Road area vegetation carbon sink assessment method and system based on Internet of Things, and storage medium

Abstract

The invention discloses a road area vegetation carbon sink assessment method and system based on the Internet of Things, and a storage medium, relates to the technical field of carbon sink monitoring, and solves the technical problem that the carbon sink assessment precision is low due to the fact that the prior art often depends on a single data source and lacks consideration of the influence of dynamic factors such as traffic interference and soil parameters. Multi-source fusion data is generated based on remote sensing data, sensor data and traffic data; inputting the multi-source fusion data corresponding to the region ID into a vegetation classification model to obtain vegetation result data corresponding to the region ID; generating a regional carbon sink total amount based on the vegetation result data; an alarm signal and an optimization scheme are generated based on the total amount of regional carbon sinks, multiple data sources are considered in real time, fusion weight coefficients are dynamically distributed for all the data sources, multi-source fusion data are more accurate, accurate vegetation result data are obtained through a pre-trained vegetation classification model, accurate data support is provided for carbon sink evaluation, and the carbon sink evaluation efficiency is improved. And the accuracy of the carbon sink evaluation method is improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of carbon sink monitoring, and in particular to a road area vegetation carbon sink evaluation method and system based on the Internet of Things and a storage medium. BACKGROUND

[0002] Under the background of global climate change intensification and increasing attention to ecological environment, the evaluation and improvement of carbon sink capacity have become an important issue of ecological protection and sustainable development. Road area vegetation, as an important connecting body between urban and natural ecosystems, its carbon sink function has gradually attracted attention in recent years. The so-called "road area vegetation carbon sink evaluation" refers to the quantitative analysis and evaluation of the ability of vegetation along the road to absorb and fix carbon dioxide in the atmosphere through photosynthesis within a certain period of time. This evaluation not only helps to fully understand the role of road greening in regional carbon cycle, but also provides a scientific basis for the improvement of the ecological benefits of transportation infrastructure. It has important practical significance and strategic value for optimizing road greening structure, improving ecological environment quality, and promoting low-carbon development in the transportation field.

[0003] The prior art often relies on a single data source when evaluating carbon sinks, and lacks consideration of the influence of dynamic factors such as traffic interference and soil parameters, resulting in low accuracy of carbon sink evaluation. Therefore, the road area vegetation carbon sink evaluation method still needs to be further improved. SUMMARY

[0004] The present application aims to at least solve one of the technical problems existing in the prior art. To this end, the present application proposes a road area vegetation carbon sink evaluation method and system based on the Internet of Things and a storage medium, which is used to solve the technical problem that the prior art often relies on a single data source, lacks consideration of the influence of dynamic factors such as traffic interference and soil parameters, and results in low accuracy of carbon sink evaluation.

[0005] To achieve the above-mentioned purpose, the first aspect of the present application provides a road area vegetation carbon sink evaluation method based on the Internet of Things, comprising: Obtaining a plurality of regional data of the road area; the regional data includes regional ID and its corresponding remote sensing data, sensor data and traffic data; the remote sensing data includes multispectral image and laser radar point cloud data; the sensor data includes soil data, environmental data and vegetation data; Generating multi-source fusion data based on remote sensing data, sensor data and traffic data; Inputting the multi-source fusion data corresponding to the regional ID into a vegetation classification model to obtain vegetation result data corresponding to the regional ID; the vegetation classification model is constructed by a machine learning model, and is used to evaluate the vegetation classification result in the regional ID; Generating regional carbon sink total based on the vegetation result data; Generating an alarm signal and an optimization scheme based on the regional carbon sink total.

[0006] This application comprehensively processes multiple data sources through the above steps and dynamically assigns fusion weight coefficients to each data source, thereby significantly improving the accuracy of multi-source data fusion. On this basis, combined with a pre-trained vegetation classification model, it further obtains high-precision vegetation result data, providing reliable data support for carbon sink assessment and effectively improving the accuracy and scientific nature of carbon sink assessment methods.

[0007] Furthermore, the generation of multi-source fusion data based on remote sensing data, sensor data, and traffic data includes: The remote sensing data includes multispectral images and lidar point cloud data; Several remote sensing parameters corresponding to multispectral images and lidar point cloud data are extracted; the remote sensing parameters include NDVI index, chlorophyll fluorescence, canopy height and LAI index. The sensor data includes soil data, environmental data, and vegetation data; Extract several sensor parameters corresponding to soil data, environmental data, and vegetation data; the sensor parameters include soil temperature, soil moisture, soil CO2 flux, light intensity, air temperature, precipitation data, and diameter at breast height (DBH) increment; Extract several traffic parameters corresponding to the traffic data; the traffic parameters include traffic flow and vehicle type. Remote sensing parameters, sensor parameters, and traffic parameters are combined using corresponding fusion weighting coefficients to obtain multi-source fused data.

[0008] Furthermore, the fusion weight coefficients are obtained through the following methods: Acquire remote sensing parameters, sensor parameters, and traffic parameters; The stability weights WQ corresponding to remote sensing parameters, sensor parameters, and traffic parameters are calculated using the entropy weight method. The entropy weight method determines the stability weight of each parameter by measuring the data stability of several parameters. The smaller the entropy value of a parameter, the higher its data stability, and the greater its corresponding reliability weight. The reliability weights KQ corresponding to remote sensing parameters, sensor parameters, and traffic parameters are calculated using the coefficient of variation method. The coefficient of variation method determines the reliability weight of each parameter by measuring the reliability of several parameters. The reliability weights are assigned based on the ratio between the standard deviation and the mean of each parameter. Through formula Calculate the fusion weight coefficients corresponding to remote sensing parameters, sensor parameters, and traffic parameters; where n represents the parameter number among the remote sensing parameters, sensor parameters, and traffic parameters; This is expressed as a proportionality coefficient. ∈(0,1).

[0009] Furthermore, the vegetation classification model is constructed using a machine learning model, including: Acquire several historical multi-source fusion data and their corresponding historical vegetation results data; Several historical multi-source fusion data and their corresponding historical vegetation results data are divided into training data, validation data and test data; and the training data, validation data and test data are preprocessed to obtain training set, validation set and test set. Choose a machine learning model as the base model; The base model is trained on the training set, and the learning rate and other hyperparameters are adjusted on the validation set to obtain the pre-trained model. By validating the pre-trained model on the test set, a vegetation classification model was finally obtained, which takes multi-source fusion data as input and outputs vegetation result data as output.

[0010] Furthermore, the generation of the regional carbon aggregate based on vegetation outcome data includes: Extract the vegetation type, its corresponding vegetation percentage (ZB), and vegetation influence coefficient from the vegetation results data; Calculate the carbon content of vegetation type based on vegetation impact coefficient; The total regional carbon content is calculated using a formula; the formula satisfies: Where j represents the vegetation type number within region ID; J represents the total number of vegetation types within region ID. It represents the biomass per unit area corresponding to the j-th vegetation type within region ID, and is a basic physical quantity reflecting the carbon sequestration capacity of the vegetation type. This represents the carbon content of the vegetation type corresponding to the j-th vegetation type within the region ID.

[0011] This application adjusts the carbon content of the corresponding vegetation type within the region ID by adjusting the vegetation influence coefficient in the vegetation result data, and accurately generates the total carbon amount in the region by combining the vegetation proportion and the biomass per unit area of ​​vegetation, making the calculation of the actual total carbon amount in the region more accurate, and providing accurate data support for the generation of subsequent alarm signals and optimization schemes.

[0012] Furthermore, the calculation of vegetation type carbon content based on vegetation influence coefficient includes: Obtain the vegetation type and its corresponding vegetation influence coefficient ZYX; Determine whether the vegetation impact coefficient is greater than the impact threshold; Yes, obtain the traffic flow change rate. And the self-recovery rate of vegetation type ZLZHV; When the rate of change in traffic flow is greater than or equal to the change threshold, the formula is used. Calculate the carbon content ZLC of vegetation type; where, Represented as the basic carbon content of vegetation type; the change threshold is set based on experience; When the rate of change in traffic flow is less than the change threshold, the carbon content of vegetation type... ; No, vegetation type carbon content .

[0013] Furthermore, the generation of alarm signals and optimization schemes based on the total regional carbon emissions includes: Get the total carbon sequestration and preset carbon sink for the region corresponding to the region ID; When the total regional carbon sequestration is less than or equal to A times the preset regional carbon sequestration, a severe carbon sequestration deviation alarm signal is generated. When the total regional carbon amount is greater than A times the regional preset carbon sink amount and the total regional carbon amount is less than or equal to B times the regional preset carbon sink amount, a carbon sink moderate deviation alarm signal is generated. When the total regional carbon sink exceeds C times the preset regional carbon sink amount, a positive carbon sink change indication signal is generated; where A, B, and C are proportionality coefficients, A and B ∈ (0, 1), C > 1; and A <B; The alarm signals include a severe carbon sink deviation alarm signal, a moderate carbon sink deviation alarm signal, and a positive carbon sink change indication signal; An optimized scheme for generating alarm signals.

[0014] Furthermore, the alarm signal generation optimization scheme includes: Obtain the region ID and its corresponding vegetation type, vegetation percentage, soil carbon background, annual increase in vegetation diameter at breast height, traffic flow, ambient temperature, ambient humidity, total regional carbon sequestration, and preset regional carbon sink; the soil carbon background is represented as the organic carbon density of the 0-100cm soil layer. The data for the scheme analysis is spliced ​​together from the vegetation type, vegetation ratio, soil carbon background, annual increase in vegetation diameter at breast height, traffic flow, environmental humidity, total regional carbon and the pre-set regional carbon sink within the region ID. Input the alarm signal and scheme analysis data corresponding to the region ID into the scheme suggestion model to obtain the optimized scheme corresponding to the alarm signal within the region ID; The proposed solution suggests that the large model be constructed in the following ways: Acquire several historical alarm signals and historical scheme analysis data, as well as the historical optimization schemes corresponding to the historical alarm signals and historical scheme analysis data; Several historical alarm signals and historical scheme analysis data, as well as the corresponding historical optimization schemes, are divided into training data, validation data, and test data; and the training data, validation data, and test data are preprocessed to obtain the training set, validation set, and test set. Choose a large model as the base model; The base model is trained on the training set, and the learning rate and other hyperparameters are adjusted on the validation set to obtain the pre-trained model. By validating the pre-trained model on the test set, a large model with alarm signals and scheme analysis data as inputs and optimized schemes corresponding to the alarm signals as outputs is finally obtained.

[0015] This application compares the total carbon sequestration in a region with its corresponding preset carbon sequestration in the region based on a region ID, thereby determining the status of vegetation carbon sequestration in the region and generating different levels of alarm signals. Subsequently, by combining various parameters in the region ID with the generated alarm signals and a pre-trained scheme suggestion model, targeted optimization schemes are generated for different alarm signals, thus improving the comprehensiveness of the roadside vegetation carbon sequestration assessment method.

[0016] A second aspect of the present invention provides an Internet of Things-based roadside vegetation carbon sequestration assessment system, comprising: a data acquisition module, a data analysis module, and an early warning module; the data acquisition module and the data analysis module are connected together; the data analysis module and the early warning module are connected together. The data acquisition module acquires data from several areas of the road network through data acquisition equipment. The area data includes area IDs and their corresponding remote sensing data, sensor data, and traffic data. The remote sensing data includes multispectral images and lidar point cloud data. The sensor data includes soil data, environmental data, and vegetation data. The data analysis module generates multi-source fusion data based on remote sensing data, sensor data, and traffic data; inputs the multi-source fusion data corresponding to the region ID into the vegetation classification model to obtain the vegetation result data corresponding to the region ID; generates the regional carbon aggregate based on the vegetation result data; and generates alarm signals and optimization schemes based on the regional carbon aggregate. The early warning module: provides prompts based on alarm signals and contacts management personnel.

[0017] Another aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the Internet of Things-based roadside vegetation carbon sequestration assessment system according to the second aspect of the present invention.

[0018] Compared with the prior art, the beneficial effects of this application are: 1. This application generates multi-source fusion data based on remote sensing data, sensor data, and traffic data; inputs the multi-source fusion data corresponding to the region ID into a vegetation classification model to obtain vegetation result data corresponding to the region ID; generates a regional carbon aggregate based on the vegetation result data; and generates an alarm signal and optimization scheme based on the regional carbon aggregate. It considers multiple data sources and dynamically assigns fusion weight coefficients to each data source, making the multi-source fusion data more accurate. It also obtains accurate vegetation result data through a pre-trained vegetation classification model, providing accurate data support for carbon sink assessment and improving the accuracy of carbon sink assessment methods.

[0019] 2. This application obtains multiple data sources and analyzes in detail the impact of several parameters in the data sources on vegetation result data. By using the entropy weight method and the coefficient of variation method, the corresponding fusion weight coefficients of several parameters are dynamically adjusted from the perspective of parameter stability and parameter reliability. This enables the multi-source fusion data to provide richer feature information in the generation of vegetation result data and improve the accuracy of vegetation result data.

[0020] 3. This application dynamically adjusts the carbon content of each vegetation type in the region by using the relationship between the vegetation influence coefficient and the influence threshold in the vegetation result data, so that the carbon content of each vegetation type in the region can be kept in an accurate state, avoiding the problem of decreased carbon sink assessment accuracy caused by using a fixed carbon content, and improving the accuracy of road area vegetation carbon sink assessment method. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a flowchart of the Internet of Things-based roadside vegetation carbon sink assessment method of this application; Figure 2 This is a schematic diagram illustrating the principle of the Internet of Things-based roadside vegetation carbon sequestration assessment system proposed in this application. Detailed Implementation

[0023] The technical solutions of this application will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0024] Please see Figure 1The first aspect of this application provides a method for assessing roadside vegetation carbon sequestration based on the Internet of Things, including: Data from several areas within the road zone is acquired. This data includes area IDs and their corresponding remote sensing data, sensor data, and traffic data. Remote sensing data includes multispectral images and lidar point cloud data. Sensor data includes soil data, environmental data, and vegetation data. In this embodiment, soil data includes soil temperature, soil moisture, and soil CO2 flux. Environmental data includes light intensity, air temperature, and precipitation data. Vegetation data includes diameter at breast height (DBH) increments, etc. Traffic data includes traffic flow and vehicle type. In this embodiment, vehicle type includes gasoline-powered vehicles and new energy vehicles. Multi-source fusion data is generated based on remote sensing data, sensor data, and traffic data; multi-source fusion data refers to the data obtained by fusing remote sensing data, sensor data, and traffic data. The multi-source fusion data corresponding to the region ID is input into the vegetation classification model to obtain the vegetation result data corresponding to the region ID; the vegetation classification model is constructed through a machine learning model and is used to evaluate the vegetation classification results within the region ID; The regional carbon aggregate is generated based on vegetation data; the regional carbon aggregate refers to the actual carbon sink within the region ID. Alarm signals and optimization schemes are generated based on the total regional carbon emissions.

[0025] In this embodiment, the generation of multi-source fused data based on remote sensing data, sensor data, and traffic data includes: Remote sensing data includes multispectral images and lidar point cloud data; Several remote sensing parameters corresponding to multispectral images and lidar point cloud data are extracted; in this embodiment, the remote sensing parameters corresponding to multispectral images include NDVI index and chlorophyll fluorescence, etc.; the remote sensing parameters corresponding to lidar point cloud data include canopy height and LAI index, etc. Sensor data includes soil data, environmental data, and vegetation data; Several sensor parameters corresponding to soil data, environmental data, and vegetation data are extracted; in this embodiment, the sensor parameters include soil temperature, soil moisture, soil CO2 flux, light intensity, air temperature, precipitation data, and diameter at breast height (DBH) increment, etc. Extract several traffic parameters corresponding to the traffic data; in this embodiment, the traffic parameters include traffic flow and vehicle type, etc. Remote sensing parameters, sensor parameters, and traffic parameters are combined using corresponding fusion weighting coefficients to obtain multi-source fused data.

[0026] The fusion weight coefficients in this embodiment are obtained through the following methods: Acquire remote sensing parameters, sensor parameters, and traffic parameters; The stability weights WQ corresponding to remote sensing parameters, sensor parameters, and traffic parameters are calculated using the entropy weight method. The entropy weight method determines the stability weight of each parameter by measuring the data stability of several parameters. The smaller the entropy value of a parameter, the higher the data stability, and the greater its corresponding reliability weight. The reliability weights KQ corresponding to remote sensing parameters, sensor parameters, and traffic parameters are calculated using the coefficient of variation method. The coefficient of variation method determines the reliability weights of each parameter by measuring the reliability of several parameters. The reliability weights are assigned based on the ratio between the standard deviation and the mean of each parameter. Through formula Calculate the fusion weight coefficients corresponding to remote sensing parameters, sensor parameters, and traffic parameters; where n represents the parameter number among the remote sensing parameters, sensor parameters, and traffic parameters; This is expressed as a proportionality coefficient. ∈(0, 1); the specific value is set based on experience, and in this embodiment it will be Setting it to 0.5 means that the fusion weight coefficient of several parameters is considered to have an equal influence from reliable weights and stable weights.

[0027] This embodiment acquires data from multiple data sources and deeply analyzes the impact of several parameters from each data source on vegetation outcome data. Based on this, it combines the entropy weight method and the coefficient of variation method to dynamically adjust the key parameters from the perspectives of parameter stability and reliability, optimizing their contribution to multi-source data fusion. This not only enhances the representativeness and information integrity of the fused data but also provides richer feature support for the generation of vegetation outcome data, thereby significantly improving the accuracy of vegetation identification and classification and providing a high-quality data foundation for subsequent carbon sink assessment.

[0028] The vegetation classification model in this embodiment is constructed using a machine learning model, including: Acquire several historical multi-source fusion data and their corresponding historical vegetation results data; Several historical multi-source fusion data and their corresponding historical vegetation results data were divided into training data, validation data, and test data; and the training data, validation data, and test data were preprocessed to obtain training set, validation set, and test set; the ratio between training set, test set, and validation set was 7:2:1. Choose a machine learning model as the base model; machine learning models include random forest models, etc. The base model is trained on the training set, and the learning rate and other hyperparameters are adjusted on the validation set to obtain the pre-trained model. By validating the pre-trained model on the test set, a vegetation classification model was finally obtained, which takes multi-source fusion data as input and outputs vegetation result data as output.

[0029] In this embodiment, the generation of regional carbon aggregate based on vegetation result data includes: Extract the vegetation type, its corresponding vegetation percentage (ZB), and vegetation influence coefficient from the vegetation results data; Calculate the carbon content of vegetation type based on vegetation impact coefficient; The total regional carbon content is calculated using a formula; the formula satisfies: Where j represents the vegetation type number within region ID; J represents the total number of vegetation types within region ID. It represents the biomass per unit area corresponding to the j-th vegetation type within region ID, and is a basic physical quantity reflecting the carbon sequestration capacity of the vegetation type. This represents the carbon content of the vegetation type corresponding to the j-th vegetation type within the region ID.

[0030] This embodiment dynamically corrects the carbon content of each vegetation type in the region based on the vegetation impact coefficient in the vegetation results data, and integrates key parameters such as vegetation coverage ratio and biomass per unit area to construct a refined carbon sink calculation model. This accurately estimates the total carbon sink in the region, improves the scientificity and accuracy of the actual carbon sink calculation, and provides reliable data support and decision-making basis for triggering subsequent carbon sink anomaly early warning signals and formulating optimized management plans.

[0031] The calculation of vegetation type carbon content based on vegetation influence coefficient in this embodiment includes: Obtain the vegetation type and its corresponding vegetation influence coefficient ZYX; Determine whether the vegetation impact coefficient is greater than the impact threshold; the impact threshold is set based on experience, and in this embodiment, the impact threshold is set to 0.1. Yes, obtain the traffic flow change rate. And the self-recovery rate of vegetation type ZLZHV; When the rate of change in traffic flow is greater than or equal to the change threshold, the formula is used. Calculate the carbon content ZLC of vegetation type; where, This represents the basic carbon content of the vegetation type; the change threshold is set based on experience, and in this embodiment, the change threshold is set to 5%. When the rate of change in traffic flow is less than the change threshold, the carbon content of vegetation type... ; No, vegetation type carbon content .

[0032] In this embodiment, by analyzing the relationship between the vegetation impact coefficient in the vegetation result data and the preset impact threshold, the dynamic adjustment of the carbon content rate of different vegetation types in the region is realized, so that the carbon content rate of various vegetation types is always in an accurate state matching their actual growth and environmental conditions, effectively avoiding the carbon sink assessment deviation caused by the use of a fixed carbon content rate in the traditional method, significantly improving the accuracy and adaptability of the road area vegetation carbon sink assessment, and further enhancing the scientificity and practicality of the assessment results.

[0033] Generating an alarm signal and an optimization plan based on the total regional carbon sink in this embodiment includes: Obtaining the total regional carbon sink corresponding to the regional ID and the preset regional carbon sink; in this embodiment, the preset regional carbon sinks of different regions will vary due to differences in regional location, vegetation quantity distribution within the region, and vegetation types; When the total regional carbon sink is less than or equal to A times the preset regional carbon sink, generating a severe carbon sink deviation alarm signal; When the total regional carbon sink is greater than A times the preset regional carbon sink and less than or equal to B times the preset regional carbon sink, generating a medium carbon sink deviation alarm signal; When the total regional carbon sink is greater than C times the preset regional carbon sink, generating a positive carbon sink change prompt signal; where A, B, and C are proportionality coefficients, A and B ∈ (0, 1), C > 1; and A < B; the specific values are set according to experience, and in this embodiment, A, B, and C are set to 0.7, 0.85, and 1.2 respectively; The alarm signal includes a severe carbon sink deviation alarm signal, a medium carbon sink deviation alarm signal, and a positive carbon sink change prompt signal; Generating an optimization plan based on the alarm signal.

[0034] Generating an optimization plan based on the alarm signal in this embodiment includes: Obtaining the regional ID and its corresponding vegetation type, vegetation proportion, soil carbon background, annual increment of vegetation diameter at breast height, traffic flow, environmental temperature, environmental humidity, total regional carbon sink, and preset regional carbon sink; the soil carbon background is expressed as the organic carbon density of the 0 - 100 cm soil layer; Concatenating the vegetation type, vegetation proportion, soil carbon background, annual increment of vegetation diameter at breast height, traffic flow, environmental humidity, total regional carbon sink, and preset regional carbon sink within the regional ID into plan analysis data; Inputting the alarm signal corresponding to the regional ID and the plan analysis data into the plan recommendation large model to obtain the optimization plan corresponding to the alarm signal within the regional ID; The plan recommendation large model is constructed through the following methods, including: Obtaining a number of historical alarm signals and historical plan analysis data, as well as the historical optimization plans corresponding to the historical alarm signals and historical plan analysis data; Several historical alarm signals and historical scheme analysis data, as well as the corresponding historical optimization schemes, are divided into training data, validation data, and test data. The training data, validation data, and test data are then preprocessed to obtain training sets, validation sets, and test sets. The ratio between the training set, test set, and validation set is 7:2:1. Choose a large model as the base model; large models include DeepSeek models, etc. The base model is trained on the training set, and the learning rate and other hyperparameters are adjusted on the validation set to obtain the pre-trained model. By validating the pre-trained model on the test set, a large model with alarm signals and scheme analysis data as inputs and optimized schemes corresponding to the alarm signals as outputs is finally obtained.

[0035] This embodiment compares and analyzes the actual carbon sequestration within a region ID with its corresponding preset carbon sequestration to determine the current state of vegetation carbon sequestration within the region and generates different levels of early warning signals accordingly. Subsequently, combining various influencing parameters within the region with the generated alarm signals, a pre-trained scheme suggestion model is used to provide targeted optimization suggestions for various alarms. This not only improves the timeliness and accuracy of identifying the status of roadside vegetation carbon sequestration but also enhances the intelligence and responsiveness of the carbon sequestration assessment system, thereby significantly improving the comprehensiveness and practicality of the roadside vegetation carbon sequestration assessment method.

[0036] Please see Figure 2 The second aspect of this application provides an Internet of Things-based roadside vegetation carbon sink assessment system, including: a data acquisition module, a data analysis module, and an early warning module; the data acquisition module and the data analysis module are connected; the data analysis module and the early warning module are connected. Data acquisition module: Acquires data from several areas of the road network through data acquisition equipment; area data includes area ID and its corresponding remote sensing data, sensor data and traffic data; remote sensing data includes multispectral images and lidar point cloud data; sensor data includes soil data, environmental data and vegetation data; data acquisition equipment includes several sensors, multispectral cameras and lidar, etc. Data analysis module: Generates multi-source fusion data based on remote sensing data, sensor data, and traffic data; inputs the multi-source fusion data corresponding to the region ID into the vegetation classification model to obtain the vegetation result data corresponding to the region ID; generates the regional carbon aggregate based on the vegetation result data; generates alarm signals and optimization schemes based on the regional carbon aggregate. Early warning module: Issues alerts based on alarm signals and contacts management personnel.

[0037] Another embodiment of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements an Internet of Things-based roadside vegetation carbon sink assessment system according to a second aspect embodiment of this application.

[0038] Some of the data in the above formula are calculated by removing dimensions and taking their numerical values. The formula is the closest to the real situation obtained by software simulation of a large amount of collected data. The preset parameters and preset thresholds in the formula are set by those skilled in the art according to the actual situation or obtained through simulation of a large amount of data.

[0039] The working principle of this application is as follows: Data from several areas within a road zone is acquired; this data includes area IDs and their corresponding remote sensing data, sensor data, and traffic data; multi-source fusion data is generated based on the remote sensing data, sensor data, and traffic data; the multi-source fusion data corresponding to the area IDs is input into a vegetation classification model to obtain vegetation result data corresponding to the area IDs; a total regional carbon footprint is generated based on the vegetation result data; an alarm signal and optimization scheme are generated based on the total regional carbon footprint. This approach considers multiple data sources and dynamically assigns fusion weight coefficients to each data source, making the multi-source fusion data more accurate. Accurate vegetation result data is obtained through a pre-trained vegetation classification model, providing accurate data support for carbon sink assessment and improving the accuracy of carbon sink assessment methods. This avoids the problem that existing technologies often rely on a single data source and lack consideration of the impact of dynamic factors such as traffic disturbance and soil parameters, leading to low accuracy in carbon sink assessment.

[0040] The above embodiments are only used to illustrate the technical methods of this application and are not intended to limit it. Although this application has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of this application without departing from the spirit and scope of the technical methods of this application.

Claims

1. A method for assessing carbon sequestration in roadside vegetation based on the Internet of Things, characterized in that, include: Data from several areas within the road zone is acquired; the area data includes area IDs and their corresponding remote sensing data, sensor data, and traffic data; the remote sensing data includes multispectral images and lidar point cloud data; the sensor data includes soil data, environmental data, and vegetation data. Multi-source fused data is generated based on remote sensing data, sensor data, and traffic data; the multi-source fused data refers to the data obtained after fusing remote sensing data, sensor data, and traffic data. The multi-source fusion data corresponding to the region ID is input into the vegetation classification model to obtain the vegetation result data corresponding to the region ID; The vegetation classification model is constructed using a machine learning model and is used to evaluate the vegetation classification results within a region ID. Generate the regional carbon aggregate based on vegetation outcome data; The total carbon sequestration of a region refers to the actual carbon sequestration within the region ID. Alarm signals and optimization schemes are generated based on the total regional carbon emissions.

2. The method for assessing roadside vegetation carbon sequestration based on the Internet of Things according to claim 1, characterized in that, The generation of multi-source fused data based on remote sensing data, sensor data, and traffic data includes: The remote sensing data includes multispectral images and lidar point cloud data; Several remote sensing parameters corresponding to multispectral images and lidar point cloud data are extracted; the remote sensing parameters include NDVI index, chlorophyll fluorescence, canopy height and LAI index. The sensor data includes soil data, environmental data, and vegetation data; Extract several sensor parameters corresponding to soil data, environmental data, and vegetation data; the sensor parameters include soil temperature, soil moisture, soil CO2 flux, light intensity, air temperature, precipitation data, and diameter at breast height (DBH) increment; Extract several traffic parameters corresponding to the traffic data; the traffic parameters include traffic flow and vehicle type. Remote sensing parameters, sensor parameters, and traffic parameters are combined using corresponding fusion weighting coefficients to obtain multi-source fused data.

3. The method for assessing roadside vegetation carbon sequestration based on the Internet of Things according to claim 2, characterized in that, The fusion weighting coefficients are obtained through the following methods: Acquire remote sensing parameters, sensor parameters, and traffic parameters; The stability weights WQ corresponding to remote sensing parameters, sensor parameters, and traffic parameters are calculated using the entropy weight method; the entropy weight method determines the stability weights of each parameter by measuring the data stability of several parameters. The reliability weights KQ corresponding to remote sensing parameters, sensor parameters, and traffic parameters are calculated using the coefficient of variation method; the coefficient of variation method determines the reliability weight of each parameter by measuring the reliability of several parameters. Through formula Calculate the fusion weight coefficients corresponding to remote sensing parameters, sensor parameters, and traffic parameters; where n represents the parameter number among the remote sensing parameters, sensor parameters, and traffic parameters; This is expressed as a proportionality coefficient. ∈(0,1).

4. The method for assessing roadside vegetation carbon sequestration based on the Internet of Things according to claim 1, characterized in that, The vegetation classification model is constructed using a machine learning model, including: Acquire several historical multi-source fusion data and their corresponding historical vegetation results data; Several historical multi-source fusion data and their corresponding historical vegetation results data are divided into training data, validation data and test data; and the training data, validation data and test data are preprocessed to obtain training set, validation set and test set. Choose a machine learning model as the base model; The base model is trained on the training set, and the learning rate and other hyperparameters are adjusted on the validation set to obtain the pre-trained model. By validating the pre-trained model on the test set, a vegetation classification model was finally obtained, which takes multi-source fusion data as input and outputs vegetation result data as output.

5. The method for assessing roadside vegetation carbon sequestration based on the Internet of Things according to claim 1, characterized in that, The generation of the regional carbon aggregate based on vegetation data includes: Extract the vegetation type, its corresponding vegetation percentage (ZB), and vegetation influence coefficient from the vegetation results data; Calculate the carbon content of vegetation type based on vegetation impact coefficient; The total regional carbon content is calculated using a formula; the formula satisfies: Where j represents the vegetation type number within region ID; J represents the total number of vegetation types within region ID. Represented as the biomass per unit area corresponding to the j-th vegetation type within region ID; This represents the carbon content of the vegetation type corresponding to the j-th vegetation type within the region ID.

6. The method for assessing roadside vegetation carbon sequestration based on the Internet of Things according to claim 5, characterized in that, The calculation of vegetation type carbon content based on vegetation influence coefficient includes: Obtain the vegetation type and its corresponding vegetation influence coefficient ZYX; Determine whether the vegetation impact coefficient is greater than the impact threshold; Yes, obtain the traffic flow change rate. And the self-recovery rate of vegetation type ZLZHV; When the rate of change in traffic flow is greater than or equal to the change threshold, the formula is used. Calculate the carbon content ZLC of vegetation type; where, Represented as the basic carbon content of vegetation type; When the rate of change in traffic flow is less than the change threshold, the carbon content of vegetation type... ; No, vegetation type carbon content .

7. The method for assessing roadside vegetation carbon sequestration based on the Internet of Things according to claim 1, characterized in that, The generation of alarm signals and optimization schemes based on regional carbon aggregates include: Get the total carbon sequestration and preset carbon sink for the region corresponding to the region ID; When the total regional carbon sequestration is less than or equal to A times the preset regional carbon sequestration, a severe carbon sequestration deviation alarm signal is generated. When the total regional carbon amount is greater than A times the regional preset carbon sink amount and the total regional carbon amount is less than or equal to B times the regional preset carbon sink amount, a carbon sink moderate deviation alarm signal is generated. When the total regional carbon sink exceeds C times the preset regional carbon sink amount, a positive carbon sink change indication signal is generated; where A, B, and C are proportionality coefficients, A and B ∈ (0, 1), C > 1; and A <B; The alarm signals include a severe carbon sink deviation alarm signal, a moderate carbon sink deviation alarm signal, and a positive carbon sink change indication signal; An optimized scheme for generating alarm signals.

8. The method for assessing roadside vegetation carbon sequestration based on the Internet of Things according to claim 7, characterized in that, The optimized scheme based on alarm signal generation includes: Obtain the region ID and its corresponding vegetation type, vegetation percentage, soil carbon background, annual increase in vegetation diameter at breast height, traffic flow, ambient temperature, ambient humidity, total regional carbon sequestration, and preset regional carbon sink. The data for the scheme analysis is spliced ​​together from the vegetation type, vegetation ratio, soil carbon background, annual increase in vegetation diameter at breast height, traffic flow, environmental humidity, total regional carbon and the pre-set regional carbon sink within the region ID. Input the alarm signal and scheme analysis data corresponding to the region ID into the scheme suggestion model to obtain the optimized scheme corresponding to the alarm signal within the region ID; The proposed solution suggests that the large model be constructed in the following ways: Acquire several historical alarm signals and historical scheme analysis data, as well as the historical optimization schemes corresponding to the historical alarm signals and historical scheme analysis data; Several historical alarm signals and historical scheme analysis data, as well as the corresponding historical optimization schemes, are divided into training data, validation data, and test data; and the training data, validation data, and test data are preprocessed to obtain the training set, validation set, and test set. Choose a large model as the base model; The base model is trained on the training set, and the learning rate and other hyperparameters are adjusted on the validation set to obtain the pre-trained model. By validating the pre-trained model on the test set, a large model with alarm signals and scheme analysis data as inputs and optimized schemes corresponding to the alarm signals as outputs is finally obtained.

9. A roadside vegetation carbon sequestration assessment system based on the Internet of Things (IoT), applied to the roadside vegetation carbon sequestration assessment method based on the IoT as described in any one of claims 1-8, characterized in that, include: Interconnected data acquisition and data analysis modules; The data acquisition module acquires data from several areas of the road network through data acquisition equipment. The area data includes area IDs and their corresponding remote sensing data, sensor data, and traffic data. The remote sensing data includes multispectral images and lidar point cloud data. The sensor data includes soil data, environmental data, and vegetation data. The data analysis module generates multi-source fusion data based on remote sensing data, sensor data, and traffic data; it inputs the multi-source fusion data corresponding to the region ID into the vegetation classification model to obtain the vegetation result data corresponding to the region ID. Generate the regional carbon aggregate based on vegetation outcome data; Alarm signals and optimization schemes are generated based on the total regional carbon emissions.

10. A computer-readable storage medium, applied to the Internet of Things-based roadside vegetation carbon sequestration assessment system as described in claim 9, characterized in that, The computer program is stored on the computer-readable storage medium.

Images