Gobi photovoltaic base ecological restoration carbon sequestration benefit evaluation method and system

Abstract

The invention discloses a gobi photovoltaic base ecological restoration carbon sequestration benefit evaluation method and system, and particularly relates to the technical field of ecological environment governance, and the method comprises the following steps: dividing a photovoltaic base region into a photovoltaic panel lower region, a photovoltaic panel inter-row channel region and a photovoltaic field peripheral region, building corresponding ecological simulation models for different regions, and constructing a photovoltaic field simulation model; simulation data are obtained and substituted into the first common calculation model and the second common calculation model respectively, calculation results of all the areas are summarized and substituted into a pre-trained machine learning model, and an output result is an ecological restoration carbon sequestration benefit value used for representing the final score of overall ecological restoration of the photovoltaic base; according to the method, the vegetation restoration and soil carbon storage process can be dynamically simulated, compared with a traditional static monitoring method, the future ecological restoration trend can be predicted, scientific support is provided for long-term ecological management, a targeted ecological optimization scheme can be provided based on regionalized simulation, and the ecological value of renewable energy sources is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of ecological environment governance, and more specifically, to a method and system for evaluating the carbon fixation benefits of ecological restoration of a Gobi photovoltaic base. Background Art

[0002] In the context of the global energy structure transformation, photovoltaic power generation has become a core component of low-carbon renewable energy, especially in areas with sufficient sunlight but fragile ecological environment (such as Gobi, desert, semi-desert), the deployment scale of photovoltaic bases continues to expand. However, the construction of large-scale photovoltaic bases may have a profound impact on the ecosystem, mainly including: Changes in the surface environment: Photovoltaic panels block solar radiation, affecting vegetation growth, soil moisture balance and local climate. Intensified soil erosion: Increased surface damage and wind and sand activities may lead to soil loss and reduced carbon storage. Uncertainty in vegetation recovery: The vegetation recovery between rows of photovoltaic panels is affected by factors such as moisture, light, and wind and sand movement, and requires scientific quantitative evaluation.

[0003] In response to these problems, existing ecological assessment methods mainly rely on remote sensing monitoring and field surveys, which make it difficult to dynamically simulate the ecological restoration process and carbon storage capacity. Therefore, the present invention proposes a method and system for evaluating the carbon fixation benefits of ecological restoration of a Gobi photovoltaic base, in order to solve the above problems. Summary of the invention

[0004] To achieve the above object, the present invention provides the following technical solutions: A method for evaluating the carbon sequestration benefits of ecological restoration of a Gobi photovoltaic base comprises the following steps: The photovoltaic base area is divided into the area under the photovoltaic panels, the channel area between the photovoltaic panels and the peripheral area of ​​the photovoltaic field, and corresponding ecological simulation models are established for different areas; For each divided area, simulation data is obtained from the corresponding ecological simulation model and substituted into the shared calculation model 1 that reflects the progress and quality of regional vegetation restoration, and the shared calculation model 2 that reflects the soil restoration effect and carbon sequestration capacity; The calculation results of all regions are summarized and substituted into the pre-trained machine learning model. The output result is the ecological restoration carbon fixation benefit value used to represent the final score of the overall ecological restoration of the photovoltaic base.

[0005] In a preferred embodiment, the logic of the ecological simulation model of the area under the photovoltaic panel is: The photovoltaic array arrangement is used to simulate the light intensity distribution and construct a light distribution model: Based on the Richards equation, water infiltration, evaporation and migration are simulated to construct a soil moisture dynamic model; The Logistic growth model is used to construct the vegetation growth model; The constructed light distribution model, soil moisture dynamics model and vegetation growth model are used for simulation.

[0006] In a preferred embodiment, the logic of the ecological simulation model of the photovoltaic panel inter-row channel area is: The Markov chain model is used to predict vegetation community succession, simulate the competition and succession process of different species under different restoration strategies, and construct a vegetation restoration model; Use soil carbon cycle models to simulate organic matter accumulation and mineralization processes, predict soil carbon storage changes, and build soil quality evolution models; The constructed vegetation restoration model and soil quality evolution model were used for simulation.

[0007] In a preferred embodiment, the logic of the ecological simulation model of the photovoltaic field peripheral area is: Based on the Saltation-Balance equation, the wind-sand movement is simulated and the wind-sand transport model is constructed; Use the RUSLE model to correct soil loss rate, evaluate the role of vegetation in soil protection, and construct a soil erosion model; The constructed wind-sand transport model and soil erosion model are used for simulation.

[0008] In a preferred embodiment, the common calculation model refers to a vegetation restoration index calculation model.

[0009] In a preferred embodiment, the shared calculation model 2 refers to a soil carbon storage index calculation model.

[0010] In a preferred embodiment, the pre-trained machine learning model is a convolutional neural network model.

[0011] In a preferred embodiment, a Gobi photovoltaic base ecological restoration carbon fixation benefit assessment system includes: The regional division module divides the photovoltaic base area into the area below the photovoltaic panels, the channel area between the photovoltaic panels and the peripheral area of ​​the photovoltaic field; Simulation module, used to establish corresponding ecological simulation models for different regions; A data acquisition module is used to acquire simulation data from the corresponding ecological simulation model for each divided area; The regional assessment module is used to substitute the obtained simulation data into the shared calculation model 1 reflecting the progress and quality of regional vegetation restoration, and the shared calculation model 2 reflecting the soil restoration effect and carbon sequestration capacity; The overall evaluation module is used to summarize the calculation results of all regions and substitute them into the pre-trained machine learning model. The output result is the ecological restoration carbon fixation benefit value used to represent the final score of the overall ecological restoration of the photovoltaic base.

[0012] Technical effects and advantages of the present invention: Traditional ecological assessment methods mainly rely on remote sensing images and field surveys, but these methods cannot dynamically predict the ecological restoration process of photovoltaic base areas, and the assessment results are greatly affected by factors such as data collection frequency. Existing carbon sink calculation methods mostly use empirical formulas or statistical models, which are difficult to accurately predict soil carbon storage changes. Based on ecological simulation modeling, the present invention accurately simulates the ecological restoration process. For the area under the photovoltaic panel, the channel area between the photovoltaic panel rows, and the peripheral area of ​​the photovoltaic field, simulation models such as light distribution, soil moisture, vegetation growth, wind and sand transport, and soil carbon cycle are constructed respectively to dynamically simulate vegetation restoration and soil carbon storage processes. Compared with traditional static monitoring methods, the present invention can predict future ecological restoration trends and provide more scientific support for long-term ecological management. Vegetation restoration index (VRI) and soil carbon storage index (SCI) are constructed to achieve cross-regional comparability. Nonlinear transformation and normalization processing are used to ensure that the index calculation results of different regions are comparable, making the assessment more scientific and objective. Compared with the single NDVI index, the VRI index of the present invention combines factors such as light, soil moisture, and species diversity, and can more comprehensively reflect the quality of vegetation restoration.

[0013] Traditional carbon sink assessment methods mostly use soil sampling analysis or empirical formula calculations, but due to the long-term evolution characteristics of soil carbon cycle, a single measurement is difficult to reflect long-term carbon storage changes. Existing carbon sink calculations do not take into account the impact of ecological restoration in the photovoltaic base area on carbon storage, resulting in large deviations in the assessment results. The present invention is based on a soil carbon cycle simulation model, dynamically calculates carbon sink capacity, and uses multiple factors such as organic carbon accumulation, nitrogen storage, soil density, and soil moisture regulation to calculate the SCI index to accurately predict soil carbon sink capacity. Through long-term simulation, the contribution of different ecological restoration measures (such as vegetation restoration and sand control) in the photovoltaic base area to soil carbon storage is evaluated, providing a scientific basis for carbon trading and ecological compensation. By accurately calculating the soil carbon storage capacity, this method can provide quantitative support for photovoltaic companies to apply for carbon credits, and promote the photovoltaic industry to develop in a low-carbon, eco-friendly way.

[0014] The present invention adopts convolutional neural network (CNN) to automatically learn the nonlinear relationship between indexes. The method processes the VRI and SCI indexes of three regions through CNN, which can automatically identify different ecological restoration modes and improve prediction accuracy. CNN can explore the potential interactive relationship between indexes in different regions, such as how vegetation restoration affects soil carbon storage and how wind and sand movement affects vegetation growth, thereby optimizing ecological management strategies.

[0015] Traditional evaluation methods usually conduct a unified analysis of the photovoltaic base area, failing to consider the differences in ecological characteristics under the photovoltaic panels, between-row channels and peripheral areas, resulting in inaccurate restoration strategies. Existing management strategies often use a "one-size-fits-all" approach to carry out ecological restoration, lacking refined management methods. Based on regional simulation, the present invention provides targeted ecological optimization solutions. Through regional modeling, the following are analyzed separately: The area under the photovoltaic panels: optimize the arrangement of photovoltaic panels, improve light transmittance, select shade-tolerant vegetation, and improve the ecological restoration effect. The area between the rows of photovoltaic panels: use the vegetation restoration model to optimize the community succession process and improve the soil carbon storage capacity. The peripheral area of ​​the photovoltaic field: combine the wind and sand transport model, reasonably arrange the sand-fixing vegetation, and improve the windproof and soil-fixing capacity. It also supports dynamic adjustment of ecological restoration strategies to improve the long-term ecological benefits of photovoltaic bases. Through long-term monitoring and simulation calculations, the ecological restoration plan can be dynamically adjusted. For example, in areas with severe wind and sand erosion, biological sand fixation measures (such as shrub belts) are added, and in areas with low carbon storage capacity, soil improvement measures (such as increasing organic matter) are optimized.

[0016] The construction of traditional photovoltaic bases mainly focuses on energy benefits, while ecological impacts and carbon sink benefits are not fully incorporated into decision-making, which may lead to long-term ecological degradation problems. The ecological value of current renewable energy has not been fully quantified, affecting its application in the carbon market and ecological compensation mechanism. The present invention establishes a coordinated optimization system of ecological restoration and photovoltaic power generation. By optimizing ecological restoration measures, it can not only reduce the negative impact of photovoltaic bases on the environment, but also improve the adaptability of ecosystems to climate change and achieve sustainable development. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to facilitate understanding by those skilled in the art, the present invention is further described below in conjunction with the accompanying drawings; Figure 1 This is a schematic diagram of a method for evaluating the carbon fixation benefits of ecological restoration of a Gobi photovoltaic base in the present invention.

[0018] Figure 2 This is a schematic diagram of a Gobi photovoltaic base ecological restoration and carbon fixation benefit assessment system in the present invention. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] Reference Figure 1 - Figure 2 The following embodiments are obtained:

[0021] Example 1: In the context of global energy structure transformation, photovoltaic power generation, as the main force of renewable energy, has been widely used in vast Gobi, desert, semi-desert and other areas. These areas have sufficient light resources and low land use value, which are very suitable for the construction of photovoltaic bases. However, the impact of large-scale deployment of photovoltaic bases on the ecological environment is still uncertain, mainly in the following aspects: Ecological and environmental challenges of Gobi photovoltaic bases: Photovoltaic panels affect the surface ecosystem: The installation of photovoltaic panels changes the original surface environment, forming shadow areas under the photovoltaic panels, affecting the photosynthesis and growth of vegetation. The evaporation of soil moisture under the photovoltaic panels is reduced, which may retain water to a certain extent, but may also inhibit the adaptability of native vegetation.

[0022] Fragile ecosystems and limited recovery capacity: The natural ecosystems in Gobi and desert areas are very fragile, with slow growth of native vegetation and extremely low soil organic carbon content. Human intervention (such as photovoltaic base construction) may destroy the original ecological balance and make ecological recovery difficult.

[0023] Wind and sand erosion and soil loss: The wind and sand movement outside the photovoltaic base is strong. If it is not properly managed, it may aggravate desertification. The exposed surface between the rows of photovoltaic panels is susceptible to wind and water erosion, resulting in soil loss and affecting the stability of the photovoltaic base.

[0024] Unclear carbon sequestration capacity: There is currently a lack of systematic evaluation of the carbon sequestration benefits of Gobi photovoltaic bases, and it is impossible to determine the vegetation restoration and soil carbon storage growth rates in the region.

[0025] Limitations of traditional ecological assessment methods: Limitations of remote sensing monitoring: Existing ecological restoration monitoring usually relies on remote sensing data (such as NDVI and vegetation coverage), but lacks refined soil moisture and carbon storage assessments, making it difficult to accurately quantify carbon sequestration capacity.

[0026] Lack of regional refined ecological simulation: Traditional ecological assessments are mostly based on historical data, but fail to build independent simulation models for the ecological characteristics of different regions, and are unable to accurately simulate the impact of photovoltaic bases on soil moisture and vegetation restoration.

[0027] Insufficient data-driven intelligent assessment: Currently, the ecological impact of photovoltaic bases is mostly assessed using empirical formulas or statistical methods, but there is a lack of a comprehensive assessment framework based on machine learning, which makes it impossible to effectively integrate multi-source data for decision support.

[0028] The invention not only fills the gap in the evaluation of the benefits of photovoltaic base ecological restoration, but also has important significance in ecological environmental protection, photovoltaic base optimization design, carbon sink calculation, etc.: Ecological and environmental protection: Through the ecological simulation model, the vegetation recovery and soil carbon storage capacity of the area under the photovoltaic panels, the channel area between the photovoltaic panels and the peripheral area of ​​the photovoltaic field are calculated to clarify the ecological recovery trend of each area and provide a scientific basis for the ecological management of the photovoltaic base. Through the dynamic modeling of soil moisture, the utilization of water resources is optimized and the efficiency of ecological restoration in arid areas is improved. Through the simulation of wind and sand transport, the impact of peripheral vegetation on wind and sand erosion is evaluated to reduce the risk of desertification in the photovoltaic base area.

[0029] Photovoltaic base optimization design: Through the light distribution model, the light environment under the photovoltaic panels is evaluated, the layout of photovoltaic components is optimized, and the ecological compatibility is improved. Through the vegetation restoration model, the construction of ecological corridors inside the photovoltaic field is guided to promote the restoration of biodiversity. Through the ecological restoration benefit evaluation, scientific quantitative standards are provided for the green certification and sustainable development evaluation of photovoltaic bases.

[0030] Carbon sink calculation and carbon market application: Soil carbon storage index (SCI) is used to calculate soil organic carbon, total nitrogen content, and soil moisture changes to accurately assess the carbon sink capacity of photovoltaic bases. Through machine learning models (CNN), the accuracy of carbon sink prediction is improved, providing a scientific basis for the carbon trading market, so that photovoltaic bases can not only contribute clean energy, but also benefit from carbon sink value-added.

[0031] This paper proposes a set of ecological restoration carbon fixation benefit assessment methods of regional simulation + index calculation + intelligent assessment, the core contents of which include the following three aspects: Ecological simulation modeling based on regional division: This method divides the photovoltaic base area into the area under the photovoltaic panels, the channel area between the photovoltaic panels and the peripheral area of ​​the photovoltaic field. According to the ecological characteristics of different areas, corresponding simulation models are established respectively: Area under photovoltaic panels: Light distribution model: simulates the effect of photovoltaic panel layout on light intensity. Soil moisture dynamic model: calculates water infiltration, evaporation and migration based on Richards equation. Vegetation growth model: uses logistic growth model to predict vegetation recovery.

[0032] Photovoltaic panel inter-row channel area: Vegetation restoration model: Predict the succession process of different vegetation species based on the Markov chain model. Soil quality evolution model: Calculate the accumulation and mineralization of organic matter based on the soil carbon cycle model.

[0033] Peripheral area of ​​the photovoltaic field: Wind and sand transport model: The Saltation-Balance equation is used to calculate the impact of wind and sand flow on surface vegetation. Soil erosion model: The role of vegetation in soil conservation is calculated based on the RUSLE formula.

[0034] Comprehensive evaluation of machine learning model: Input data: The photovoltaic base area is divided into the area under the photovoltaic panel (Area A), the area between the photovoltaic panel rows (Area B), and the area outside the photovoltaic field (Area C). The vegetation restoration index (VRI) and soil carbon storage index (SCI) of Area A, Area B, and Area C are used as input features. Evaluation model: Convolutional neural network (CNN) is used to extract cross-regional ecological restoration features. Final output: Calculate the ecological restoration and carbon sequestration benefit value of the photovoltaic base as a whole, which is used for ecological benefit evaluation and carbon sink trading support.

[0035] The present invention discloses a method for evaluating the carbon fixation benefits of ecological restoration of a Gobi photovoltaic base, comprising the following steps: The photovoltaic base area is divided into the area under the photovoltaic panels, the area between the photovoltaic panels and the area outside the photovoltaic field, and corresponding ecological simulation models are established for different areas; due to the significant differences in light, water, vegetation succession, soil environment and other aspects in different areas of the photovoltaic base, unified modeling cannot accurately reflect local ecological changes. Through regional division, the phenomena of photovoltaic panel shading, soil moisture migration, and wind and sand transport are simulated respectively to improve the applicability of the simulation model. The area under the photovoltaic panels (area A) is affected by the shading of photovoltaic panels, and soil evaporation is reduced, but the photosynthetic efficiency is reduced. It is necessary to study a suitable shade-tolerant plant restoration plan. The photovoltaic panel inter-row channel (area B) has good light and precipitation conditions and is suitable as an ecological restoration zone. It is necessary to simulate vegetation succession and soil carbon sink changes. The periphery of the photovoltaic field (area C) is significantly affected by wind and sand erosion, and it is necessary to simulate the role of sand-fixing vegetation in wind and sand control and soil conservation.

[0036] For each divided area, the simulation data is obtained from the corresponding ecological simulation model and substituted into the common calculation model 1 reflecting the progress and quality of regional vegetation restoration, and the common calculation model 2 reflecting the soil restoration effect and carbon sequestration capacity; traditional ecological assessment methods are mostly based on historical statistics or remote sensing data. This method can dynamically predict the changing trends of different ecological factors, including vegetation recovery rate, soil moisture dynamics, wind and sand transport, etc., by establishing a physical ecological model. VRI and SCI are calculated by simulation results to avoid data missing problems and improve the scientificity and operability of the assessment. Two general indexes (VRI and SCI) are used to ensure that ecological restoration and carbon sequestration capacity can be uniformly quantified under different environmental conditions. Although the simulation data of the area under the photovoltaic panel, the inter-row channel, and the peripheral area are different, they can all be calculated by VRI and SCI, making the model more applicable. Multi-scale data fusion improves the prediction ability of ecological restoration.

[0037] Simulation data from different regions can be analyzed on multiple time scales, for example: Short-term effects (1-3 years): focus on the impact of photovoltaic panel shading on vegetation and soil moisture. Medium-term effects (5-10 years): analyze the improvement of carbon sequestration capacity by ecological restoration measures (such as vegetation planting and irrigation). Long-term effects (more than 10 years): predict the accumulation of soil organic carbon and the effect of wind and sand control during the operation period of the photovoltaic base. VRI (vegetation recovery index): measures the recovery speed, health and diversity of vegetation. SCI (soil carbon storage index): measures the accumulation of soil organic carbon. Moreover, cross-regional comparison can optimize ecological restoration strategies. For example, if the VRI of a certain area is lower than expected, it indicates that the vegetation recovery is not ideal, and the vegetation type, density and irrigation strategy can be adjusted. If the SCI is low, it means that the soil carbon storage capacity is insufficient, and the input of organic matter (such as biological cover) can be increased and fertilization measures can be optimized. The trend changes of the same index can help users observe the long-term effectiveness of ecological restoration measures.

[0038] The calculation results of all regions are summarized and substituted into the pre-trained machine learning model, and the output result is the ecological restoration carbon fixation benefit value used to represent the final score of the overall ecological restoration of the photovoltaic base. The calculation results of all regions are summarized and substituted into the pre-trained convolutional neural network (CNN) model, and the ecological restoration carbon fixation benefit value is finally output to represent the overall ecological restoration effect of the photovoltaic base. CNN is used to automatically extract cross-regional ecological characteristics to avoid the uncertainty of artificially set weights and improve calculation accuracy. Combined with data from multiple regions for comprehensive learning, the evaluation results are more in line with the actual laws of ecological changes. Ecological restoration involves multivariate nonlinear interactions (light, precipitation, soil moisture, wind and sand, etc.), and traditional linear models are difficult to accurately predict long-term changes. CNN extracts multidimensional features through convolution kernels, which can better predict vegetation and soil restoration trends. It has strong scalability and can be applied to photovoltaic bases in different regions, including desert type, grassland type, and agricultural photovoltaic complementary type. It can also be combined with remote monitoring systems to facilitate subsequent automated ecological monitoring and dynamic adjustment of management strategies.

[0039] The logic of the ecological simulation model for the area under the photovoltaic panels is: The photovoltaic array arrangement is used to simulate the light intensity distribution and construct a light distribution model: Based on the Richards equation, water infiltration, evaporation and migration are simulated to construct a soil moisture dynamic model; The Logistic growth model is used to construct the vegetation growth model; The constructed light distribution model, soil moisture dynamics model and vegetation growth model are used for simulation.

[0040] The light distribution model uses the photovoltaic array arrangement to simulate the spatial distribution of light intensity and calculates the photosynthetically active radiation (PAR) under and around the photovoltaic panels as an input variable for vegetation growth and soil evaporation calculations. The formula is: ; is the effective light below the photovoltaic panel (W / m²), is the direct sunlight intensity (W / m²), is the shading ratio of the photovoltaic panel, defined as: ; is the shadow length, which is determined by the height of the photovoltaic panel and the sun's altitude Calculated, ; is the row spacing of photovoltaic panels; is the scattered light part, calculated as: ; is the diffuse light ratio, is the total incident light. By quantifying the impact of photovoltaic panels on light and determining the potential for vegetation growth and evaporation, ecological restoration strategies can be optimized by adjusting the spacing or height of photovoltaic panels.

[0041] The soil moisture dynamic model calculates water infiltration, evaporation and migration based on the Richards equation, simulates the soil moisture distribution under the influence of photovoltaic panels, and provides the necessary moisture information for vegetation growth. The formula is: ; is the soil moisture content (m³ / m³), is (time variable, unit: seconds or days), which represents the time evolution process of soil moisture, that is, the change of soil moisture content over time. It can be used to describe the dynamic changes of moisture, including the evolution of precipitation infiltration, soil moisture evaporation, plant absorption and other processes over time. It is a depth variable, unit: meter, which indicates the depth in the soil profile. It usually takes a positive value from the surface downward, that is, the value of 0 represents the surface. It is used to describe the distribution and migration of water in the vertical direction, for example: surface water infiltrates downward (infiltration), capillary water migrates upward from deep soil (transpiration), and water in deep soil may further seep into the underground due to gravity. is the hydraulic conductivity (m / s), which indicates the rate at which water penetrates the soil and depends on the water content: ; is the saturation permeability, is the saturated water content, n is the empirical index, where is the soil matrix potential (m), representing the water adsorption capacity in the soil, is the evaporation rate (m / s), calculated as: ; is the maximum evaporation capacity, The maximum light intensity is calculated by calculating the soil moisture changes to determine whether vegetation can survive and grow, which is helpful for evaluating the potential water resource benefits of photovoltaic panels in reducing evaporation.

[0042] The vegetation growth model is based on the Logistic growth model, combined with the influencing factors of light (PAR) and soil moisture (SWC), to simulate vegetation recovery, including changes in vegetation coverage, leaf area index (LAI) and biomass. The formula is: ; is the vegetation coverage (%), is the growth rate (% / day), which determines the growth rate of vegetation. is the maximum vegetation coverage (%), is a time variable, indicating vegetation coverage The growth process over time. This variable is used to describe the growth rate of vegetation, that is, how the vegetation grows from the initial state to the stable value (maximum coverage) within a certain period of time. ). is the light adaptation factor: ; For the optimal light level, is the soil moisture adaptation factor: ; The optimal soil moisture content. By predicting the vegetation recovery speed under photovoltaic panels, scientific guidance can be provided for ecological restoration, and the appropriate vegetation types and management plans can be optimized by combining the effects of light and water.

[0043] The logic of the ecological simulation model of the inter-row channel area of ​​photovoltaic panels is: The Markov chain model is used to predict vegetation community succession, simulate the competition and succession process of different species under different restoration strategies, and construct a vegetation restoration model; Use soil carbon cycle models to simulate organic matter accumulation and mineralization processes, predict soil carbon storage changes, and build soil quality evolution models; The constructed vegetation restoration model and soil quality evolution model were used for simulation.

[0044] The photovoltaic panel inter-row channel area (area B) is a key area for ecological restoration within the photovoltaic base. Due to its good light and water conditions, this area is suitable as an ecological restoration zone to promote vegetation restoration and soil carbon storage accumulation. Therefore, the ecological simulation of this area is mainly composed of a vegetation restoration model and a soil quality evolution model. The vegetation restoration model uses the Markov chain model to predict the succession of vegetation communities, simulates the competition, adaptation and succession process of different species under different ecological restoration strategies, and evaluates the vegetation restoration in this area. The formula is: ; For time The vegetation distribution state vector at the moment indicates the coverage ratio of different vegetation types: ; For the Plants in The coverage ratio at the moment, For time The vegetation distribution state vector at time , is the vegetation succession transition probability matrix: ; Indicates vegetation type Transform to type The probability of the PV panels being replaced by the PV panels is affected by factors such as climate, soil nutrients, and competition. Description of simulation steps: Combine the light distribution model, soil moisture dynamic model, and vegetation growth model to perform spatiotemporal dynamic simulation and predict the ecological restoration process in the area under the PV panels. It can predict the dynamic changes of different vegetation types, evaluate the ecological restoration path, and help consider the long-term impact of species competition and restoration measures on vegetation restoration, and optimize ecological restoration strategies.

[0045] The soil quality evolution model uses the soil carbon cycle model to calculate the accumulation and mineralization process of organic matter, predict the changes in soil carbon storage (SOC), and construct a soil quality evolution model. The formula is: ; is the soil organic carbon content, unit: g / m², indicating the organic carbon stored in the soil. is the input carbon amount, unit: g / m² / year, ; For the The coverage ratio of the vegetation type (from the vegetation restoration model), For the The litter input of this type of vegetation (g / m² / year) indicates the contribution of this species to soil organic carbon, and n indicates the total number of vegetation types. is the mineralization loss rate, ; is the baseline mineralization rate, is the current soil temperature (℃), is the reference temperature (°C), is the preset temperature sensitivity coefficient (determines the effect of temperature on mineralization rate), is the time step of soil carbon change. Since soil carbon changes slowly, it is different from the time step of water or vegetation growth model. Description of simulation steps: Combine the vegetation restoration model (Markov chain) and the soil quality evolution model (carbon cycle) to perform multi-time scale simulation to predict the long-term ecological restoration trend of the inter-row channel area of ​​photovoltaic panels. By optimizing the carbon sequestration capacity of the soil in the photovoltaic base area, a scientific basis can be provided for carbon trading. Combined with vegetation restoration, the contribution of different vegetation types to soil carbon storage can be evaluated and vegetation selection can be optimized. Input the initial vegetation distribution, calculate the vegetation succession path, calculate the soil carbon input according to the vegetation type and coverage ratio, input the current soil temperature, calculate the carbon mineralization rate, iteratively update the soil organic carbon content, and predict long-term carbon storage changes.

[0046] The logic of the ecological simulation model in the peripheral area of ​​the photovoltaic field is as follows: Based on the Saltation-Balance equation, the wind and sand movement is simulated and the wind and sand transport model is constructed: Use the RUSLE model to correct soil loss rate, evaluate the role of vegetation in soil protection, and construct a soil erosion model; The constructed wind-sand transport model and soil erosion model are used for simulation.

[0047] The area outside the photovoltaic field (area C) is very prone to wind erosion and soil loss due to exposure to strong winds, which affects the stability of photovoltaic equipment and aggravates regional desertification. Therefore, the ecological simulation of this area mainly includes: wind and sand transport model (based on the Saltation-Balance equation) to simulate the movement characteristics of wind and sand. Soil erosion model (based on the RUSLE formula) to evaluate the role of vegetation in soil conservation and predict the soil loss rate. Wind and sand transport + soil erosion simulation to predict the long-term ecological evolution trend outside the photovoltaic base.

[0048] In the wind-sand transport model (based on the Saltation-Balance equation), wind-sand transport refers to the process in which surface sand is blown up, collided and transported under the action of wind. This process is affected by factors such as wind speed, sand size, and surface roughness. The Saltation-Balance equation is used to describe the dynamic balance of wind-sand transport. The formula is: ; is the wind-sand transport flux per unit width (i.e. the amount of sand passing through a certain cross section per unit time), is the horizontal coordinate of the wind direction, is the wind erosion supply, which indicates the amount of sand blown up by the wind on the surface. The calculation formula is: ; is the air density (kg / m³), is the preset surface wind erosion coefficient, dimensionless, is the wind speed (m / s), is the sand-raising threshold wind speed (m / s), which determines whether wind erosion occurs. is the amount of sand sedimentation, which means the amount of sand that settles due to gravity or vegetation after being transported to a certain place. The calculation formula is: ; It is the preset sedimentation coefficient, which is related to factors such as surface roughness and vegetation density.

[0049] In the soil erosion model (based on the RUSLE formula), soil erosion refers to the loss of soil particles under the action of water erosion and wind erosion. The presence of vegetation can effectively reduce the soil loss rate. This study uses RUSLE to correct the soil erosion rate and quantify the role of vegetation in soil conservation.

[0050] ; A (unit: t / ha / year) is the soil erosion rate, which indicates the amount of soil lost per unit area per year; R (unit: MJ·mm / ha·h·year) is the rainfall erosivity factor, which indicates the impact of rainfall on soil erosion. The calculation formula is: ; For the Kinetic energy of rainfall (MJ / ha), is the maximum rainfall intensity in 30 minutes (mm / h), n is the total number of rainfalls, (Unit: t·ha·h / MJ·mm), the preset soil erodibility factor, indicates the susceptibility of soil to erosion. and (dimensionless), is the preset slope length factor and slope factor, which affects the erosion effect of water flow on soil. (dimensionless), is the vegetation cover factor, which indicates the role of vegetation in soil conservation. ; is an empirical parameter, is the leaf area index, which comes from the vegetation restoration model. (dimensionless) is the soil and water conservation measures factor, which reflects the impact of human governance measures. It is assigned according to the specific soil and water conservation measures and regional characteristics. The assignment (based on empirical values) is given as an example: If the periphery of the photovoltaic base is completely exposed and there are no soil and water conservation measures, then P=1. If shrubs or herbaceous vegetation (such as desert green belts) are planted, then P=0.1-0.5. If combined with windbreak and sand fixation projects (such as grass grids and sand barriers), then P=0.05-0.2. n represents the number of rainfalls in rainfall calculations, and represents the number of vegetation types in the Markov chain model. Description of simulation steps: Combine the wind and sand transport model and the soil erosion model to perform multi-factor simulation, predict the ecological evolution trend of the area outside the photovoltaic field, input wind speed and surface characteristics, and calculate the wind and sand transport flux , input rainfall and vegetation conditions, calculate soil erosion rate A, simulate wind and sand deposition and vegetation blocking effects, and predict soil conservation effects. By predicting the wind and sand erosion trend around the photovoltaic base and optimizing the protection system, the inhibitory effect of vegetation on wind and water erosion can be quantified, thus improving the efficiency of ecological restoration.

[0051] The common calculation model 1 refers to the vegetation restoration index calculation model. Specifically: NDVI is the normalized vegetation index, i.e., vegetation coverage, and is calculated as follows: ; NIR is near infrared reflectivity, REDRED is red light reflectivity.

[0052] The reason for using logarithmic transformation is to prevent NDVI from changing too drastically when it is close to 0 and to increase the sensitivity of the small value interval. LAI is the leaf area index, which is the leaf area ratio obtained by simulation, obtained by dividing the total leaf area by the surface area; the power function transformation is used ( is a preset empirical coefficient) to enhance the impact of LAI under different vegetation density conditions. H is the vegetation height, which reflects the vegetation growth. It is obtained by calculating the average height of vegetation in the region. Fractional transformation is used to stabilize the contribution of height to avoid unreasonable impact of excessive height of individual plants on the index. S is the species diversity index, which measures the ecological stability of vegetation communities. It is calculated using the Shannon diversity index. All of them are preset weight coefficients, which are determined through data training. The values ​​in different regions may be different. The tanh function is used for normalization. The final vegetation restoration index It is limited to the interval [-1,1] to prevent extreme values ​​from affecting the stability of the model.

[0053] The second common calculation model refers to the soil carbon storage index calculation model. The soil carbon storage index is used to measure the soil carbon sequestration capacity in different regions. Its calculation is based on key factors such as organic carbon accumulation, nitrogen storage, soil structure stability and water retention capacity, combined with nonlinear transformation, normalization processing and soil carbon cycle model to ensure that the calculation results can comprehensively and scientifically reflect the soil carbon storage capacity.

[0054] Based on the soil quality evolution model and ecological simulation model, the following data are obtained: Soil organic carbon content: calculated by soil carbon cycle simulation, reflecting the soil carbon fixation capacity. Soil total nitrogen content: derived from nutrient dynamic simulation, determines soil fertility and organic matter reserves. Soil density: calculated by physical simulation, affects soil carbon sequestration capacity. Soil moisture content: calculated by soil moisture dynamic model, determines soil carbon decomposition rate. For soil carbon state at time t: ; is the current soil organic carbon stock, For time The soil organic carbon stock at the time is the mineralization loss of soil carbon per unit time, is the elapsed time interval.

[0055] Soil carbon storage depends not only on organic carbon input, but also on the physical structure of the soil. Soil density and porosity determine the soil carbon sequestration capacity, which can be calculated as follows: ; is the contribution of soil structure to carbon storage capacity (dimensionless), BD is soil density (g / cm³), the greater the density, the smaller the porosity, and the weaker the carbon sequestration capacity.

[0056] Water affects two key processes of soil carbon storage: an overly dry environment will inhibit microbial activity and reduce carbon decomposition, while an overly wet environment will promote carbon decomposition, leading to faster carbon loss, which can be calculated as follows: ; is the moisture adjustment factor (dimensionless), is the current soil moisture content (%), The most suitable soil moisture content for carbon fixation (%) It is a preset adjustment parameter that determines the degree of influence of moisture on carbon storage capacity; Integrating the above calculation results, nonlinear transformation is used to obtain the final soil carbon storage index: ; N represents the total nitrogen content of the soil (g / m²), which indicates the impact of the current nitrogen reserve on the stability of carbon storage; C represents the current soil organic carbon stock, , , and These are preset influence coefficients used to measure the effects of different parameters on soil carbon storage index. The degree of impact may vary in different regional settings.

[0057] The pre-trained machine learning model is a convolutional neural network model. Convolutional neural network (CNN) is a deep learning model that can automatically extract features, identify patterns, and make nonlinear decisions. It is suitable for complex environmental data analysis. In this method, CNN is used to comprehensively process three soil carbon storage indices and three vegetation restoration indices to calculate the overall ecological restoration carbon sequestration benefit value of the photovoltaic base. The model can capture the nonlinear relationship between the indices of each region and provide a more accurate evaluation than the traditional weighted calculation.

[0058] Structuring of input data: Each photovoltaic base area (the area under the photovoltaic panel, the channel area between the photovoltaic panel rows, and the peripheral area of ​​the photovoltaic field) generates two core indexes: the vegetation restoration index (which measures the growth and coverage of vegetation) and the soil carbon storage index (which measures the carbon fixation capacity of the soil), forming a data matrix with six input features (three areas × two indices), which are input into CNN for calculation.

[0059] Hierarchical structure analysis: CNN is mainly composed of input layer, convolution layer, activation layer, pooling layer and fully connected layer. The functions of each layer are as follows: Input layer: This layer accepts six input data, representing the soil carbon storage index and vegetation restoration index of three regions. These data will be converted into a multidimensional feature matrix for subsequent feature extraction. Convolution layer: The convolution layer is the core part of CNN. It uses multiple filters (also called convolution kernels) to scan the input data to capture the patterns and interactions between the indices of different regions. The main goal of this layer: Identify the relationship between soil carbon storage and vegetation restoration in a specific area. Identify the mutual influence of indices between different regions (such as the potential promotion of soil carbon storage in a certain area by vegetation restoration in a certain area) and discover hidden trends and characteristics of ecological restoration. Activation layer: This layer processes the output of the convolution layer through nonlinear activation functions (such as ReLU), so that the model can learn complex nonlinear relationships. For example: When the vegetation restoration index of a certain area is high, but the soil carbon storage index is low, the model will automatically adjust the weights to evaluate the long-term carbon sequestration potential of the area. If the indices of all regions are on an upward trend, CNN may assign a higher ecological restoration score.

[0060] Pooling layer: The role of this layer is to reduce the data dimension, retain the most critical information, and reduce the computational complexity. In this application scenario, the pooling layer helps to: Aggregate the index data of each region and eliminate small fluctuations between regions. Extract the most representative ecological restoration pattern to improve prediction accuracy. Fully connected layer: This layer expands the pooled features and passes them to the final decision layer. The neurons here perform a comprehensive calculation of the weights of different indexes to generate a final ecological restoration and carbon fixation benefit score. Output layer: This layer is responsible for generating the final scoring result, which is used to quantify the overall ecological restoration and carbon fixation benefits of the photovoltaic base. The score is usually between 0 and 1 (can be normalized to 0-100) to intuitively represent the ecological restoration effect.

[0061] Embodiment 2: A Gobi photovoltaic base ecological restoration carbon fixation benefit assessment system, comprising: The regional division module divides the photovoltaic base area into the area below the photovoltaic panels, the channel area between the photovoltaic panels and the peripheral area of ​​the photovoltaic field; Simulation module, used to establish corresponding ecological simulation models for different regions; A data acquisition module is used to acquire simulation data from the corresponding ecological simulation model for each divided area; The regional assessment module is used to substitute the obtained simulation data into the shared calculation model 1 reflecting the progress and quality of regional vegetation restoration, and the shared calculation model 2 reflecting the soil restoration effect and carbon sequestration capacity; The overall evaluation module is used to summarize the calculation results of all regions and substitute them into the pre-trained machine learning model. The output result is the ecological restoration carbon fixation benefit value used to represent the final score of the overall ecological restoration of the photovoltaic base.

[0062] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0063] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0064] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0065] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0066] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for evaluating the carbon fixation benefits of ecological restoration of a Gobi photovoltaic base, characterized in that: The following steps are involved: The photovoltaic base area is divided into the area under the photovoltaic panels, the channel area between the photovoltaic panels and the peripheral area of ​​the photovoltaic field, and corresponding ecological simulation models are established for different areas; For each divided area, simulation data is obtained from the corresponding ecological simulation model and substituted into the shared calculation model 1 that reflects the progress and quality of regional vegetation restoration, and the shared calculation model 2 that reflects the soil restoration effect and carbon sequestration capacity; The calculation results of all regions are summarized and substituted into the pre-trained machine learning model. The output result is the ecological restoration carbon fixation benefit value used to represent the final score of the overall ecological restoration of the photovoltaic base.

2. A method for evaluating the carbon fixation benefits of Gobi photovoltaic base ecological restoration according to claim 1, characterized in that: The logic of the ecological simulation model for the area under the photovoltaic panels is: The photovoltaic array arrangement is used to simulate the light intensity distribution and construct a light distribution model: Based on the Richards equation, water infiltration, evaporation and migration are simulated to construct a soil moisture dynamic model; The Logistic growth model is used to construct the vegetation growth model; The constructed light distribution model, soil moisture dynamics model and vegetation growth model are used for simulation.

3. A method for evaluating the carbon sequestration benefits of Gobi photovoltaic base ecological restoration according to claim 2, characterized in that: The logic of the ecological simulation model of the inter-row channel area of ​​photovoltaic panels is: The Markov chain model is used to predict vegetation community succession, simulate the competition and succession process of different species under different restoration strategies, and construct a vegetation restoration model; Use soil carbon cycle models to simulate organic matter accumulation and mineralization processes, predict soil carbon storage changes, and build soil quality evolution models; The constructed vegetation restoration model and soil quality evolution model were used for simulation.

4. A method for evaluating the carbon sequestration benefits of Gobi photovoltaic base ecological restoration according to claim 3, characterized in that: The logic of the ecological simulation model in the peripheral area of ​​the photovoltaic field is: Based on the Saltation-Balance equation, the wind-sand movement is simulated and the wind-sand transport model is constructed; Use the RUSLE model to correct soil loss rate, evaluate the role of vegetation in soil protection, and construct a soil erosion model; The constructed wind-sand transport model and soil erosion model are used for simulation.

5. A method for evaluating the carbon fixation benefits of Gobi photovoltaic base ecological restoration according to claim 4, characterized in that: The common calculation model one refers to the vegetation restoration index calculation model.

6. A method for evaluating the carbon fixation benefits of Gobi photovoltaic base ecological restoration according to claim 5, characterized in that: The second shared calculation model refers to the soil carbon storage index calculation model.

7. A method for evaluating the carbon sequestration benefits of Gobi photovoltaic base ecological restoration according to claim 6, characterized in that: The pre-trained machine learning model is a convolutional neural network model.

8. A Gobi photovoltaic base ecological restoration carbon fixation benefit assessment system, used to implement a Gobi photovoltaic base ecological restoration carbon fixation benefit assessment method as described in any one of claims 1 to 7, characterized in that: include: The regional division module divides the photovoltaic base area into the area below the photovoltaic panels, the channel area between the photovoltaic panels and the peripheral area of ​​the photovoltaic field; Simulation module, used to establish corresponding ecological simulation models for different regions; A data acquisition module is used to acquire simulation data from the corresponding ecological simulation model for each divided area; The regional assessment module is used to substitute the obtained simulation data into the shared calculation model 1 reflecting the progress and quality of regional vegetation restoration, and the shared calculation model 2 reflecting the soil restoration effect and carbon sequestration capacity; The overall evaluation module is used to summarize the calculation results of all regions and substitute them into the pre-trained machine learning model. The output result is the ecological restoration carbon fixation benefit value used to represent the final score of the overall ecological restoration of the photovoltaic base.

Images