A carbon emission calculation method and device based on dynamic simulation and cruise measurement

Abstract

A method and apparatus for calculating carbon emissions based on dynamic simulation and mobile monitoring is disclosed. The method includes: controlling a mobile monitoring vehicle to collect real-time carbon dioxide data, real-time panoramic image data, and real-time meteorological data from road monitoring points; identifying and analyzing the real-time panoramic image data to obtain environmental analysis results; inputting the real-time meteorological data and environmental analysis results into a background concentration simulation and prediction model to obtain the predicted background carbon dioxide concentration at the monitoring points; calculating the carbon dioxide increment at the monitoring points based on the real-time carbon dioxide data and the predicted background carbon dioxide concentration at the monitoring points; and performing spatiotemporal integration of the carbon dioxide increment at multiple road monitoring points based on a Gaussian plume model to obtain the road carbon emissions. This method can effectively improve the accuracy of road carbon emission calculation.

Description

Technical Field

[0001] This invention relates to the field of carbon emission calculation technology, and in particular to a carbon emission calculation method and apparatus based on dynamic simulation and mobile field measurement. Background Technology

[0002] Currently, many cities and transportation management agencies rely on emission models based on vehicle type, quantity, and operating status to estimate transportation carbon emissions. These models typically estimate carbon emissions by statistically analyzing information such as vehicle type, quantity, and speed, combined with fuel consumption data and emission factors. A representative model is the MOVES (Motor Vehicle Emission Simulator) model, which is now widely used both domestically and internationally for estimating carbon dioxide (CO2) and other greenhouse gases and pollutants in the transportation sector.

[0003] The MOVES model calculates carbon emissions from road traffic in four steps: (1) First, researchers collect and statistically analyze data on road traffic flow and vehicle types (such as light vehicles, heavy vehicles, and trucks) using road intersection monitoring equipment, traffic surveys, or navigation map software; (2) Second, researchers calculate the traffic flow status, including acceleration, deceleration, idling, and stable driving, using monitoring equipment and navigation software; (3) Based on the vehicle's operating status and weather conditions, the MOVES model is used to calculate the fuel consumption (including different types of fuel, such as #92 and #95 gasoline) for each type of vehicle; (4) Finally, the emission factors for each vehicle type and fuel type are applied to convert fuel consumption into carbon dioxide emissions. Patent CN119128334A discloses a method for calculating carbon emissions from urban roads under dynamic calibration conditions. Referring to the MOVES model, it constructs basic carbon emission factors for different vehicle models under different operating conditions, calculates the operating conditions second by second based on small sample GPS data, uses an improved NSGA-II algorithm to divide the optimal speed range, obtains the operating condition distribution of different speed ranges, and dynamically updates the operating condition correction parameters of the carbon emission factor library.

[0004] While the MOVES model and similar models based on traffic flow and emission factors can provide macroscopic estimates of traffic carbon emissions, they also have significant drawbacks. Specifically, existing energy consumption-based emission estimation models (such as MOVES) rely on statistical calculations of empirical data and emission factor calculations to obtain carbon emissions, failing to consider the influence of background carbon dioxide concentrations, resulting in low accuracy of the calculated carbon emissions. Currently, existing methods for calculating urban background carbon dioxide concentrations mainly include station observation and mobile observation methods. Station observation primarily involves setting up urban atmospheric carbon dioxide concentration background monitoring stations to record carbon dioxide concentrations at locations without carbon emission sources as background concentration values. Mobile observation, on the other hand, uses the minimum value of a relatively long mobile monitoring segment that includes the field area as the background concentration value. However, both methods are characterized by temporal variation and spatial non-heterogeneity; that is, the background concentration value for the entire city has only one value at any given time, failing to highlight the differences between different areas. Summary of the Invention

[0005] This invention provides a carbon emission calculation method and apparatus based on dynamic simulation and mobile field measurement, which can effectively improve the accuracy of road carbon emission calculation.

[0006] A carbon emission calculation method based on dynamic simulation and mobile surveys includes:

[0007] Pre-control mobile vehicles to collect carbon dioxide concentration data, panoramic image data and meteorological data at road sampling points, and perform multi-dimensional coupling and filtering to obtain background carbon dioxide concentration data;

[0008] The pre-established deep learning model is trained based on the background carbon dioxide concentration data to obtain a background concentration simulation prediction model.

[0009] The mobile monitoring vehicle collects real-time carbon dioxide data, real-time panoramic image data, and real-time meteorological data from road monitoring points. The real-time panoramic image data is then identified and analyzed to obtain environmental analysis results.

[0010] The real-time meteorological data and environmental analysis results are input into the background concentration simulation and prediction model to obtain the predicted background carbon dioxide concentration at the monitoring point.

[0011] The carbon dioxide increment at the monitoring point is calculated based on the real-time carbon dioxide data and the predicted background concentration of carbon dioxide at the monitoring point.

[0012] The carbon dioxide increment at multiple monitoring points along the road is calculated based on the Gaussian plume model to obtain the road carbon emissions.

[0013] Furthermore, after collecting carbon dioxide concentration data and meteorological data from road sampling points, the process also includes:

[0014] The carbon dioxide concentration data and meteorological data were cleaned and standardized.

[0015] Furthermore, carbon dioxide concentration data, panoramic imagery data, and meteorological data are coupled and filtered in multiple dimensions, including:

[0016] Time matching is performed based on the acquisition time of the carbon dioxide concentration data, panoramic image data, and meteorological data.

[0017] The panoramic image data is identified using a pre-trained image recognition model. Based on the recognition results, the non-traffic-affected area and its corresponding panoramic image acquisition time and environmental information are determined.

[0018] Based on the meteorological data, calculate the turbulence intensity in the non-traffic-affected area at the corresponding panoramic image acquisition time;

[0019] Obtain the road congestion index of the non-traffic-affected area at the corresponding panoramic image acquisition time, and calculate the background credibility score of the non-traffic-affected area at the corresponding panoramic image acquisition time based on the turbulence intensity and the road congestion index.

[0020] Based on the background credibility score, the carbon dioxide concentration data of the non-traffic-affected area at the corresponding panoramic image acquisition time are filtered to obtain filtered carbon dioxide concentration data.

[0021] The screened carbon dioxide concentration data is subjected to cluster analysis, and the clustering results are correlated with the corresponding environmental information and meteorological data to obtain the background carbon dioxide concentration data.

[0022] Furthermore, the identification results include vehicles, vegetation, and building facades;

[0023] Based on the identification results, the non-traffic-affected areas and their corresponding panoramic image acquisition times and environmental information are determined, including:

[0024] Calculate the pixel percentage of the vehicle, vegetation, and building facade in the panoramic image respectively. Determine the sampling point corresponding to the panoramic image where the pixel percentage of the vehicle is less than the preset value as the non-traffic-affected area, and record the corresponding panoramic image acquisition time, the pixel percentage of the vegetation, and the pixel percentage of the building facade.

[0025] Furthermore, the meteorological data includes wind speed;

[0026] Based on the meteorological data, the turbulence intensity in the non-traffic-affected area at the corresponding panoramic image acquisition time is calculated, including:

[0027] Based on the time matching results, obtain the wind speed data of the non-traffic-affected area within a preset time period before the corresponding panoramic image acquisition time;

[0028] Calculate the standard deviation of wind speed within the preset time period based on the wind speed data;

[0029] The turbulence intensity is calculated based on the wind speed standard deviation and the wind speed in the non-traffic-affected area at the corresponding panoramic image acquisition time.

[0030] Furthermore, the screened carbon dioxide data is subjected to cluster analysis, and the clustering results are correlated with corresponding environmental information and meteorological data to obtain the background carbon dioxide concentration data, including:

[0031] The carbon dioxide data was Z-score standardized, and outlier data points were removed based on the Z-score standardization results.

[0032] Calculate the coefficient of variation for the remaining screened carbon dioxide data after removing outliers, and select candidate carbon dioxide background concentration data based on the coefficient of variation.

[0033] The DBSCAN clustering algorithm was used to perform density analysis on the candidate carbon dioxide background concentration data, and the cluster center value with a density value greater than the preset density value was taken as the clustering result.

[0034] Based on the time matching results, the clustering results are correlated with the corresponding meteorological data, the pixel proportion of vegetation, and the pixel proportion of building facades to obtain the background carbon dioxide concentration data.

[0035] Furthermore, the real-time panoramic image data is identified and analyzed to obtain environmental analysis results, including:

[0036] The real-time panoramic image data is input into the image recognition model for recognition and analysis to obtain the pixel ratio of vegetation and the pixel ratio of building facades at the monitoring points.

[0037] Furthermore, the carbon dioxide increment at the monitoring point is the difference between the real-time carbon dioxide concentration and the predicted background carbon dioxide concentration at the monitoring point.

[0038] Furthermore, the real-time meteorological data includes real-time wind speed;

[0039] Based on the Gaussian plume model, the spatiotemporal integration of carbon dioxide increments at multiple monitoring points along the road is used to obtain road carbon emissions, including:

[0040] Calculate the area of ​​the monitoring region based on the monitoring points;

[0041] Based on the area of ​​the monitoring area and the real-time wind speed, the carbon dioxide emission rate of the road segment corresponding to the monitoring point is obtained using the Gaussian plume model.

[0042] The carbon dioxide emission rates calculated from various monitoring points along the same road are summed to obtain the total carbon dioxide emission rate of the road.

[0043] The road carbon emissions are calculated based on the total carbon dioxide emission rate.

[0044] A carbon emission calculation device based on dynamic simulation and mobile survey, comprising:

[0045] The sample collection module is used to pre-control the mobile vehicle to collect carbon dioxide concentration data, panoramic image data and meteorological data at road sampling points and perform multi-dimensional coupling and filtering to obtain background carbon dioxide concentration data.

[0046] The training module is used to train a pre-established deep learning model based on the background carbon dioxide concentration data to obtain a background concentration simulation prediction model.

[0047] The real-time data acquisition module is used to control the mobile vehicle to collect real-time carbon dioxide data, real-time panoramic image data and real-time meteorological data from road monitoring points, and to identify and analyze the real-time panoramic image data to obtain environmental analysis results.

[0048] The prediction module is used to input the real-time meteorological data and environmental analysis results into the background concentration simulation prediction model to obtain the predicted background concentration of carbon dioxide at the monitoring point.

[0049] The incremental calculation module is used to calculate the carbon dioxide increment at the monitoring point based on the real-time carbon dioxide data and the predicted background concentration of carbon dioxide at the monitoring point.

[0050] The emissions calculation module is used to calculate the carbon dioxide increment at multiple monitoring points along the road based on a Gaussian plume model, thereby obtaining the road carbon emissions.

[0051] Furthermore, the sample acquisition module is also used for:

[0052] The carbon dioxide concentration data and meteorological data were cleaned and standardized.

[0053] Furthermore, the sample acquisition module performs multi-dimensional coupled filtering of carbon dioxide concentration data, panoramic image data, and meteorological data, including:

[0054] Time matching is performed based on the acquisition time of the carbon dioxide concentration data, panoramic image data, and meteorological data.

[0055] The panoramic image data is identified using a pre-trained image recognition model. Based on the recognition results, the non-traffic-affected area and its corresponding panoramic image acquisition time and environmental information are determined.

[0056] Based on the meteorological data, calculate the turbulence intensity in the non-traffic-affected area at the corresponding panoramic image acquisition time;

[0057] Obtain the road congestion index of the non-traffic-affected area at the corresponding panoramic image acquisition time, and calculate the background credibility score of the non-traffic-affected area at the corresponding panoramic image acquisition time based on the turbulence intensity and the road congestion index.

[0058] Based on the background credibility score, the carbon dioxide concentration data of the non-traffic-affected area at the corresponding panoramic image acquisition time are filtered to obtain filtered carbon dioxide concentration data.

[0059] The screened carbon dioxide concentration data is subjected to cluster analysis, and the clustering results are correlated with the corresponding environmental information and meteorological data to obtain the background carbon dioxide concentration data.

[0060] Furthermore, the identification results include vehicles, vegetation, and building facades;

[0061] The sample acquisition module determines the non-traffic-affected area and its corresponding panoramic image acquisition time based on the recognition results, including:

[0062] Calculate the pixel percentage of the vehicle, vegetation, and building facade in the panoramic image respectively. Determine the sampling point corresponding to the panoramic image where the pixel percentage of the vehicle is less than the preset value as the non-traffic-affected area, and record the corresponding panoramic image acquisition time, the pixel percentage of the vegetation, and the pixel percentage of the building facade.

[0063] Furthermore, the meteorological data includes wind speed;

[0064] The sample acquisition module calculates the turbulence intensity in the non-traffic-affected area at the corresponding panoramic image acquisition time based on the meteorological data, including:

[0065] Based on the time matching results, obtain the wind speed data of the non-traffic-affected area within a preset time period before the corresponding panoramic image acquisition time;

[0066] Calculate the standard deviation of wind speed within the preset time period based on the wind speed data;

[0067] The turbulence intensity is calculated based on the wind speed standard deviation and the wind speed in the non-traffic-affected area at the corresponding panoramic image acquisition time.

[0068] Furthermore, the sample collection module performs cluster analysis on the screened carbon dioxide data, and correlates the clustering results with corresponding environmental information and meteorological data to obtain the background carbon dioxide concentration data, including:

[0069] The carbon dioxide data was Z-score standardized, and outlier data points were removed based on the Z-score standardization results.

[0070] Calculate the coefficient of variation for the remaining screened carbon dioxide data after removing outliers, and select candidate carbon dioxide background concentration data based on the coefficient of variation.

[0071] The DBSCAN clustering algorithm was used to perform density analysis on the candidate carbon dioxide background concentration data, and the cluster center value with a density value greater than the preset density value was taken as the clustering result.

[0072] Based on the time matching results, the clustering results are correlated with the corresponding meteorological data, the pixel proportion of vegetation, and the pixel proportion of building facades to obtain the background carbon dioxide concentration data.

[0073] Furthermore, the real-time data acquisition module identifies and analyzes the real-time panoramic image data to obtain environmental analysis results, including:

[0074] The real-time panoramic image data is input into the image recognition model for recognition and analysis to obtain the pixel ratio of vegetation and the pixel ratio of building facades at the monitoring points.

[0075] Furthermore, the carbon dioxide increment at the monitoring point is the difference between the real-time carbon dioxide concentration and the predicted background carbon dioxide concentration at the monitoring point.

[0076] Furthermore, the real-time meteorological data includes real-time wind speed;

[0077] The emission calculation module, based on a Gaussian plume model, performs spatiotemporal integration of carbon dioxide increments at multiple monitoring points along the road to obtain road carbon emissions, including:

[0078] Calculate the area of ​​the monitoring region based on the monitoring points;

[0079] Based on the area of ​​the monitoring area and the real-time wind speed, the carbon dioxide emission rate of the road segment corresponding to the monitoring point is obtained using the Gaussian plume model.

[0080] The carbon dioxide emission rates calculated from various monitoring points along the same road are summed to obtain the total carbon dioxide emission rate of the road.

[0081] The road carbon emissions are calculated based on the total carbon dioxide emission rate.

[0082] The carbon emission calculation method and apparatus based on dynamic simulation and mobile field measurement provided by this invention have at least the following beneficial effects:

[0083] (1) The carbon dioxide background concentration is taken into account in the calculation of road carbon emissions. The carbon dioxide increment is calculated based on the carbon dioxide background concentration, and the carbon emissions are calculated based on the carbon dioxide increment, so that the obtained carbon dioxide emissions are more accurate.

[0084] (2) The model is trained based on the data obtained from the mobile survey to obtain a background concentration simulation prediction model. The background concentration prediction model is used to predict the background concentration of carbon dioxide. It is applicable to various road scenarios and different traffic conditions. There is no need to build an observation station, which effectively reduces costs and has wide applicability.

[0085] (3) The background concentration simulation and prediction model takes into account the influence of wind speed, buildings and vegetation on the background concentration of carbon dioxide, which effectively improves the accuracy of the prediction of background concentration of carbon dioxide, and thus improves the accuracy of subsequent calculation of carbon dioxide emissions.

[0086] (4) Spatiotemporal matching and screening of carbon dioxide concentration data, panoramic image data and meteorological data used for model training are performed to effectively improve the accuracy of background concentration simulation prediction model. Attached Figure Description

[0087] Figure 1 This is a flowchart of one embodiment of the carbon emission calculation method based on dynamic simulation and mobile survey provided by the present invention.

[0088] Figure 2 This is a flowchart illustrating an embodiment of the data coupling screening method for carbon emission calculation based on dynamic simulation and mobile surveys provided by the present invention.

[0089] Figure 3 This is a flowchart of an embodiment of the carbon emission calculation method based on dynamic simulation and mobile survey provided by the present invention, which calculates turbulence intensity.

[0090] Figure 4 This is a flowchart of an embodiment of cluster analysis in the carbon emission calculation method based on dynamic simulation and mobile survey provided by the present invention.

[0091] Figure 5 This is a flowchart of one embodiment of the carbon emission calculation method based on dynamic simulation and mobile survey provided by the present invention.

[0092] Figure 6 This is a schematic diagram of one embodiment of the carbon emission calculation device based on dynamic simulation and mobile field measurement provided by the present invention. Detailed Implementation

[0093] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0094] refer to Figure 1 In some embodiments, a carbon emission calculation method based on dynamic simulation and mobile survey is provided, including:

[0095] S1. Pre-control the mobile vehicle to collect carbon dioxide concentration data, panoramic image data and meteorological data at road sampling points and perform multi-dimensional coupling and filtering to obtain carbon dioxide background concentration data.

[0096] S2. Based on the background carbon dioxide concentration data, train the pre-established deep learning model to obtain a background concentration simulation prediction model.

[0097] S3. Control the mobile vehicle to collect real-time carbon dioxide data, real-time panoramic image data and real-time meteorological data from road monitoring points, and identify and analyze the real-time panoramic image data to obtain environmental analysis results.

[0098] S4. Input the real-time meteorological data and environmental analysis results into the background concentration simulation and prediction model to obtain the predicted background carbon dioxide concentration at the monitoring point;

[0099] S5. Calculate the carbon dioxide increment at the monitoring point based on the real-time carbon dioxide data and the predicted background concentration of carbon dioxide at the monitoring point;

[0100] S6. Based on the Gaussian plume model, the carbon dioxide increment at multiple monitoring points along the road is calculated to obtain the road carbon emissions.

[0101] Specifically, in step S1, a mobile monitoring vehicle collects carbon dioxide concentration data, panoramic image data, and meteorological data from sampling points along the road. A vehicle-mounted mobile power supply is used in conjunction with a high-precision carbon dioxide concentration analyzer (e.g., Picarro2401) to monitor the carbon dioxide concentration at the road sampling points. This analyzer employs advanced cavity ring-down technology, achieving a high-precision measurement result of 0.1 ppm. Panoramic image data is collected via a panoramic camera. The vehicle-mounted meteorological instrument records atmospheric parameters such as temperature, humidity, pressure, wind speed, and wind direction during the vehicle's operation. This data provides information on the environment, weather, and traffic flow—variables closely related to carbon emissions.

[0102] Furthermore, in step S1, after collecting carbon dioxide concentration data and meteorological data from road sampling points, the following steps are also included:

[0103] The carbon dioxide concentration data and meteorological data were cleaned and standardized.

[0104] Specifically, the data cleaning process removes invalid data, such as sensor malfunctions, lost signals, or abnormal data points (data anomalies caused by vibration). By removing this invalid data, the reliability and accuracy of the calculation results can be improved.

[0105] Standardizing carbon dioxide concentration and meteorological data enables data from different sources to be analyzed and calculated under a unified standard.

[0106] Further, refer to Figure 2 In step S1, carbon dioxide concentration data, panoramic image data, and meteorological data are coupled and filtered in multiple dimensions, including:

[0107] S11. Time matching is performed based on the acquisition time of the carbon dioxide concentration data, panoramic image data, and meteorological data.

[0108] S12. The pre-trained image recognition model is used to identify the panoramic image data, and the non-traffic-affected area and its corresponding panoramic image acquisition time are determined based on the recognition results.

[0109] S13. Based on the meteorological data, calculate the turbulence intensity in the non-traffic-affected area at the corresponding panoramic image acquisition time;

[0110] S14. Obtain the road congestion index of the non-traffic-affected area at the corresponding panoramic image acquisition time, and calculate the background credibility score of the non-traffic-affected area at the corresponding panoramic image acquisition time based on the turbulence intensity and the road congestion index.

[0111] S15. Based on the background credibility score, filter the carbon dioxide concentration data of the non-traffic-affected area at the corresponding panoramic image acquisition time to obtain the filtered carbon dioxide concentration data.

[0112] S16. Perform cluster analysis on the screened carbon dioxide concentration data, and correlate the clustering results with the corresponding environmental information and meteorological data to obtain the background carbon dioxide concentration data.

[0113] Specifically, in step S11, carbon dioxide concentration data, panoramic image data, and meteorological data all have time tags. Time matching is performed based on the time tags of each data. That is, each carbon dioxide concentration data corresponds to panoramic image data and meteorological data collected at the same time, ensuring that the time points of each corresponding data are consistent, thus providing a basis for subsequent analysis.

[0114] Further, in step S12, a pre-trained image recognition model is used to identify the panoramic image data. The image recognition model can be a DeepLabV3+ model, and the recognition results include vehicles, vegetation, and building facades.

[0115] Based on the identification results, the non-traffic-affected areas and their corresponding panoramic image acquisition times and environmental information are determined, including:

[0116] Calculate the pixel percentage of the vehicle, vegetation, and building facade in the panoramic image respectively. Determine the sampling point corresponding to the panoramic image where the pixel percentage of the vehicle is less than the preset value as the non-traffic-affected area, and record the corresponding panoramic image acquisition time, the pixel percentage of the vegetation, and the pixel percentage of the building facade.

[0117] Specifically, the pixel ratio is calculated using the following formula:

[0118] (1)

[0119] in, Let A represent the i-th pixel of the vehicle / vegetation / building facade in the panoramic image, let A represent the pixel percentage of the vehicle / vegetation / building facade in the panoramic image, and let S represent the total number of pixels in the panoramic image.

[0120] Specifically, vehicles are the main source of carbon emissions, while buildings represent human activities. Therefore, both buildings and vegetation can be considered as factors affecting the background concentration of carbon dioxide. The sampling points corresponding to panoramic images where the pixel ratio of vehicles is less than a preset value are identified as non-traffic-affected areas. That is, at the corresponding time, there are few or no vehicles at this sampling point, and the carbon dioxide concentration obtained at this time is closer to the background concentration without emission sources.

[0121] Further, refer to Figure 3 In step S13, the meteorological data includes wind speed;

[0122] Based on the meteorological data, the turbulence intensity in the non-traffic-affected area at the corresponding panoramic image acquisition time is calculated, including:

[0123] S13a. Based on the time matching results, obtain the wind speed data of the non-traffic-affected area within a preset time period before the corresponding panoramic image acquisition time.

[0124] S13b: Calculate the standard deviation of wind speed within the preset time period based on the wind speed data;

[0125] S13c. The turbulence intensity is calculated based on the wind speed standard deviation and the wind speed in the non-traffic-affected area at the corresponding panoramic image acquisition time.

[0126] Specifically, wind speed is also an important factor affecting the concentration of carbon dioxide in the air. Based on wind speed data in non-traffic-affected areas, turbulence intensity is calculated for subsequent data screening.

[0127] The turbulence intensity is calculated using the following formula:

[0128] (2)

[0129] Where TI represents turbulence intensity, σ represents the standard deviation of wind speed within a preset time period, and v represents the wind speed in the non-traffic-affected area at the corresponding panoramic image acquisition time.

[0130] Further, in step S14, the road congestion index of the non-traffic-affected area at the corresponding panoramic image acquisition time is obtained. Based on the turbulence intensity and the road congestion index, the background confidence score of the non-traffic-affected area at the corresponding panoramic image acquisition time is calculated. This background confidence score is calculated using the following formula:

[0131] (3)

[0132] Where Wtime represents the background confidence score, P represents the road congestion index, TI represents the turbulence intensity, and Q represents the boundary layer stability coefficient of wind speed.

[0133] The boundary layer stability coefficient Q of wind speed can be solved using the Moning-Obukhoff similarity theory. In the atmospheric boundary layer, the boundary layer stability coefficient can be represented by the Moning-Obukhoff length.

[0134] Further, in step S15, the carbon dioxide concentration data of the non-traffic-affected area at the corresponding panoramic image acquisition time is filtered according to the background confidence score. Specifically, the carbon dioxide concentration data with a background confidence score greater than a preset score are selected to obtain the filtered carbon dioxide concentration data.

[0135] Further, refer to Figure 4 In step S16, the screened carbon dioxide data is subjected to cluster analysis. The clustering results are then correlated with corresponding environmental information and meteorological data to obtain the background carbon dioxide concentration data, including:

[0136] S16a. Perform Z-score standardization on the screened carbon dioxide data, and remove outlier data points based on the Z-score standardization results.

[0137] S16b: Calculate the coefficient of variation for the remaining screened carbon dioxide data after removing outliers, and select candidate carbon dioxide background concentration data based on the coefficient of variation.

[0138] S16c. The DBSCAN clustering algorithm is used to perform density analysis on the candidate carbon dioxide background concentration data, and the cluster center value with a density value greater than the preset density value is taken as the clustering result.

[0139] S16d. Based on the time matching results, the clustering results are associated with the corresponding meteorological data, the pixel ratio of vegetation, and the pixel ratio of building facades to obtain the carbon dioxide background concentration data.

[0140] Specifically, in step S16a, outliers are first removed by setting a time window and calculating the mean of the filtered carbon dioxide data according to a preset sliding time step. and standard deviation Furthermore, Z-score standardization is performed on the data points within the time window to remove outliers with Z scores greater than a preset score.

[0141] Further, in step S16b, the coefficient of variation (CV) is calculated for the remaining screened carbon dioxide data after removing outliers. The coefficient of variation (CV) is the ratio of the mean to the standard deviation. The coefficient of variation (CV) of the remaining screened carbon dioxide data in each time window is calculated. Window data with a coefficient of variation (CV) less than a preset value are retained, and a preset percentage of data is selected from the retained window data as the candidate carbon dioxide background concentration data for that window.

[0142] Further, in step S16c, the DBSCAN clustering algorithm is used to perform density analysis on the candidate carbon dioxide background concentration data, specifically including:

[0143] For each candidate carbon dioxide background concentration data point, based on the preset neighborhood radius, all data points within the neighborhood are searched, and based on the number of data points within the neighborhood, a core point is determined. Starting from the core point, each data point is added to the corresponding cluster in a recursive manner.

[0144] In the final clusters obtained, sparsely distributed clusters (which may be affected by instantaneous emissions) are removed, and the center values ​​of high-density clusters are retained as the final clustering result. The density determination value can be set according to the actual situation.

[0145] Finally, in step S16d, the obtained clustering results, i.e., the carbon dioxide concentration data identified as background concentration, are correlated with the pixel proportion of vegetation, the pixel proportion of building facades, and the wind speed of the corresponding panoramic image to obtain carbon dioxide background concentration data. That is, the carbon dioxide background concentration data includes the carbon dioxide concentration data identified as background concentration, the pixel proportion of vegetation, the pixel proportion of building facades, and the wind speed of the corresponding panoramic image. This carbon dioxide background concentration data will be used as a sample set to train the deep learning model.

[0146] Further, in step S2, a pre-established deep learning model is trained based on the background carbon dioxide concentration data. The deep learning model may include an LSTM network and a random forest spatial model, used for modeling time-series dependencies and handling high-dimensional nonlinear features, respectively. The LSTM network employs a network structure with 128 hidden units in the LSTM layer, a dropout layer with a dropout rate of 0.2, and a fully connected layer with 64 neurons. Temporal features are extracted using a 24-hour sliding window, and the model is trained using the Adam optimizer. The random forest model uses 200 decision trees, a maximum tree depth of 12 layers, a minimum leaf node sample size of 10, a feature sampling ratio of 0.8, and evaluates the importance of feature variables using the Gini index. Parameter optimization is performed using 5-fold cross-validation during model training.

[0147] LSTM networks employ an attention mechanism to adaptively weight meta-features, enabling dynamic fusion of different feature information. The dynamic weight adjustment mechanism utilizes a multi-dimensional adaptive strategy: calculating the predictive power of each model at different time scales based on temporal correlation; evaluating the model's applicability in different regions based on spatial distance; and dynamically adjusting model weights according to the magnitude of prediction error. When significant changes in meteorological elements are detected (e.g., a 24-hour change exceeding twice the historical standard deviation), the attention weights are adjusted, automatically reducing the LSTM model's contribution weight to 70% of its original value, while simultaneously increasing the influence of the random forest model in prediction.

[0148] Further, in step S3, the mobile monitoring vehicle is controlled to collect real-time carbon dioxide data, real-time panoramic image data, and real-time meteorological data from road monitoring points. The real-time panoramic image data is then analyzed to obtain environmental analysis results, including:

[0149] The real-time panoramic image data is input into the image recognition model for recognition and analysis to obtain the pixel ratio of vegetation and the pixel ratio of building facades at the monitoring points.

[0150] In step S4, the real-time meteorological data and environmental analysis results are input into the background concentration simulation prediction model. The background concentration simulation prediction model makes predictions based on the pixel ratio of vegetation and the pixel ratio of building facades in the real-time panoramic image data and meteorological data to obtain the predicted background concentration of carbon dioxide at the monitoring point, where the meteorological data is wind speed.

[0151] Further, in step S5, based on the real-time carbon dioxide data and the predicted background concentration of carbon dioxide at the monitoring point, the carbon dioxide increment at the monitoring point is calculated. The carbon dioxide increment at the monitoring point is the difference between the real-time carbon dioxide concentration and the predicted background concentration of carbon dioxide at the monitoring point, i.e.:

[0152] (4)

[0153] in, This is the real-time carbon dioxide concentration value. To predict background concentrations of carbon dioxide, This represents the increase in carbon dioxide.

[0154] Furthermore, the real-time meteorological data includes real-time wind speed.

[0155] refer to Figure 5 In step S6, the carbon dioxide increment at multiple monitoring points along the road is calculated based on the Gaussian plume model to obtain the road carbon emissions, including:

[0156] S61. Calculate the area of ​​the monitoring region based on the monitoring points;

[0157] S62. Based on the area of ​​the monitoring area and the real-time wind speed, the carbon dioxide emission rate of the road section corresponding to the monitoring point is obtained using the Gaussian plume model.

[0158] S63. The carbon dioxide emission rates calculated from each monitoring point on the same road are summed to obtain the total carbon dioxide emission rate of the road.

[0159] S64. Calculate the road carbon emissions based on the total carbon dioxide emission rate.

[0160] Specifically, in step S61, half the distance between every two monitoring points is used as the radius, and the area of ​​the circular region centered on the monitoring point is calculated as the monitoring area.

[0161] Further, in step S62, the carbon dioxide emission rate of the road segment corresponding to the monitoring point is obtained using the Gaussian plume model, and the specific calculation formula is as follows:

[0162] (5)

[0163] Where Aj represents the monitoring area of ​​the walking monitoring point, σ z Calculated using the Pasquill-Gifford empirical formula, where u is the real-time wind speed, H is the height of the emission source (i.e., the height of the vehicle's exhaust pipe), with a default value of 0.3 m, and Q... j Let J be the carbon dioxide emission rate of the j-th road segment corresponding to the monitoring point. This represents the increase in carbon dioxide.

[0164] In step S63, the total CO2 emission rate of the road is calculated using the following formula:

[0165] (6)

[0166] Where m represents the number of monitoring points on each road, Then, j represents the length of the road segment represented by each monitoring point j, and Qtotal represents the total carbon dioxide emission rate of the road.

[0167] In step S64, the total carbon dioxide emission rate of the road is multiplied by the time coefficient to obtain the road carbon emissions for the corresponding time period.

[0168] refer to Figure 6 In some embodiments, a carbon emission calculation device based on dynamic simulation and mobile survey is provided, comprising:

[0169] The sample acquisition module 201 is used to pre-control the mobile vehicle to collect carbon dioxide concentration data, panoramic image data and meteorological data at road sampling points and perform multi-dimensional coupling and filtering to obtain carbon dioxide background concentration data.

[0170] Training module 202 is used to train a pre-established deep learning model based on the carbon dioxide background concentration data to obtain a background concentration simulation prediction model.

[0171] The real-time data acquisition module 203 is used to control the mobile vehicle to collect real-time carbon dioxide data, real-time panoramic image data and real-time meteorological data from road monitoring points, and to identify and analyze the real-time panoramic image data to obtain environmental analysis results.

[0172] Prediction module 204 is used to input the real-time meteorological data and environmental analysis results into the background concentration simulation prediction model to obtain the predicted background concentration of carbon dioxide at the monitoring point;

[0173] The incremental calculation module 205 is used to calculate the carbon dioxide increment at the monitoring point based on the real-time carbon dioxide data and the predicted background concentration of carbon dioxide at the monitoring point.

[0174] The emission calculation module 206 is used to calculate the carbon dioxide increment at multiple monitoring points along the road based on the Gaussian plume model, thereby obtaining the road carbon emissions.

[0175] Furthermore, the sample acquisition module 201 is also used for:

[0176] The carbon dioxide concentration data and meteorological data were cleaned and standardized.

[0177] Furthermore, the sample acquisition module 201 performs multi-dimensional coupled filtering of carbon dioxide concentration data, panoramic image data, and meteorological data, including:

[0178] Time matching is performed based on the acquisition time of the carbon dioxide concentration data, panoramic image data, and meteorological data.

[0179] The panoramic image data is identified using a pre-trained image recognition model. Based on the recognition results, the non-traffic-affected area and its corresponding panoramic image acquisition time and environmental information are determined.

[0180] Based on the meteorological data, calculate the turbulence intensity in the non-traffic-affected area at the corresponding panoramic image acquisition time;

[0181] Obtain the road congestion index of the non-traffic-affected area at the corresponding panoramic image acquisition time, and calculate the background credibility score of the non-traffic-affected area at the corresponding panoramic image acquisition time based on the turbulence intensity and the road congestion index.

[0182] Based on the background credibility score, the carbon dioxide concentration data of the non-traffic-affected area at the corresponding panoramic image acquisition time are filtered to obtain filtered carbon dioxide concentration data.

[0183] The screened carbon dioxide concentration data is subjected to cluster analysis, and the clustering results are correlated with the corresponding environmental information and meteorological data to obtain the background carbon dioxide concentration data.

[0184] Furthermore, the identification results include vehicles, vegetation, and building facades;

[0185] The sample acquisition module 201 determines the non-traffic-affected area and its corresponding panoramic image acquisition time based on the recognition results, including:

[0186] Calculate the pixel percentage of the vehicle, vegetation, and building facade in the panoramic image respectively. Determine the sampling point corresponding to the panoramic image where the pixel percentage of the vehicle is less than the preset value as the non-traffic-affected area, and record the corresponding panoramic image acquisition time, the pixel percentage of the vegetation, and the pixel percentage of the building facade.

[0187] Furthermore, the meteorological data includes wind speed;

[0188] The sample acquisition module 201 calculates the turbulence intensity in the non-traffic-affected area at the corresponding panoramic image acquisition time based on the meteorological data, including:

[0189] Based on the time matching results, obtain the wind speed data of the non-traffic-affected area within a preset time period before the corresponding panoramic image acquisition time;

[0190] Calculate the standard deviation of wind speed within the preset time period based on the wind speed data;

[0191] The turbulence intensity is calculated based on the wind speed standard deviation and the wind speed in the non-traffic-affected area at the corresponding panoramic image acquisition time.

[0192] Furthermore, the sample acquisition module 201 performs cluster analysis on the screened carbon dioxide data, and correlates the clustering results with corresponding environmental information and meteorological data to obtain the background carbon dioxide concentration data, including:

[0193] The carbon dioxide data was Z-score standardized, and outlier data points were removed based on the Z-score standardization results.

[0194] Calculate the coefficient of variation for the remaining screened carbon dioxide data after removing outliers, and select candidate carbon dioxide background concentration data based on the coefficient of variation.

[0195] The DBSCAN clustering algorithm was used to perform density analysis on the candidate carbon dioxide background concentration data, and the cluster center value with a density value greater than the preset density value was taken as the clustering result.

[0196] Based on the time matching results, the clustering results are correlated with the corresponding meteorological data, the pixel proportion of vegetation, and the pixel proportion of building facades to obtain the background carbon dioxide concentration data.

[0197] Furthermore, the real-time data acquisition module 203 identifies and analyzes the real-time panoramic image data to obtain environmental analysis results, including:

[0198] The real-time panoramic image data is input into the image recognition model for recognition and analysis to obtain the pixel ratio of vegetation and the pixel ratio of building facades at the monitoring points.

[0199] Furthermore, the carbon dioxide increment at the monitoring point is the difference between the real-time carbon dioxide concentration and the predicted background carbon dioxide concentration at the monitoring point.

[0200] Furthermore, the real-time meteorological data includes real-time wind speed;

[0201] The emission calculation module 206 calculates the carbon dioxide increments at multiple monitoring points along the road based on a Gaussian plume model to obtain the road carbon emissions, including:

[0202] Calculate the area of ​​the monitoring region based on the monitoring points;

[0203] Based on the area of ​​the monitoring area and the real-time wind speed, the carbon dioxide emission rate of the road segment corresponding to the monitoring point is obtained using the Gaussian plume model.

[0204] The carbon dioxide emission rates calculated from various monitoring points along the same road are summed to obtain the total carbon dioxide emission rate of the road.

[0205] The road carbon emissions are calculated based on the total carbon dioxide emission rate.

[0206] The carbon emission calculation method and apparatus based on dynamic simulation and mobile survey provided in the above embodiments have at least the following beneficial effects:

[0207] (1) The carbon dioxide background concentration is taken into account in the calculation of road carbon emissions. The carbon dioxide increment is calculated based on the carbon dioxide background concentration, and the carbon emissions are calculated based on the carbon dioxide increment, so that the obtained carbon dioxide emissions are more accurate.

[0208] (2) The model is trained based on the data obtained from the mobile survey to obtain a background concentration simulation prediction model. The background concentration prediction model is used to predict the background concentration of carbon dioxide. It is applicable to various road scenarios and different traffic conditions. There is no need to build an observation station, which effectively reduces costs and has wide applicability.

[0209] (3) The background concentration simulation and prediction model takes into account the influence of wind speed, buildings and vegetation on the background concentration of carbon dioxide, which effectively improves the accuracy of the prediction of background concentration of carbon dioxide, and thus improves the accuracy of subsequent calculation of carbon dioxide emissions.

[0210] (4) Spatiotemporal matching and screening of carbon dioxide concentration data, panoramic image data and meteorological data used for model training are performed to effectively improve the accuracy of background concentration simulation prediction model.

[0211] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention. Clearly, those skilled in the art can make various alterations and modifications to the invention without departing from its spirit and scope. Thus, if these modifications and modifications of the invention fall within the scope of the claims and their equivalents, the invention is also intended to include these modifications and modifications.

Claims

1. A carbon emission calculation method based on dynamic simulation and mobile survey, characterized in that, include: The process involves: Pre-controlling mobile monitoring vehicles to collect carbon dioxide concentration data, panoramic image data, and meteorological data from road sampling points, performing multi-dimensional coupling and filtering to obtain background carbon dioxide concentration data; time matching based on the collection times of the carbon dioxide concentration data, panoramic image data, and meteorological data; using a pre-trained image recognition model to identify the panoramic image data, determining the non-traffic-affected area and its corresponding panoramic image collection time and environmental information based on the identification results; calculating the turbulence intensity of the non-traffic-affected area at the corresponding panoramic image collection time based on the meteorological data; obtaining the road congestion index of the non-traffic-affected area at the corresponding panoramic image collection time, and calculating the background credibility score of the non-traffic-affected area at the corresponding panoramic image collection time based on the turbulence intensity and road congestion index; filtering the carbon dioxide concentration data of the non-traffic-affected area at the corresponding panoramic image collection time based on the background credibility score to obtain filtered carbon dioxide concentration data; performing cluster analysis on the filtered carbon dioxide concentration data, and correlating the clustering results with the corresponding environmental information and meteorological data to obtain the background carbon dioxide concentration data. The pre-established deep learning model is trained based on the background carbon dioxide concentration data to obtain a background concentration simulation prediction model. The mobile monitoring vehicle collects real-time carbon dioxide data, real-time panoramic image data, and real-time meteorological data from road monitoring points. The real-time panoramic image data is then identified and analyzed to obtain environmental analysis results. The real-time meteorological data and environmental analysis results are input into the background concentration simulation and prediction model to obtain the predicted background carbon dioxide concentration at the monitoring point. The carbon dioxide increment at the monitoring point is calculated based on the real-time carbon dioxide data and the predicted background concentration of carbon dioxide at the monitoring point. The carbon dioxide increment at multiple monitoring points along the road is calculated based on the Gaussian plume model to obtain the road carbon emissions.

2. The method according to claim 1, characterized in that, After collecting carbon dioxide concentration data and meteorological data from road sampling points, the process also includes: The carbon dioxide concentration data and meteorological data were cleaned and standardized.

3. The method according to claim 1, characterized in that, The identification results include vehicles, vegetation, and building facades; Based on the identification results, the non-traffic-affected areas and their corresponding panoramic image acquisition times and environmental information are determined, including: Calculate the pixel percentage of the vehicle, vegetation, and building facade in the panoramic image respectively. Determine the sampling point corresponding to the panoramic image where the pixel percentage of the vehicle is less than the preset value as the non-traffic-affected area, and record the corresponding panoramic image acquisition time, the pixel percentage of the vegetation, and the pixel percentage of the building facade.

4. The method according to claim 1, characterized in that, The meteorological data includes wind speed; Based on the meteorological data, the turbulence intensity in the non-traffic-affected area at the corresponding panoramic image acquisition time is calculated, including: Based on the time matching results, obtain the wind speed data of the non-traffic-affected area within a preset time period before the corresponding panoramic image acquisition time; Calculate the standard deviation of wind speed within the preset time period based on the wind speed data; The turbulence intensity is calculated based on the wind speed standard deviation and the wind speed in the non-traffic-affected area at the corresponding panoramic image acquisition time.

5. The method according to claim 3, characterized in that, The screened carbon dioxide concentration data is subjected to cluster analysis. Based on the clustering results, it is correlated with corresponding environmental information and meteorological data to obtain the background carbon dioxide concentration data, including: The carbon dioxide data was Z-score standardized, and outlier data points were removed based on the Z-score standardization results. Calculate the coefficient of variation for the remaining screened carbon dioxide data after removing outliers, and select candidate carbon dioxide background concentration data based on the coefficient of variation. The DBSCAN clustering algorithm was used to perform density analysis on the candidate carbon dioxide background concentration data, and the cluster center value with a density value greater than the preset density value was taken as the clustering result. Based on the time matching results, the clustering results are correlated with the corresponding meteorological data, the pixel proportion of vegetation, and the pixel proportion of building facades to obtain the background carbon dioxide concentration data.

6. The method according to claim 3, characterized in that, The real-time panoramic image data is identified and analyzed to obtain environmental analysis results, including: The real-time panoramic image data is input into the image recognition model for recognition and analysis to obtain the pixel ratio of vegetation and the pixel ratio of building facades at the monitoring points.

7. The method according to claim 1, characterized in that, The carbon dioxide increment at the monitoring point is the difference between the real-time carbon dioxide concentration and the predicted background carbon dioxide concentration at the monitoring point.

8. The method according to claim 1, characterized in that, The real-time meteorological data includes real-time wind speed; Based on the Gaussian plume model, the spatiotemporal integration of carbon dioxide increments at multiple monitoring points along the road is used to obtain road carbon emissions, including: Calculate the area of ​​the monitoring region based on the monitoring points; Based on the area of ​​the monitoring area and the real-time wind speed, the carbon dioxide emission rate of the road segment corresponding to the monitoring point is obtained using the Gaussian plume model. The carbon dioxide emission rates calculated from various monitoring points along the same road are summed to obtain the total carbon dioxide emission rate of the road. The road carbon emissions are calculated based on the total carbon dioxide emission rate.

9. A carbon emission calculation device based on dynamic simulation and mobile field measurement, characterized in that, include: The sample acquisition module is used to pre-control mobile vehicles to collect carbon dioxide concentration data, panoramic image data, and meteorological data from road sampling points, and perform multi-dimensional coupling and filtering to obtain carbon dioxide background concentration data. This involves: time matching based on the acquisition time of the carbon dioxide concentration data, panoramic image data, and meteorological data; using a pre-trained image recognition model to identify the panoramic image data; determining the non-traffic-affected area and its corresponding panoramic image acquisition time and environmental information based on the identification results; calculating the turbulence intensity of the non-traffic-affected area at the corresponding panoramic image acquisition time based on the meteorological data; obtaining the road congestion index of the non-traffic-affected area at the corresponding panoramic image acquisition time; calculating the background credibility score of the non-traffic-affected area at the corresponding panoramic image acquisition time based on the turbulence intensity and road congestion index; filtering the carbon dioxide concentration data of the non-traffic-affected area at the corresponding panoramic image acquisition time based on the background credibility score to obtain filtered carbon dioxide concentration data; and performing cluster analysis on the filtered carbon dioxide concentration data, and associating the clustering results with the corresponding environmental information and meteorological data to obtain the carbon dioxide background concentration data. The training module is used to train a pre-established deep learning model based on the background carbon dioxide concentration data to obtain a background concentration simulation prediction model. The real-time data acquisition module is used to control the mobile vehicle to collect real-time carbon dioxide data, real-time panoramic image data and real-time meteorological data from road monitoring points, and input them into the background concentration simulation and prediction model to obtain the predicted background carbon dioxide concentration at the monitoring points. The incremental calculation module is used to calculate the carbon dioxide increment at the monitoring point based on the real-time carbon dioxide data and the predicted background concentration of carbon dioxide at the monitoring point. The emissions calculation module is used to calculate the carbon dioxide increment at multiple monitoring points along the road based on a Gaussian plume model, thereby obtaining the road carbon emissions.

Images