Carbon emission rights confirmation and labeling method based on spatiotemporal carbon map

Abstract

The present invention discloses a carbon emission confirmation and labeling method based on a spatiotemporal carbon map, characterized in that the method comprises the following steps: (1) determining the starting point of the carbon confirmation investigation based on the spatiotemporal carbon map; (2) conducting drone detection, and the drone automatically plans the detection path; (3) the drone autonomously determines the carbon emission source through a sensor; (4) combining the spatiotemporal carbon map, determining the carbon emission subject information. The present invention can be combined with the existing spatiotemporal carbon map to further enhance the practicality of the spatiotemporal carbon map; there is no need to install specific monitoring equipment at the carbon emission site, which is convenient for third parties to carry out carbon auditing; the drone automatically plans the detection path and autonomously confirms the carbon emission source without manual intervention, and the method is easy to deploy and implement.

Description

Technical Field

[0001] The present invention belongs to the fields of carbon neutrality, carbon auditing, etc., and specifically relates to a carbon emission rights confirmation and labeling method based on spatiotemporal carbon maps. Background Art

[0002] Global climate change is a severe challenge facing the sustainable development of mankind. The spatiotemporal carbon map technology that counts and displays carbon emissions has become a current research hotspot. For example, in 2021, Liu Zhu's research group at Tsinghua University released the world's first near-real-time carbon emissions map, providing real-time key geographic location information of carbon emissions. Nanjing Audit University has successfully developed a spatiotemporal carbon map system mainly based on the field of carbon auditing. The system contains important data such as carbon emission monitoring and geographic location information.

[0003] However, the existing spatiotemporal carbon map makes it difficult to accurately confirm the ownership of carbon emissions. Carbon auditing can audit the carbon quota verification, carbon quota allocation and carbon quota compliance issues in the process of carbon emission rights trading of enterprises, which will help enterprises innovate green technologies, reduce the total amount of carbon emissions and strictly fulfill their quota payment obligations, promote the effective improvement of enterprises' awareness and ability to reduce emissions, and promote the sustainable and healthy development of the carbon emission rights trading market. The first step to achieve carbon auditing is to determine the main information of carbon emissions, that is, carbon confirmation.

[0004] With respect to spatiotemporal carbon maps, existing research has not proposed an effective solution for carbon rights confirmation. Summary of the invention

[0005] Based on the problems existing in the background technology, the present invention proposes a carbon emission rights identification and labeling method which is easy to deploy and implement, does not require the installation of specific monitoring equipment at the carbon emission site, and is convenient for third parties to conduct carbon audits.

[0006] It includes the following steps:

[0007] (1) Determine the starting point of the carbon emission rights confirmation survey based on the spatiotemporal carbon map;

[0008] (2) Carry out drone detection, and the drone automatically plans the detection path;

[0009] (3) UAVs autonomously determine carbon emission sources through sensors;

[0010] (4) Combine the spatiotemporal carbon map to determine the main carbon emission information.

[0011] The details are as follows:

[0012] (1) Determine the starting point of the carbon titling survey based on the spatiotemporal carbon map.

[0013] In step (1), the starting point of the carbon rights confirmation investigation can be determined based on the corresponding geographical coordinates in the spatiotemporal carbon map. The auditor can also specify the geographical coordinates as the starting point of the investigation.

[0014] (2) Carry out drone detection, and the drone automatically plans the detection path.

[0015] The step (2) specifically comprises the following steps:

[0016] (2-1) Configure the geographic coordinates corresponding to the survey starting point in the drone;

[0017] (2-2) The drone flies to the survey starting point;

[0018] (2-3) Drones conduct carbon emission detection and automatically plan detection routes.

[0019] In step (2-1), the configured geographic coordinates may be a geographic coordinate system, an ECEF coordinate system, a local NED coordinate system, etc.

[0020] In step (2-3), the drone can use a gas monitor to detect carbon emissions at the current location.

[0021] In step (2-3), the UAV detection path planning process is as described in steps (2-3-1) to (2-3-5).

[0022] (2-3-1) The drone detects carbon emissions from its current location as the starting point S. The detected carbon emissions are recorded as c1;

[0023] (2-3-2) The drone rotates α degrees to detect the carbon emissions at the current location;

[0024] (2-3-3) Repeat step (2-3-2) until the drone returns to the starting point S;

[0025] (2-3-4) The drone compares the carbon emissions c1, c2, …, c n , take the direction corresponding to the maximum value as the flight direction;

[0026] (2-3-5) The flight distance is set to β. After the drone flies β distance, execute step (3).

[0027] In step (2-3-2), the drone can be set to rotate clockwise or counterclockwise, but the rotation direction should remain unchanged after being set.

[0028] In step (2-3-4),

[0029] In step (2-3-5), β can be set to different values ​​according to the actual situation. For example, when the wind is strong, it can be set to a smaller value, such as 100 meters; when the weather is clear, it can be set to a larger value, such as 500 meters, etc.

[0030] (3) Drones autonomously determine the source of carbon emissions.

[0031] The step (3) specifically comprises the following steps:

[0032] (3-1) For the ground, the drone starts video shooting;

[0033] (3-2) The drone performs video recognition and automatically identifies the source of carbon emissions when shooting video;

[0034] (3-3) If the carbon emission source point is identified successfully, execute step (4); otherwise, execute step (2).

[0035] In step (3-1), when the drone shoots ground video, it can rotate 360 ​​degrees to shoot a panoramic view of the ground.

[0036] In step (3-2), the drone can automatically locate the carbon emission source through deep learning image classification method. The specific steps can be shown as (3-2-1) to (3-2-6).

[0037] (3-2-1) Collect smoke emission images and other types of images;

[0038] (3-2-2) The collected images are used as training samples to train deep learning models such as ResNet. The trained deep learning models such as ResNet are binary classification models, that is, the output result is "yes" or "no". If it is "yes", the current image is a smoke emission image, otherwise it is another type of image;

[0039] (3-2-3) Deploy the trained binary classification model to the drone through programming;

[0040] (3-2-4) When the drone shoots ground video, the video is extracted frame by frame as images;

[0041] (3-2-5) Use the extracted images as input to the classification model for classification;

[0042] (3-2-6) If the classification result is "yes", the carbon emission source point is identified successfully; otherwise, continue to execute step (3-2-5) until all images are classified.

[0043] (4) Combine the spatiotemporal carbon map to determine the main carbon emission information.

[0044] The step (4) specifically comprises the following steps:

[0045] (4-1) The drone flies to the identified carbon emission source;

[0046] (4-2) The drone sends the current geographic location to the background spatiotemporal carbon map server;

[0047] (4-3) Based on the received geographic location information, the server queries the “location-ownership” correspondence in the spatiotemporal carbon map to determine the person responsible for carbon emissions, and then completes the carbon emission rights confirmation and marking in the spatiotemporal carbon map.

[0048] In step (4-3), the “location-ownership” correspondence in the spatiotemporal carbon map can be determined by national land survey data.

[0049] In step (4-3), the carbon emission rights confirmation and labeling can be completed by associating the carbon emission responsible person information with the carbon emission monitoring data and drone monitoring data in the spatiotemporal carbon map.

[0050] Beneficial Effects of the Invention

[0051] (1) It can be combined with existing spatiotemporal carbon maps to further enhance the practicality of spatiotemporal carbon maps;

[0052] (2) There is no need to install specific monitoring equipment at the carbon emission site, making it easier for third parties to conduct carbon audits;

[0053] (3) The drone automatically plans the detection path, autonomously identifies the source of carbon emissions, and monitors carbon emissions without the need for human intervention. The method is easy to deploy and implement. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 The overall execution flow chart of the method is given.

[0055] Figure 2 This is an example diagram of the spatiotemporal carbon map system.

[0056] Figure 3 Sample instance for ResNet model training.

[0057] Figure 4 This is an example diagram of the “position-ownership” correspondence.

[0058] Figure 5 An example diagram of carbon titling is given. DETAILED DESCRIPTION

[0059] The present invention will be further described below in conjunction with embodiments, but the protection scope of the present invention is not limited thereto:

[0060] The overall operation process of the method is as follows Figure 1 The “Carbon Trace Auditable” spatiotemporal carbon map system developed by Nanjing Audit University is shown in Figure 2 As shown. Figure 2 In the system shown, the carbon emissions in area A are the highest. Assuming that the carbon emissions in area A need to be confirmed, refer to Figure 1 The execution steps are as follows:

[0061] (1) Determine the starting point of the survey. According to the “Carbon Trace Auditable” spatiotemporal carbon map system, the longitude and latitude coordinates of the starting point of the survey are set to (118.614913, 32.068569).

[0062] (2) The drone conducts carbon emission detection and autonomously determines the source of carbon emissions. The drone is set to rotate 45 degrees. At the starting point, the drone measures 8 carbon content values, including c1=0.03%, c2=0.06%, c3=0.06%, c4=0.08%, c5=0.05%, c6=0.04%, c7=0.03%, and c8=0.03%. Among them, c4 has the largest value, and the drone continues to fly 200 meters in the direction of c4.

[0063] (3) The drone automatically identifies the carbon emission source. The ResNet model is used to classify smoke emission images. The training sample examples of the ResNet model are as follows: Figure 3 The core structure construction code of the ResNet model is shown below.

[0064] self.in_channel = 64

[0065] self.conv1=nn.Conv2d(3, self.in_channel, kernel_size=7, stride=2, padding=3, bias=False)

[0066] self.bn1=nn.BatchNorm2d(self.in_channel)

[0067] self.relu=nn.ReLU(inplace=True)

[0068] self.maxpool=nn.MaxPool2d(kernel_size=3, stride=2, padding=1)

[0069] self.layer1=self._make_layer(block,64,blocks_num[0])

[0070] self.layer2=self._make_layer(block,128,blocks_num[1],stride=2)

[0071] self.layer3=self._make_layer(block,256,blocks_num[2],stride=2)

[0072] self.layer4=self._make_layer(block,512,blocks_num[3],stride=2)

[0073] (4) The drone successfully identifies the carbon emission source and sends the geographic coordinates of the corresponding location (118.613872, 32.045606) to the "Carbon Trace Auditable" spatiotemporal carbon map system server. After receiving the geographic coordinates, the server determines the "location-ownership" relationship based on the land survey data. An example of the determined "location-ownership" relationship is as follows: Figure 4 shown.

[0074] Based on the "location-ownership" relationship, the "Carbon Trace Auditable" spatiotemporal carbon map system server associates the carbon emissions of place A with the corresponding responsible persons, thereby completing the carbon emission rights confirmation and labeling. Figure 5 shown.

[0075] It should be stated that the content and specific implementation methods of the present invention are intended to demonstrate the practical application of the technical solution provided by the present invention and should not be interpreted as limiting the scope of protection of the present invention. Those skilled in the art may make various modifications, equivalent substitutions, or improvements inspired by the spirit and principles of the present invention. However, these changes or modifications are within the scope of protection of the application to be approved.

Claims

1. A carbon emission rights confirmation and labeling method based on spatiotemporal carbon map, characterized in that: The method comprises the following steps: (1) Determine the starting point of the carbon emission rights confirmation survey based on the spatiotemporal carbon map; (2) Carrying out drone detection, and the drone automatically plans a detection path; the step (2) comprises the following steps: (2-1) Configure the geographic coordinates corresponding to the survey starting point in the drone; (2-2) The drone flies to the survey starting point; (2-3) The drone conducts carbon emission detection and automatically plans the detection path; the drone detection path planning process is as described in steps (2-3-1) to (2-3-5): (2-3-1) The drone takes the current position as the starting point S , detect carbon emissions, and the detected carbon emissions are recorded as c 1 ; (2-3-2) Drone rotation Degree, detect the carbon emissions at the current location; (2-3-3) Repeat step (2-3-2) until the drone returns to the starting point S ; (2-3-4) Comparison of carbon emissions detected by drones c 1 , c 2 , …, c i , c n , take the direction corresponding to the maximum value as the flight direction; c i Indicates that the drone i The amount of carbon emissions detected; (2-3-5) The flight distance is set to , drone flight After the distance is determined, execute step (3); (3) The drone autonomously determines the carbon emission source through sensors; step (3) comprises the following steps: (3-1) For the ground, the drone starts video shooting and sensors; (3-2) The drone performs video recognition and automatically identifies the carbon emission source and carbon emission amount when shooting the video; in step (3-2), the drone automatically identifies the carbon emission source through deep learning image classification method; the specific steps are shown in (3-2-1) to (3-2-6): (3-2-1) Manually collect smoke emission images and other types of images; (3-2-2) Use the collected images as training samples to train the ResNet deep learning model; The trained ResNet deep learning model is a binary classification model, that is, the output result is "yes" or "no"; if it is "yes", the current picture is a smoke emission picture, otherwise it is another type of picture; (3-2-3) Deploy the trained binary classification model to the drone programmatically; (3-2-4) When the drone shoots ground video, the video is extracted frame by frame as images; (3-2-5) Use the extracted images as input to the classification model for classification; (3-2-6) If the classification result is "yes", the carbon emission source point is identified successfully; otherwise, continue to execute step (3-2-5) until all images are classified; (3-3) If the carbon emission source is successfully identified, proceed to step (4); otherwise, proceed to step (2); (4) Determine the carbon emission subject information in combination with the spatiotemporal carbon map; the step (4) comprises the following steps: (4-1) The drone flies to the identified carbon emission source; (4-2) The drone sends its current geographic location and carbon emissions to the background spatiotemporal carbon map server; (4-3) Based on the received geographic location information, the server queries the "location-ownership" correspondence in the spatiotemporal carbon map to determine the person responsible for carbon emissions, and then completes the carbon emission rights confirmation and marking in the spatiotemporal carbon map.

Images