mankind. As proposed in the 2021 "China's Policies and Actions on Climate Change", China should promote the coordinated and integrated planning of pollution reduction and carbon reduction, and embark on a path of greenhouse gas emission reduction that suits national conditions. The acceleration of urbanization and the growth of transportation demand have made transportation carbon emissions one of the main causes of global climate change and environmental pollution, and road transportation carbon emissions account for about three-quarters of transportation carbon emissions and are an important source of greenhouse gas emissions. Road transportation carbon emissions refer to the emissions of greenhouse gases such as carbon dioxide generated by road transportation.

[0003] The built environment is composed of various buildings and places that are artificially constructed and transformed. It is a combination of land use patterns, transportation systems, and a series of elements related to urban design that can affect residents' activities and behaviors, and has an important impact on people's quality of life, health, social interaction, etc. The operation of road traffic is significantly related to the built environment. The urban built environment is highly artificial and complex, and its various functional areas need to be connected through the transportation system. Therefore, reasonable urban built environment planning can reduce traffic congestion, improve travel conditions, reduce road traffic carbon emissions, and reduce pollution.

[0004] The emergence of big data such as vehicle trajectory data and street view data has provided direction for the quantification of carbon emissions and the characterization of the built environment. In this context, it is of great significance to accurately assess road traffic carbon emissions and analyze the impact of the built environment on road traffic carbon emissions. It can make new contributions to the assessment of road traffic carbon emissions and the analysis of the impact of the urban built environment on road traffic carbon emissions, provide new directions for urban construction and planning, and promote the construction of low-carbon cities and the development of ecological civilization. Summary of the invention

[0005] The main purpose of the present invention is to overcome the shortcomings and deficiencies of the prior art and to provide a method and system for assessing the environmental impact of road carbon emissions based on multi-source trajectories.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions: In one aspect of the present invention, a method for assessing the environmental impact of road carbon emissions based on multi-source trajectories is provided, comprising the following steps: Obtain multi-source vehicle trajectory data, urban road network data, street view data, and point of interest data and perform data integration, cleaning, denoising, matching, conversion, and correction; Calculate traffic carbon emissions of different pollutants at the road scale based on multi-source vehicle trajectory data; Based on urban road network data, street view image data and point of interest data, multi-dimensional characterization of urban built environment is carried out, and built environment indicators at the road scale are calculated; Construct an assessment model of the impact of urban built environment on road carbon emissions and use built environment indicators and road carbon emissions data for training and prediction to obtain the importance ranking of built environment indicators; Conduct correlation analysis between road carbon emissions and urban built environment; The built environmental impact of road carbon emissions is evaluated based on the results of importance ranking and correlation analysis of built environmental indicators.

[0007] As a preferred technical solution, the acquisition of multi-source vehicle trajectory data, urban road network data, street view data, and point of interest data and the data integration, cleaning, denoising, matching, conversion, and correction are specifically as follows: Obtain street view data, point of interest data, road network data, and multi-source vehicle trajectory data, and integrate the data to align various types of data according to a unified spatial coordinate system; Clean and process the integrated data, including: deleting duplicate data, filling missing values, and processing abnormal data; performing noise filtering and trajectory matching on street view data and multi-source vehicle trajectory data; The processed road network data is converted into a road network model. The roads are abstracted into line segments and these line segments are connected using a topological structure to represent the relationship between roads. The road network is topologically corrected to check whether there are closed loops, identify intersection nodes, and correct the connection relationship between road segments.

[0008] As a preferred technical solution, the traffic carbon emissions of different pollutants on the road scale are calculated based on multi-source vehicle trajectory data, specifically: Calculate the time interval between consecutive points in different types of vehicle trajectory data, compare it with the set threshold, and filter out trajectory points with a time interval lower than the set threshold; Calculate the instantaneous speed and acceleration of the vehicle at the trajectory point according to the time interval and distance between each point of the vehicle trajectory data, and filter out the drift points and abnormal points of speed and acceleration in the vehicle trajectory points, where the abnormal points are trajectory points whose speed or acceleration is greater than a set threshold; According to the instantaneous speed, acceleration and fuel type of each vehicle GPS track point after filtering, the carbon emissions of each vehicle GPS track point are calculated using micro emission model, emission factor method, life cycle analysis, macro emission model or traffic flow model, specifically: Place the vehicle at the GPS point p The emission at E p , the instantaneous velocity and acceleration are recorded as S p and a p , E p The calculation formula is: E p = f 1 + f 2 S p + f 3 S p 2 + f 4 a p + f 5 a p 2 + f 6 S p a p ; in, f 1 arrive f 6 Determined by the vehicle fuel type and the type of pollutants calculated. The vehicle fuel type is determined by the type of vehicle; Use the nearest neighbor algorithm, buffer construction, inverse distance weighted or cluster analysis method to assign each point to an urban road based on the urban road data, find the road closest to the current point, add up the carbon emissions of all points assigned to each road section, and get the emissions of the four pollutants on the road section. Perform principal component analysis on the same pollutant data of different types of vehicles to get the emissions of each pollutant.

[0009] As a preferred technical solution, the urban built environment is characterized based on street view image data, specifically: The urban built environment is characterized by using street view image data, and the semantic segmentation model is used to perform semantic segmentation on the street view images. The street landscape elements such as buildings, roads, vehicles, sidewalks, trees, sky, and pedestrians in the street view pictures are extracted, and the street landscape index from the street view perspective is constructed. The proportion of various street landscape elements in the street view images is calculated, and the street view built environment indicators such as sky openness, street enclosure, green view rate, building enclosure, crowd concentration, grassland proportion, tree proportion, plant proportion, sign density index, and cycling suitability index are obtained. Finally, the street view sampling points are spatially connected with the road network data, and the data of all street view sampling points on the road section are averaged to obtain the street landscape index at the road scale.

[0010] As a preferred technical solution, the urban built environment is characterized based on urban road network data, specifically: Road network data are used to construct road-related built environment indicators such as road linearity index, betweenness centrality, eigenvector centrality, closeness centrality, degree centrality, and clustering coefficient. The road linearity index is obtained by the ratio of the distance from the starting point to the end point of the road section to the distance, and the betweenness centrality, eigenvector centrality, closeness centrality, degree centrality, and clustering coefficient are obtained by taking the average of the indicators of the starting point and the end point of the road section.

[0011] As a preferred technical solution, the urban built environment is characterized based on the point of interest data, specifically: Use point of interest data to construct urban built environment indicators related to POIs, including land use richness, Shannon diversity index, Simpson diversity index, and proportion of road land use types; The road network data is analyzed by buffer zone analysis, and the 100-meter buffer zone of the road is extracted as the road space. Secondly, the road space is spatially connected with the POI point data, and the urban built environment index of each road section is calculated using the POI point data in the road space of the road section. Among them, when calculating land use richness, Shannon diversity index, and Simpson diversity index, POIs are divided into different subcategories; when calculating the proportion of road land use types, POIs are divided into four major categories: commercial, residential, transportation, and cultural, and the proportion of POI points in each category is calculated as the proportion of land use types on the road section.

[0012] As a preferred technical solution, the evaluation model of the impact of urban built environment on road carbon emissions is constructed and built environment indicators and road carbon emissions data are used for training and prediction to obtain the importance ranking of built environment indicators, specifically: Geographically weighted regression, random forest model, geographically weighted-random forest model and graph convolutional neural network are used for modeling. All streets are divided into training samples and verification samples according to a certain ratio; the training sample ratio is 75%; the above-mentioned built environment indicators and road carbon emissions are input into geographically weighted regression, random forest model, geographically weighted-random forest model and graph convolutional neural network, and iterative training is carried out to find the optimal parameters, generate the best fitting model, and rank the importance of each built environment indicator.

[0013] As a preferred technical solution, the correlation analysis between road carbon emissions and urban built environment is specifically as follows: Use geographic detectors to conduct correlation analysis between road carbon emissions and urban built environment, including single factor impact analysis and two-factor interactive impact analysis; The single factor impact analysis divides the study area and calculates the variance of the target variable in each sub-area, and finally obtains a detector value to reflect the contribution of the factor to the spatial phenomenon; The two-factor interaction analysis divides the region of factor combinations, calculates the variance of the target variable under different factor combinations, and then obtains the detector value of the interaction.

[0014] Another aspect of the present invention further provides a road carbon emission built environment impact assessment system based on multi-source trajectories, which is applied to the above-mentioned road carbon emission built environment impact assessment method based on multi-source trajectories, including a data acquisition and preprocessing module, a carbon emission quantification module, a built environment characterization module and an impact analysis module; The data acquisition and preprocessing module is used to obtain multi-source vehicle trajectory data, urban road network data, street view data, and point of interest data and to perform data integration, cleaning, denoising, matching, conversion, and correction; The carbon emission quantification module is used to calculate the traffic carbon emissions of different pollutants on the road scale based on multi-source vehicle trajectory data; The built environment characterization module is used to perform multi-dimensional characterization of the urban built environment based on urban road network data, street view image data and point of interest data, and calculate built environment indicators on a road scale; The impact analysis module is used to construct an assessment model of the impact of the urban built environment on road carbon emissions and use built environment indicators and road carbon emissions data for training and prediction, obtain the importance ranking of built environment indicators, and perform correlation analysis on road carbon emissions and urban built environment; based on the results of the importance ranking and correlation analysis of built environment indicators, the impact of road carbon emissions on the built environment is assessed.

[0015] Another aspect of the present invention provides a storage medium storing a program, which, when executed by a processor, implements the above-mentioned method for assessing the environmental impact of road carbon emissions based on multi-source trajectories.

[0016] Compared with the prior art, the present invention has the following advantages and beneficial effects: (1) Based on multi-source trajectory vehicle data, the present invention uses a micro emission model, combined with urban road network data, street view data, and POI data, to provide a road carbon emission built environment impact assessment method based on multi-source trajectory, construct a road carbon emission assessment index system, explore the degree and impact mechanism of various built environment indicators on road carbon emissions, and put forward suggestions for emission reduction policies. Compared with traditional social survey methods, this study saves a lot of material and financial resources and can be used for large-scale, automated road carbon emission quantification and its built environment impact assessment, with good benefits in transportation carbon emission reduction, urban transportation planning, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a flow chart of a method for assessing the environmental impact of road carbon emissions based on multi-source trajectories according to an embodiment of the present invention; Figure 2 is a schematic diagram of the structure of a road carbon emission built environmental impact assessment system based on multi-source trajectories according to an embodiment of the present invention; Figure 3 It is a schematic diagram of the structure of the storage medium of an embodiment of the present invention. DETAILED DESCRIPTION

[0018] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without making creative work are within the scope of protection of the present application.

[0019] Example: like Figure 1 As shown, this embodiment provides a method for assessing the environmental impact of road carbon emissions based on multi-source trajectories, including the following steps: S1: Obtain multi-source vehicle trajectory data, urban road network data, street view data, and point of interest data and perform preprocessing, specifically: S11: Obtain street view data, point of interest data, road network data and multi-source vehicle trajectory data. In this embodiment, data integration is performed through geographic information system professional software, and various types of data are aligned according to a unified spatial coordinate system to ensure that different data sources can be effectively compared and analyzed in the same coordinate system.

[0020] S12: Clean and process the integrated data, including deleting duplicate data, filling missing values, processing abnormal data, etc. Perform noise filtering and trajectory matching on street view data and multi-source vehicle trajectory data to ensure data quality and consistency.

[0021] S13: Convert the preprocessed road network data into a road network model. Usually, the roads are abstracted into line segments and these line segments are connected using a topological structure to represent the relationship between roads. The road network is topologically corrected to check whether there are closed loops, identify intersection nodes, and correct the connection relationship between road segments to ensure the accuracy and completeness of the road network structure.

[0022] S2: Based on multi-source vehicle trajectory data, calculate the traffic carbon emissions of various pollutants at the road scale, specifically: S21: Process the trajectory data of different types of vehicles (in this embodiment, taxis, buses, and trucks are taken as examples), calculate the time interval between consecutive points, and compare it with a set threshold, and filter out trajectory points whose time interval is lower than the set threshold (set to 120 seconds in this embodiment).

[0023] S22: Calculate the instantaneous speed and acceleration of the vehicle at the trajectory point according to the time interval and distance between each point of the vehicle trajectory data, and filter out the drift points and abnormal points of speed and acceleration in the vehicle trajectory points (in this embodiment, the speed is greater than 300km / h and the acceleration is greater than 10m / s 2 points are identified as outliers).

[0024] S23: Calculate the carbon emissions of each point based on the speed and acceleration of each vehicle GPS trajectory point after filtering using a micro emission model, emission factor method, life cycle analysis, macro emission model or traffic flow model.

[0025] Furthermore, the micro emission model is a method for estimating carbon dioxide (CO 2 ), nitrogen oxides (NO X ), particulate matter (PM) and volatile organic compound (VOC) instantaneous emission models have certain advantages in small-scale carbon emission calculations, such as the road-scale carbon emission prediction in this application. When using micro-emission models to calculate emissions, Place the vehicle at the GPS pointp The emission at E p , the instantaneous velocity and acceleration are recorded as S p and a p , E p The calculation formula is: E p = f 1 + f 2 S p + f 3 S p 2 + f 4 a p + f 5 a p 2 + f 6 S p a p ; in, f 1 arrive f 6 The vehicle's fuel type is determined by the type of vehicle and the type of pollutants calculated. The vehicle's fuel type is determined by the type of vehicle.

[0026] S24: Use the nearest neighbor algorithm to assign each point to an urban road based on the urban road data, find the road closest to the current point, add up the carbon emissions of all points assigned to each road section, and obtain the emissions of the four pollutants on the road section. Perform principal component analysis on the same pollutant data of different types of vehicles to obtain the emissions of each pollutant. In particular, in addition to the nearest neighbor algorithm, the allocation can also be done by building a buffer, inverse distance weighting or cluster analysis. The nearest neighbor algorithm can make full use of the trajectory point data without repeated application.

[0027] S3: Based on urban road network data, street view image data and point of interest data, a multi-dimensional representation of the urban built environment is performed, and built environment indicators at the road scale are calculated, specifically: S31: Use massive street view data to characterize the urban built environment, use a semantic segmentation model (in this embodiment, a deep fully convolutional neural network structure SegNet is selected) to perform semantic segmentation on street view images, extract street view elements such as buildings, roads, vehicles, sidewalks, trees, sky, pedestrians, etc. in street view images, construct a street view index from the street view perspective, calculate the proportion of various street view elements in the street view image, and obtain 10 street view built environment indicators such as sky openness, street enclosure, green view rate, building enclosure, crowd concentration, grass proportion, tree proportion, plant proportion, sign density index, and cycling suitability index. Finally, the street view sampling points are spatially connected with the road network data, and the data of all street view sampling points on the road section are averaged to obtain the street view index on the road scale.

[0028] Specifically, the calculation method of streetscape built environment indicators is as follows: (1) Green View Scale (GSV) refers to the ratio of tree, grass and other plant pixels to the total pixels in a street view image and is calculated using the following formula: ; in, is the green view rate, is the number of pixels of trees in the street view image in the front, back, left, and right directions, is the number of pixels of grass in the street view image in the front, back, left, and right directions, is the number of pixels of other plants in the street view images in the front, back, left, and right directions, It is the total number of pixels of street view images in the four directions of front, back, left and right.

[0029] (2) Sky freedom fraction (SVF) refers to the ratio of sky pixels to total pixels in street view images and is calculated using the following formula: ; in, For the openness of the sky, is the number of sky pixels in the street view image in the front, back, left, and right directions, It is the total number of pixels of street view images in the four directions of front, back, left and right.

[0030] (3) Interface Enclosure (IF) refers to the proportion of building, wall and tree pixels to the total pixels in the street view image and is calculated using the following formula: ; in, is the number of pixels of buildings in the street view images in the front, back, left, and right directions, is the number of pixels of the wall in the street view image in the front, back, left and right directions, is the number of pixels of trees in the street view image in the front, back, left, and right directions, It is the total number of pixels of street view images in the four directions of front, back, left and right.

[0031] (4) Building enclosure (BF) refers to the ratio of building pixels to total pixels in street view images and is calculated using the following formula: ; in, is the number of pixels of buildings in the street view images in the front, back, left, and right directions, It is the total number of pixels of street view images in the four directions of front, back, left and right.

[0032] (5) Crowd density (PERSONI) indicates the proportion of pedestrian pixels to total pixels in the street view image and is calculated using the following formula: ; in, is the number of pixels of pedestrians in the street view images in the front, back, left, and right directions, It is the total number of pixels of street view images in the four directions of front, back, left and right.

[0033] (6) Grass fraction (GI) refers to the ratio of grass pixels to total pixels in the street view image and is calculated using the following formula: ; in, is the number of pixels of grass in the street view image in the front, back, left, and right directions, It is the total number of pixels of street view images in the four directions of front, back, left and right.

[0034] (7) Tree percentage (TI) refers to the ratio of tree pixels to total pixels in the street view image and is calculated using the following formula: ; in, is the number of pixels of trees in the street view image in the front, back, left, and right directions, It is the total number of pixels of street view images in the four directions of front, back, left and right.

[0035] (8) Plant proportion (PI) refers to the proportion of plant pixels to total pixels in the street view image and is calculated using the following formula: ; in, is the number of pixels of other plants in the street view images in the front, back, left, and right directions, It is the total number of pixels of street view images in the four directions of front, back, left and right.

[0036] (9) Signage Density Index (SDI) represents the ratio of streetlight and road sign pixels to the total pixels in a street view image. It is calculated using the following formula: ; in, is the number of pixels of street lights in the street view images in the four directions of front, back, left, and right. is the number of pixels of road signs in the street view images in the four directions of front, back, left and right. It is the total number of pixels of street view images in the four directions of front, back, left and right.

[0037] (10) The Cycling Suitability Index (CSI) represents the ratio of road and sidewalk pixels to the total pixels in a street view image and is calculated using the following formula: ; in, is the number of pixels of the road in the street view image in the four directions of front, back, left and right. is the number of pixels of the sidewalk in the street view images in the four directions of front, back, left and right, It is the total number of pixels of street view images in the four directions of front, back, left and right.

[0038] S32: Use road network data to construct six road-related built environment indicators, including road linearity index, betweenness centrality, eigenvector centrality, closeness centrality, degree centrality, and clustering coefficient. The road linearity index is obtained by the ratio of the distance from the starting point to the end point of the road section to the distance. The betweenness centrality, eigenvector centrality, closeness centrality, degree centrality, and clustering coefficient are obtained by taking the average of the indicators of the starting point and the end point of the road section.

[0039] Specifically, the calculation method of road-related built environment indicators is as follows: (1) Betweenness centrality refers to the number of shortest paths through a node in a network and is calculated using the following formula: ; in, is the betweenness centrality of the node, Indicates passing through node i , and is the number of paths of the shortest path, Indicates connection s and t The number of shortest paths.

[0040] (2) The linearity index represents the ratio of the distance from the starting point to the end point of the road segment to the distance traveled, and is calculated using the following formula: ; in,L is the linear index, is the distance from the starting point to the end point of the road segment, It is the distance from the starting point to the end point of the road segment.

[0041] (3) Closeness refers to the average value of the closeness of the two nodes at the starting point and the end point of the road section, which is calculated by the following formula: ; in, To approach centrality, For Node v To Node u The distance For Node u The sum of the shortest distances to all nodes, n is the total number of nodes in the road network.

[0042] (4) Degree centrality (Degree) represents the average degree centrality of the two nodes at the starting point and the end point of the road section, and is calculated using the following formula: ; in, is the degree centrality, For the node i The number of directly connected edges, N is the total number of nodes in the road network.

[0043] (5) Eigenvector centrality represents the average value of the eigenvector centrality of the two nodes at the start and end of the road section, and is calculated using the following formula: ; Among them, the eigenvector centrality is calculated by a recursive method, For Node i The eigenvector centrality of c is the normalization factor, g is the total number of nodes in the road network, For Node i and j The adjacency relationship is 1 or 0. For Node j The eigenvector centrality of .

[0044] (6) Local clustering refers to the average value of the clustering coefficients of the two nodes at the start and end of the road segment, which is calculated using the following formula: ; in, is the clustering coefficient, For Node u The number of triangles that can be formed between neighboring nodes, For Node u The number of neighbor nodes.

[0045] S33: Use point of interest data to construct urban built environment indicators related to POIs, including eight indicators such as land use richness, Shannon diversity index, Simpson diversity index, and road land use type ratio. Perform buffer analysis on the road network data and extract the 100-meter buffer of the road as the road space. Secondly, spatially connect the road space with the POI point data, and use the POI point data in the road space of each road section to calculate the urban built environment indicators of the road section. Among them, when calculating land use richness, Shannon index, and Simpson index, POIs are divided into 22 small categories; when calculating the proportion of road land use types, POIs are divided into four major categories: commercial, residential, transportation, and cultural, and the proportion of POI points in each category is calculated as the proportion of land use types in the road section.

[0046] Specifically, the calculation method of built environment indicators related to points of interest (POI points) is as follows: (1) Richness index refers to the richness index of POI points within the 100m buffer zone of the road section, which is calculated by the following formula: ;in, R is the richness index, For the i The number of class POIs, N is the total number of categories.

[0047] (2) Shannon diversity index (Shannon) represents the Shannon diversity index of POI points within the 100m buffer zone of the road section and is calculated using the following formula: ; in, SN is the Shannon diversity index, It is the proportion of POI of each type within the 100m buffer zone.

[0048] (3) Simpson diversity index (Simpson) represents the Simpson diversity index of POI points within the 100m buffer zone of the road section and is calculated using the following formula: ; in, SP is the Simpson diversity index, is the number of POIs of each type within the 100m buffer zone of the road section, N It is the number of categories of all POIs within the 100m buffer zone of the road section.

[0049] (4) Commercial area ratio (Commercial) refers to the ratio of commercial POI points to the total POI points within the 100m buffer zone of the road section, which is calculated by the following formula: ; in, As a percentage of commercial areas, The number of commercial POI points within the 100m buffer zone of the road section. It is the total number of POI points within the 100m buffer zone of the road section.

[0050] (5) Residential ratio refers to the ratio of residential POI points to the total POI points within the 100m buffer zone of the road section, which is calculated by the following formula: ; in, The proportion of residential areas, is the number of residential POI points within the 100m buffer zone of the road section. It is the total number of POI points within the 100m buffer zone of the road section.

[0051] (6) Traffic zone ratio (Traffic) refers to the ratio of traffic type POI points to the total POI points within the 100m buffer zone of the road section, which is calculated by the following formula: ; in, is the proportion of traffic area, The number of POI points of the traffic type within the 100m buffer zone of the road section. It is the total number of POI points within the 100m buffer zone of the road section.

[0052] (7) Culture area ratio (Culture) refers to the ratio of cultural POI points to the total POI points within the 100m buffer zone of the road section, which is calculated by the following formula: ; in, The proportion of cultural areas is The number of cultural POI points within the 100m buffer zone of the road section. It is the total number of POI points within the 100m buffer zone of the road section.

[0053] S4: Construct an assessment model for the impact of urban built environment on road carbon emissions and use built environment indicators and road carbon emissions data for training and prediction to obtain the importance ranking of built environment indicators, specifically: S41: Use geographically weighted regression, random forest model, geographically weighted-random forest model and graph convolutional neural network to model, and divide all streets into training samples and verification samples according to a certain ratio; in this embodiment, the training sample ratio is 75%; input the above-mentioned built environment indicators and road carbon emissions into geographically weighted regression, random forest model, geographically weighted-random forest model and graph convolutional neural network, and perform iterative training to find the optimal parameters, generate the best fitting model, and rank the importance of each built environment indicator.

[0054] S5: Conduct correlation analysis between road carbon emissions and urban built environment, specifically: S51: Use geographic detectors to conduct correlation analysis between road carbon emissions and urban built environment, including single factor impact analysis and two-factor interactive impact analysis.

[0055] The single factor impact analysis of geographic detectors mainly reveals the influence of a single factor in different regions by quantifying its explanatory power on the spatial heterogeneity of the target variable. It divides the study area and calculates the variance of the target variable in each sub-region, and finally obtains a detector value that reflects the contribution of the factor to the spatial phenomenon. The higher the detector value, the stronger the explanatory power of the factor on the target variable. This method is intuitive and easy to apply, and is suitable for analyzing the role of a single factor in spatial distribution.

[0056] The two-factor interaction analysis focuses on the interaction between two factors and evaluates their joint impact on the target variable. By dividing the area of ​​the factor combination, the variance of the target variable under different factor combinations is calculated, and then the detector value of the interaction is obtained. If the interaction is significant, it means that there is a synergistic or inhibitory effect between the two factors, which affects the spatial distribution of the target variable. This method is suitable for exploring complex geographical phenomena under the joint action of multiple factors, and helps to deeply understand the relationship and spatial pattern between factors. In particular, in this embodiment, correlation analysis, spatial autocorrelation analysis, random forest model, support vector machine, etc. can also be used for association analysis.

[0057] S6: Evaluate the built environmental impact of road carbon emissions based on the importance ranking of built environmental indicators and the results of correlation analysis, specifically: Based on the results of model construction and correlation analysis, the degree of influence of various built environment indicators on road carbon emissions is obtained, and their influence mechanisms are analyzed to provide suggestions for urban planning oriented towards transportation carbon emission reduction. The above three types of urban built environment indicators and the optimal fitting model are used to construct a road carbon emission indicator evaluation system and prediction model to improve the generalization of the model and apply it to the assessment and prediction of road carbon emissions in various regions.

[0058] like Figure 2As shown, in another embodiment of the present application, a road carbon emission built environment impact assessment system based on multi-source trajectories is provided, the system comprising a data acquisition and preprocessing module, a carbon emission quantification module, a built environment characterization module and an impact analysis module; The data acquisition and preprocessing module is used to obtain multi-source vehicle trajectory data, urban road network data, street view data, and point of interest data and to perform data integration, cleaning, denoising, matching, conversion, and correction; The carbon emission quantification module is used to calculate the traffic carbon emissions of different pollutants on the road scale based on multi-source vehicle trajectory data; The built environment characterization module is used to perform multi-dimensional characterization of the urban built environment based on urban road network data, street view image data and point of interest data, and calculate built environment indicators on a road scale; The impact analysis module is used to construct an assessment model of the impact of the urban built environment on road carbon emissions and use built environment indicators and road carbon emissions data for training and prediction, obtain the importance ranking of built environment indicators, and perform correlation analysis on road carbon emissions and urban built environment; based on the results of the importance ranking and correlation analysis of built environment indicators, the impact of road carbon emissions on the built environment is assessed.

[0059] It should be noted here that the system provided in the above embodiment is only illustrated by the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure is divided into different functional modules to complete all or part of the functions described above. The system is an environmental impact assessment method for road carbon emissions based on multi-source trajectories applied to the above embodiment.

[0060] like Figure 3 As shown, in another embodiment of the present application, a storage medium is further provided, storing a program, and when the program is executed by a processor, a method for assessing the environmental impact of road carbon emissions based on multi-source trajectories is implemented, specifically: S1: Obtain multi-source vehicle trajectory data, urban road network data, street view data, and point of interest data and perform data integration, cleaning, denoising, matching, conversion, and correction; S2: Calculate traffic carbon emissions of different pollutants at the road scale based on multi-source vehicle trajectory data; S3: Based on urban road network data, street view image data and point of interest data, multi-dimensional characterization of urban built environment is carried out, and built environment indicators at the road scale are calculated; S4: Construct an assessment model of the impact of urban built environment on road carbon emissions and use built environment indicators and road carbon emissions data for training and prediction to obtain the importance ranking of built environment indicators; S5: Correlation analysis between road carbon emissions and urban built environment; S6: Evaluate the built environment impact of road carbon emissions based on the results of importance ranking and correlation analysis of built environment indicators.

[0061] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0062] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.

Claims

1. A road carbon emission environmental impact assessment method based on multi-source trajectories is characterized by: The steps include: Obtain multi-source vehicle trajectory data, urban road network data, street view data, and point of interest data and perform data integration, cleaning, denoising, matching, conversion, and correction; Calculate traffic carbon emissions of different pollutants at the road scale based on multi-source vehicle trajectory data; Based on urban road network data, street view image data and point of interest data, a multi-dimensional representation of the urban built environment is performed, and built environment indicators at the road scale are calculated; Construct an assessment model of the impact of urban built environment on road carbon emissions and use built environment indicators and road carbon emissions data for training and prediction to obtain the importance ranking of built environment indicators; Conduct correlation analysis between road carbon emissions and urban built environment; The built environmental impact of road carbon emissions is evaluated based on the results of importance ranking and correlation analysis of built environmental indicators.

2. The method for evaluating the environmental impact of road carbon emissions based on multi-source trajectories according to claim 1 is characterized in that: The acquisition of multi-source vehicle trajectory data, urban road network data, street view data, and point of interest data and the data integration, cleaning, denoising, matching, conversion, and correction are specifically as follows: Obtain street view data, point of interest data, road network data, and multi-source vehicle trajectory data, and integrate the data to align various types of data according to a unified spatial coordinate system; Clean and process the integrated data, including: deleting duplicate data, filling missing values, and processing abnormal data; performing noise filtering and trajectory matching on street view data and multi-source vehicle trajectory data; The processed road network data is converted into a road network model. The roads are abstracted into line segments and these line segments are connected using a topological structure to represent the relationship between roads. The road network is topologically corrected to check whether there are closed loops, identify intersection nodes, and correct the connection relationship between road segments.

3. The method for evaluating the environmental impact of road carbon emissions based on multi-source trajectories according to claim 1 is characterized in that: The method of calculating the traffic carbon emissions of different pollutants on the road scale based on multi-source vehicle trajectory data is as follows: Calculate the time interval between consecutive points in different types of vehicle trajectory data, compare it with the set threshold, and filter out trajectory points with a time interval lower than the set threshold; Calculate the instantaneous speed and acceleration of the vehicle at the trajectory point according to the time interval and distance between each point of the vehicle trajectory data, and filter out the drift points and abnormal points of speed and acceleration in the vehicle trajectory points, where the abnormal points are trajectory points whose speed or acceleration is greater than a set threshold; According to the instantaneous speed, acceleration and fuel type of each vehicle GPS track point after filtering, the carbon emissions of each vehicle GPS track point are calculated using micro emission model, emission factor method, life cycle analysis, macro emission model or traffic flow model, specifically: Place the vehicle at the GPS point p The emission at E p , the instantaneous velocity and acceleration are recorded as S p and a p , E p The calculation formula is: E p = f 1+ f 2 S p + f 3 S p 2 + f 4 a p + f 5 a p 2 + f 6 S p a p ； in, f 1 to f 6 Determined by the vehicle fuel type and the type of pollutants calculated. The vehicle fuel type is determined by the type of vehicle; Use the nearest neighbor algorithm, buffer construction, inverse distance weighted or cluster analysis method to assign each point to an urban road based on the urban road data, find the road closest to the current point, add up the carbon emissions of all points assigned to each road section, and get the emissions of the four pollutants on the road section. Perform principal component analysis on the same pollutant data of different types of vehicles to get the emissions of each pollutant.

4. The method for evaluating the environmental impact of road carbon emissions based on multi-source trajectories according to claim 1 is characterized in that: Characterize the urban built environment based on street view image data, specifically: The urban built environment is characterized by using street view image data, and the semantic segmentation model is used to perform semantic segmentation on the street view images. The street landscape elements such as buildings, roads, vehicles, sidewalks, trees, sky, and pedestrians in the street view pictures are extracted, and the street landscape index from the street view perspective is constructed. The proportion of various street landscape elements in the street view images is calculated, and the street view built environment indicators such as sky openness, street enclosure, green view rate, building enclosure, crowd concentration, grassland proportion, tree proportion, plant proportion, sign density index, and cycling suitability index are obtained. Finally, the street view sampling points are spatially connected with the road network data, and the data of all street view sampling points on the road section are averaged to obtain the street landscape index at the road scale.

5. The method for evaluating the environmental impact of road carbon emissions based on multi-source trajectories according to claim 1 is characterized in that: The urban built environment is characterized based on urban road network data, specifically: Road network data are used to construct road-related built environment indicators such as road linearity index, betweenness centrality, eigenvector centrality, closeness centrality, degree centrality, and clustering coefficient. The road linearity index is obtained by the ratio of the distance from the starting point to the end point of the road section to the distance, and the betweenness centrality, eigenvector centrality, closeness centrality, degree centrality, and clustering coefficient are obtained by taking the average of the indicators of the starting point and the end point of the road section.

6. The method for evaluating the environmental impact of road carbon emissions based on multi-source trajectories according to claim 1 is characterized in that: Characterize the urban built environment based on point of interest data, specifically: Use point of interest data to construct urban built environment indicators related to POIs, including land use richness, Shannon diversity index, Simpson diversity index, and proportion of road land use types; The road network data is analyzed by buffer zone analysis, and the 100-meter buffer zone of the road is extracted as the road space. Secondly, the road space is spatially connected with the POI point data, and the urban built environment index of each road section is calculated using the POI point data in the road space of the road section. Among them, when calculating land use richness, Shannon diversity index, and Simpson diversity index, POIs are divided into different subcategories; when calculating the proportion of road land use types, POIs are divided into four major categories: commercial, residential, transportation, and cultural, and the proportion of POI points in each category is calculated as the proportion of land use types on the road section.

7. The method for evaluating the environmental impact of road carbon emissions based on multi-source trajectories according to claim 1 is characterized in that: The evaluation model of the impact of urban built environment on road carbon emissions is constructed and the built environment indicators and road carbon emissions data are used for training and prediction to obtain the importance ranking of built environment indicators, specifically: Geographically weighted regression, random forest model, geographically weighted-random forest model, and graph convolutional neural network are used for modeling. All streets are divided into training samples and validation samples according to a certain ratio. The proportion of training samples is 75%. The above built environment indicators and road carbon emissions are input into geographically weighted regression, random forest model, geographically weighted-random forest model and graph convolutional neural network, and iterative training is carried out to find the optimal parameters, generate the best fitting model, and rank the importance of each built environment indicator.

8. The method for evaluating the environmental impact of road carbon emissions based on multi-source trajectories according to claim 1 is characterized in that: The correlation analysis between road carbon emissions and urban built environment is as follows: Use geographic detectors to conduct correlation analysis between road carbon emissions and urban built environment, including single factor impact analysis and two-factor interactive impact analysis; The single factor impact analysis divides the study area and calculates the variance of the target variable in each sub-area, and finally obtains a detector value to reflect the contribution of the factor to the spatial phenomenon; The two-factor interaction analysis divides the region of factor combinations, calculates the variance of the target variable under different factor combinations, and then obtains the detector value of the interaction.

9. A road carbon emission environmental impact assessment system based on multi-source trajectories is constructed, characterized in that: A method for assessing the built environment impact of road carbon emissions based on multi-source trajectories applied to any one of claims 1-8, comprising a data acquisition and preprocessing module, a carbon emission quantification module, a built environment characterization module, and an impact analysis module; The data acquisition and preprocessing module is used to obtain multi-source vehicle trajectory data, urban road network data, street view data, and point of interest data and to perform data integration, cleaning, denoising, matching, conversion, and correction; The carbon emission quantification module is used to calculate the traffic carbon emissions of different pollutants on the road scale based on multi-source vehicle trajectory data; The built environment characterization module is used to perform multi-dimensional characterization of the urban built environment based on urban road network data, street view image data and point of interest data, and calculate built environment indicators on a road scale; The impact analysis module is used to construct an assessment model of the impact of the urban built environment on road carbon emissions and use built environment indicators and road carbon emissions data for training and prediction, obtain the importance ranking of built environment indicators, and perform correlation analysis on road carbon emissions and urban built environment; based on the results of the importance ranking and correlation analysis of built environment indicators, the impact of road carbon emissions on the built environment is assessed.

10. A storage medium storing a program, characterized in that: When the program is executed by a processor, the method for assessing the built environment impact of road carbon emissions based on multi-source trajectories as described in any one of claims 1 to 8 is implemented.

Images