A dust source management method, device and system based on a high-resolution satellite

Abstract

The application relates to the field of environmental governance, and discloses a dust source management and control method, device and system based on a high-resolution satellite, wherein the method comprises the following steps: S1, acquiring high-resolution satellite remote sensing image data covering a target area, sequentially performing radiation correction, atmospheric correction and geometric correction on the satellite remote sensing image data, extracting spectral features related to a bare land surface area, and inputting the extracted spectral features into a preset land class classification model; the device comprises a remote sensing data processing module, a meteorological data acquisition and processing module, a dust diffusion calculation module, a management optimization and decision-making module, and a management implementation and feedback module; and the system comprises a remote sensing data acquisition unit, a bare land identification and classification unit, a meteorological data processing unit, a dust diffusion prediction unit and an optimization decision-making unit. The application realizes fine regulation and control and efficient resource allocation of dust management by fusing remote sensing images, measured data and a dynamic optimization model.

Description

Technical Field

[0001] This invention relates to the field of environmental governance, specifically to a method, device, and system for controlling dust sources based on high-resolution satellites. Background Technology

[0002] With the acceleration of urbanization and the increase in large-scale construction activities, dust pollution has become increasingly serious, especially in bare land areas. Due to the lack of vegetation cover, bare land areas easily become sources of dust pollution, severely impacting air quality and residents' health. Traditional methods for effectively addressing dust pollution are proving inadequate.

[0003] Currently, traditional dust control mainly relies on manual monitoring and fixed control plans; these methods can identify pollution sources to some extent and formulate control plans based on human experience. In some areas, simulation calculations based on meteorological data have been introduced to predict the pollution diffusion trend of dust sources and take some basic countermeasures. Through these traditional technologies, pollution can be controlled to a certain extent.

[0004] However, existing technologies have some shortcomings. First, they rely on fixed monitoring points, failing to achieve comprehensive and efficient dust monitoring, resulting in the failure to detect pollution sources in a timely manner. Second, traditional treatment schemes often lack flexibility, unable to quickly adjust based on real-time meteorological data and environmental changes, leading to treatment intensity that does not meet actual needs. More importantly, existing technologies also have problems with resource allocation and optimization, resulting in low resource utilization efficiency and a tendency to over- or under-treatment, increasing costs and workload. In contrast, this invention, by combining high-resolution satellite remote sensing imagery, real-time meteorological data, and a dynamic optimization model, can accurately monitor dust sources, flexibly adjust treatment strategies, and rationally allocate treatment resources. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method, device, and system for dust source control based on high-resolution satellites, which solves the problems of limited monitoring range, static control schemes, extensive resource allocation, and lagging feedback mechanisms in existing technologies.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for controlling dust sources based on high-resolution satellites, comprising the following steps:

[0007] S1. Acquire high-resolution satellite remote sensing image data covering the target area, perform radiometric correction, atmospheric correction and geometric correction on the satellite remote sensing image data in sequence, extract spectral features related to the exposed land surface area, and input the extracted spectral features into a preset land type classification model to generate a spatial distribution map containing various types of bare land labels.

[0008] S2. Collect meteorological observation data corresponding to the acquisition time and spatial location of the target area of ​​the satellite remote sensing image data. The meteorological observation data includes wind speed, wind direction, temperature, and relative humidity. The meteorological data is rasterized using a unified spatial and temporal resolution interpolation method to construct a rasterized structure corresponding to the spatial distribution. Figure 1 Meteorological data grid;

[0009] S3. Spatial registration is performed between the spatial distribution maps of various types of bare land and the meteorological data grid. Based on the spatial registration, a dust source diffusion model is constructed. The dust source diffusion model uses the spatial location and type information of the bare land units and the meteorological parameters of their corresponding grids as input variables to calculate the amount of dust transmitted from each bare land unit to the target receiving point.

[0010] S4. Based on the calculated dust transmission amount from each bare land unit to the target receiving point, and combined with the manageability parameters of each bare land unit and the total amount of management resources, a management optimization model is constructed. The management optimization model is used to allocate the management intensity coefficient of each bare land unit.

[0011] S5. Based on the governance intensity coefficient of each bare land unit, generate a corresponding governance operation instruction dataset, and periodically receive the measured concentration data collected by the dust monitoring equipment deployed in the target area. Feed the measured concentration data back to the governance optimization model to update relevant parameters and solve iteratively.

[0012] Preferably, in step S1, performing radiometric correction, atmospheric correction, and geometric correction on the satellite remote sensing image data sequentially includes:

[0013] Radiometric correction is performed on the original image data using absolute calibration coefficients to obtain apparent radiance;

[0014] Atmospheric correction was performed using an atmospheric transport model to eliminate atmospheric aerosols and water vapor, and surface reflectance images were obtained.

[0015] Geometric correction based on ground control points and digital terrain models ensures that satellite remote sensing image data meets spatial accuracy requirements.

[0016] Preferably, in step S1, generating a spatial distribution map containing various types of bare land markings includes:

[0017] Typical band combinations of remote sensing images are extracted as classification features;

[0018] The classification features are input into a pre-trained supervised classification model, which is built based on support vector machines; the spatial distribution map is classified pixel by pixel, and the bare land type labeling results are output, which include the bare land type and its geographic location encoding information.

[0019] Preferably, in step S2, the rasterization processing of the meteorological data includes:

[0020] The discrete meteorological observation data is converted into a continuous spatial meteorological field based on inverse distance weighting interpolation.

[0021] The inverse distance weighted interpolation results are aligned with the spatial distribution map at a uniform spatial resolution to form a multi-channel meteorological raster dataset containing variables such as wind speed, wind direction, temperature, and relative humidity.

[0022] Preferably, in step S3, the spatial registration of the spatial distribution maps of various types of bare land with the meteorological data grid includes:

[0023] A unified spatial projection coordinate system is applied to the spatial distribution map and meteorological data raster, and the data is resampled to a consistent spatial resolution. Based on the spatial projection coordinates, each bare land unit is matched with its corresponding meteorological raster unit to achieve a one-to-one spatial correspondence.

[0024] Preferably, in step S3, calculating the dust transmission amount from each bare land unit to the target receiving point includes:

[0025] The location and type of the bare land unit and its corresponding meteorological parameters are input into the dust source diffusion model, which is constructed based on the principles of mass conservation and convection diffusion.

[0026] The dust transport volume is expressed as an integral as follows:

[0027]

[0028] In the formula, T i Q represents the amount of dust transported from the i-th bare land unit to the target point P; i The intensity of dust sources in the bare land unit; Its wind speed vector; d iP is the transmission path distance from the target point; f is a function of wind speed direction and path.

[0029] Preferably, in step S4, constructing the governance optimization model includes:

[0030] Set the objective function to minimize the total dust concentration at the target receiving point:

[0031]

[0032] In the formula, α i T is the treatment intensity coefficient for the i-th bare land unit, ranging from 0 to 1, representing the treatment intensity ratio; i Let be the amount of dust transported from the i-th bare land unit to the target point P; n is the number of observation points.

[0033] And set the following constraints:

[0034] Overall governance resource constraints:

[0035] Unit governance boundary constraint: 0 ≤ α i ≤1;

[0036] In the formula, c i Let C be the governance cost of the i-th unit. max This represents the total available governance resources.

[0037] Preferably, in step S5, updating the relevant parameters and iteratively solving includes:

[0038] The error is calculated between the measured concentration data of the target receiving point collected by the dust monitoring equipment and the model prediction value.

[0039] The dust source intensity estimate is corrected based on error calculation;

[0040] The governance optimization model is rerun using a rolling optimization approach to solve for the updated optimal governance strength coefficient.

[0041] The present invention also provides a dust source control device based on high-resolution satellites, comprising:

[0042] The remote sensing data processing module is used to acquire remote sensing image data from high-resolution satellites, perform radiometric correction, atmospheric correction and geometric correction on the remote sensing image data, extract spectral features related to bare land, and generate a spatial distribution map containing bare land type labels.

[0043] The meteorological data acquisition and processing module is used to acquire meteorological data of the spatial location of the target area in real time, and to perform rasterization processing on the meteorological data to construct meteorological grid data consistent with the remote sensing data;

[0044] The dust source diffusion calculation module calculates the amount of dust transmitted from each bare land unit to the target receiving point based on the remote sensing image data and the meteorological grid data, and evaluates the contribution of each bare land unit to dust pollution based on the amount of dust transmitted.

[0045] The governance optimization and decision-making module calculates and outputs the optimal dust control scheme based on the calculated dust pollution contribution and total amount constraints, and generates a governance intensity coefficient.

[0046] The governance implementation and feedback module is used to implement governance measures according to the governance intensity coefficient scheme, and to receive the measured concentration data of dust monitoring equipment in the target area in real time, and to feed the monitoring data back to the governance optimization model to update the governance strategy.

[0047] This invention also provides a dust source control system based on high-resolution satellites, comprising:

[0048] The remote sensing data acquisition unit is used to acquire remote sensing image data from high-resolution satellites and preprocess the remote sensing image data to extract spectral features related to bare land areas.

[0049] The bare land identification and classification unit classifies remote sensing images based on the spectral features, identifies various types of bare land within the target area, and generates a spatial distribution map.

[0050] The meteorological data processing unit is used to acquire real-time meteorological data of the target area and perform rasterization processing on the meteorological data to construct a spatial distribution of the bare land. Figure 1 Meteorological data grid;

[0051] The dust source diffusion prediction unit is used to construct a dust source diffusion model based on the bare land type and meteorological data grid, and to predict the contribution of each bare land unit to dust pollution in the target area.

[0052] The optimization decision-making unit is used to optimize dust source control schemes and output corresponding control measures based on dust source diffusion models and total emission constraints.

[0053] This invention provides a method, device, and system for dust source control based on high-resolution satellite imagery. It offers the following advantages:

[0054] 1. This invention employs a technical solution combining high-resolution satellite remote sensing imagery with real-time meteorological data, achieving the technical effect of accurately identifying bare land areas and monitoring dust concentration in real time. Compared with traditional manual monitoring methods in the prior art, this invention can significantly improve the monitoring efficiency of dust sources and reduce the complexity and error of manual operation.

[0055] 2. This invention introduces a feedback mechanism for dynamically updating the governance optimization model. It adjusts the governance intensity coefficient using periodically collected dust monitoring data, achieving an adaptive optimization of the governance strategy. Unlike the fixed approach of traditional methods, this invention can automatically optimize and adjust governance measures based on measured concentration data. This not only improves the governance effect but also effectively avoids the problem of ineffective governance due to environmental changes, ensuring the long-term stability of dust control.

[0056] 3. This invention combines an optimization model of governance intensity coefficient and resource constraints to rationally allocate governance resources, achieving the effect of maximizing the utilization of governance resources. Compared with existing governance methods that suffer from significant resource waste, this invention, through intelligent resource allocation, ensures that the governance intensity of each bare land unit meets actual needs, effectively saving governance costs and improving resource utilization efficiency.

[0057] 4. This invention achieves refined dust pollution source control through precise dust transport calculation and iterative model optimization. Unlike traditional methods that rely solely on localized treatment, this invention considers the dust transport volume of each bare land unit and its impact on the target receiving point, making dust control more targeted and comprehensive. This not only improves the accuracy of the control effect but also reduces ineffective treatment. Attached Figure Description

[0058] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0059] Figure 2 This is a diagram of the device architecture of the present invention;

[0060] Figure 3 This is a system architecture diagram of the present invention. Detailed Implementation

[0061] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0062] Please see the appendix Figure 1 This invention provides a method for controlling dust sources based on high-resolution satellites, comprising the following steps:

[0063] S1. Acquire high-resolution satellite remote sensing image data covering the target area, perform radiometric correction, atmospheric correction and geometric correction on the satellite remote sensing image data in sequence, extract spectral features related to the exposed land surface area, and input the extracted spectral features into a preset land type classification model to generate a spatial distribution map containing various types of bare land labels.

[0064] First, high-resolution remote sensing image data covering the target area is acquired using domestically produced high-resolution satellites (such as the Gaofen series satellites, the Ziyuan series satellites, Beijing-2, and Jilin-1). Generally, the spatial resolution of the image data is no less than 1 meter to ensure effective identification of the spatial distribution of bare land within the target area.

[0065] After receiving satellite remote sensing image data, radiometric correction is first performed to remove radiometric distortions caused by satellite sensors, observation conditions, etc. Generally, radiometric correction is achieved using absolute calibration coefficients. Specifically, the following radiometric correction formula is used:

[0066] L calibrated =G·(DN-B);

[0067] In the formula, L calibrated Here, denoted as DN, represents the radiometrically corrected image brightness value; G is the sensor's gain factor; B is the offset value; and DN is the original digital value. This formula converts the digital value (DN) captured by the sensor into the corresponding radiometric brightness to eliminate radiometric errors in satellite imagery.

[0068] After radiometric correction, atmospheric correction is performed. The purpose of atmospheric correction is to remove the influence of atmospheric aerosols, water vapor, and other factors on the remote sensing image, ensuring that the acquired image reflectance more closely approximates the actual surface features. In this embodiment, an atmospheric transfer model, such as the 6S model or the MODTRAN model, is used to perform atmospheric correction. The model calculates the atmospheric radiative transfer factor based on atmospheric parameters such as aerosol optical thickness, humidity, and air pressure, removing atmospheric influences and ultimately obtaining the actual surface reflectance image. Specifically, the surface reflectance R... surface The following formula can be used for calculation:

[0069]

[0070] In the formula, L measured These are satellite measurements after atmospheric effects; L atmospheric The light is scattered and reflected by the atmosphere; θ is the angle of incidence of the sun; T sun R is the surface transfer factor of solar radiation; surface It represents the surface reflectance.

[0071] After the satellite imagery data quality processing is completed, geometric correction is performed. Geometric correction is used to eliminate geometric distortions caused by satellite sensor attitude, Earth curvature, and ground undulations. By aligning with ground control points (GCPs) and combining distortion correction with a digital elevation model (DEM), accurate spatial matching of the imagery is achieved. In this embodiment, methods such as linear transformation and quadratic curve fitting are used for coordinate transformation to ensure that the satellite imagery has high spatial accuracy and meets the requirements of the target area's geographic coordinate system.

[0072] After the radiometric, atmospheric, and geometric correction steps for the image are completed, the next step is to extract spectral features from the processed remote sensing image. Specifically, representative spectral bands (such as red, near-infrared, and short-wave infrared) are selected and combined to construct a feature space. Generally, the reflectance characteristics of these spectral bands are used to analyze the spectral responses of different materials on the Earth's surface, thereby extracting spectral features related to bare land.

[0073] For the classification of spectral features, this invention employs a pre-defined land use classification model to automatically identify bare land types. Alternatively, the classification model can be a supervised classification model, such as a Support Vector Machine (SVM). This model uses samples from the training set data to classify based on spectral features. By analyzing the spectral information of each pixel in the input image, the model can identify different types of bare land, including but not limited to construction bare land, unpaved bare land, and green bare land. The model output is a land use classification map, which labels the spatial distribution of different bare land types.

[0074] Generally, land use classification models determine the land use attribute of each pixel using extracted spectral features, such as the ratio of red to near-infrared reflectance and the Normalized Difference Vegetation Index (NDVI), through pixel-by-pixel classification. Specifically, NDVI can be calculated using the following formula:

[0075]

[0076] In the formula, NIR is the near-infrared reflectance; R is the red reflectance; and NDVI is the normalized vegetation index. Based on the NDVI value, the model can determine whether the pixel is bare land, vegetation, or other types of land cover.

[0077] The bare land type information obtained from the classification model can be used to generate a spatial distribution map containing annotations for various types of bare land. In this map, each cell is not only labeled with land type information, but also identified by color or label to indicate the type and location of the bare land. For the target area, the spatial distribution map can intuitively show the spatial distribution of different types of bare land, effectively providing an important reference for subsequent dust source analysis and control.

[0078] S2. Collect meteorological observation data corresponding to the acquisition time and spatial location of the target area of ​​the satellite remote sensing image data. The meteorological observation data includes wind speed, wind direction, temperature, and relative humidity. The meteorological data is rasterized using a unified spatial and temporal resolution interpolation method to construct a rasterized structure corresponding to the spatial distribution. Figure 1 The acquisition time of satellite remote sensing imagery is closely related to the spatial location of the target area; therefore, accurate meteorological observation data is crucial for subsequent dust dispersion prediction and pollution contribution calculation. Specifically, the goal of this step is to collect meteorological observation data that matches the acquisition time of the satellite remote sensing imagery and the spatial location of the target area, and to transform this data into a grid that corresponds to the spatial distribution of the remote sensing imagery through rasterization. Figure 1 A consistent meteorological data grid is used. This method ensures the spatial and temporal consistency between meteorological data and remote sensing imagery data, providing necessary support for the subsequent construction of diffusion models.

[0079] Generally, meteorological data includes variables such as wind speed, wind direction, temperature, and relative humidity. These meteorological parameters are key factors affecting dust dispersion and pollution propagation. To ensure the spatiotemporal consistency of the data, the acquisition of meteorological data must be closely matched with the acquisition time of remote sensing imagery and the spatial location of the target area.

[0080] In this embodiment, meteorological observation data is collected in real time by multiple meteorological stations or sensors, covering the target area and its surrounding environment. The collected meteorological data must not only meet spatial distribution requirements but also temporal synchronization requirements. Therefore, meteorological data from several hours before and after the satellite image acquisition time are typically collected and processed to ensure accuracy.

[0081] Rasterization is a key technical step in this embodiment. Since meteorological observation data usually comes from scattered observation points and is often discrete, interpolation methods are needed to convert the discrete observation data into consistent gridded data so that it can be registered and analyzed with remote sensing image data.

[0082] In this embodiment, a unified spatial resolution interpolation method is selected to interpolate discrete meteorological data points into a regular grid. Commonly used interpolation methods include inverse distance weighted (IDW) interpolation and Kriging interpolation. This interpolation method ensures a smooth spatial transition for each meteorological parameter (wind speed, wind direction, temperature, and humidity, etc.), allowing each grid cell to obtain the corresponding meteorological data.

[0083] For example, the inverse distance weighted interpolation method performs interpolation calculations based on the following formula:

[0084]

[0085] In the formula, Z(x,y) represents the meteorological value at the target point; w i Let d be the meteorological value at the i-th observation point; i is the distance between the target point and the i-th observation point; n is the number of observation points.

[0086] To ensure temporal consistency between meteorological data and satellite remote sensing imagery, some embodiments employ temporal interpolation methods. Specifically, if the temporal resolution of the meteorological data does not perfectly match that of the remote sensing imagery, linear interpolation or spline interpolation can be used to adjust the time of the meteorological data. For example, assuming the satellite imagery was acquired at 10:00, while the meteorological data exists between 09:30 and 10:30, the following linear interpolation formula can be used to adjust the meteorological data:

[0087] In the formula, T interpThe interpolation result is for the target time point; T1 and T2 are the meteorological observation data from the two consecutive observations; t1 and t2 are the corresponding times; t target The target time point (e.g., satellite image acquisition time 10:00).

[0088] After interpolation, the meteorological data will be converted into a raster format, resulting in meteorological data with higher spatial resolution. For example, if the spatial resolution of the remote sensing image data is 500 meters, the spatial resolution of the meteorological data will also be adjusted to 500 meters. The rasterization result of each meteorological parameter will become a two-dimensional array (meteorological data grid), with each cell representing the meteorological data of the target area at that location.

[0089] The spatial distribution maps of these meteorological data grids and remote sensing images are registered in subsequent steps. In this way, the meteorological data can be processed and analyzed in the spatial coordinate system corresponding to the remote sensing image data, thereby providing accurate data support for subsequent dust dispersion modeling, pollution source contribution calculation, and other tasks.

[0090] In one possible implementation, in addition to basic wind speed, wind direction, temperature, and humidity, other meteorological parameters (such as air pressure and solar radiation) can be added as needed to enhance the model's accuracy. These additional meteorological data also require the same interpolation processing and synchronization with the same spatial and temporal resolution as the remote sensing imagery.

[0091] As an option, the accuracy and method of meteorological data acquisition and interpolation can be flexibly adjusted according to the requirements of the actual application scenario. For example, in some special cases, the temporal resolution of meteorological data may be low. In this case, historical meteorological data and empirical models can be combined to extrapolate the missing meteorological data to avoid model bias caused by missing data.

[0092] S3. Spatial registration is performed between the spatial distribution maps of various types of bare land and the meteorological data grid. Based on the spatial registration, a dust source diffusion model is constructed. The dust source diffusion model uses the spatial location and type information of the bare land units and the meteorological parameters of their corresponding grids as input variables to calculate the amount of dust transmitted from each bare land unit to the target receiving point.

[0093] In general, to ensure data consistency, it is necessary to unify the coordinate projection system and spatial resolution of the two types of data.

[0094] Specifically, the remote sensing classification results and meteorological data are subjected to projection transformation using the same spatial reference system (such as WGS-84 or UTM projection). Subsequently, based on the resolution of the spatial distribution map of the remote sensing image, a resampling operation is performed on the meteorological data grid. This resampling can use bilinear interpolation or nearest-neighbor interpolation methods to preserve the spatial continuity and numerical stability of meteorological parameters.

[0095] In one possible implementation, a spatial index matrix is ​​constructed to achieve a cell-by-cell mapping between bare land cells and meteorological grids. That is, each bare land cell is bound to the meteorological grid cell at its location, thereby obtaining the meteorological attributes corresponding to that cell, including wind speed vector, wind direction angle, temperature, and relative humidity.

[0096] After spatial registration is completed, the next step is to construct the dust source diffusion model.

[0097] In this embodiment, the dust source diffusion model uses the location, land type attributes, and meteorological conditions of each bare land unit as input parameters to simulate the amount of pollutants transmitted to a specific target receiving point downwind.

[0098] Generally, this model follows the mass conservation and convection-diffusion theory of atmospheric diffusion processes.

[0099] The dust transport volume is expressed as an integral as follows:

[0100]

[0101] In the formula, T i Q represents the amount of dust transported from the i-th bare land unit to the target point P; i The intensity of dust sources in the bare land unit; Its wind speed vector; d iP is the transmission path distance from the target point; f is a function of wind speed direction and path.

[0102] Furthermore, in another implementation, if the meteorological data has a high temporal resolution (e.g., hourly), the above integral calculation can be discretized using a piecewise accumulation method, i.e.:

[0103]

[0104] In the formula, T i Δt represents the dust transport amount from the i-th bare land unit to target point P; N represents the number of time segments; Δt k T represents the k-th time interval; i Let be the amount of dust transported from the i-th bare land unit to the target point P; Its wind speed vector; d iPLet f be the transmission path distance from the target point; f is a function of wind speed direction and path. Let be the amount of dust transported by the i-th bare land unit during the k-th time interval.

[0105] In practical applications, the target receiving point P can be set as a sensitive point, a residential area, or a monitoring point. The final sequence is the assessment result of the contribution of each bare land unit to the pollution at that point, providing input basis for subsequent governance decisions and resource allocation.

[0106] S4. Based on the calculated dust transmission amount from each bare land unit to the target receiving point, and combined with the manageability parameters of each bare land unit and the total amount of management resources, a management optimization model is constructed. The management optimization model is used to allocate the management intensity coefficient of each bare land unit.

[0107] Having calculated the dust transmission volume from each bare land unit to the target receiving point, and considering variables such as wind speed, bare land type, and meteorological conditions, these calculations provide crucial information for subsequent remediation decisions. However, the sheer amount of pollution contribution cannot directly guide the formulation of remediation plans, as remediation measures are constrained by remediation resources and manageability parameters. Therefore, based on this, we further constructed a remediation optimization model to minimize dust pollution and rationally allocate remediation resources.

[0108] In this embodiment, the core objective of the governance optimization model is to maximize the governance effect of dust pollution by rationally configuring the governance intensity coefficient, under the constraints of limited governance resources and the governance potential of each bare land unit. Specifically, the model will comprehensively consider the pollution contribution, manageability, and resource allocation constraints of each bare land unit to generate an optimal governance intensity allocation scheme, thereby guiding actual governance operations.

[0109] First, based on the dust transport volume T calculated in the previous steps... i Based on the manageability parameters of bare land units, a management optimization model is constructed. Specifically, the goal of the management optimization model is to achieve the optimal management scheme by minimizing the total dust concentration at the target receiving point.

[0110] Generally, the objective function is the weighted sum of dust concentrations, specifically defined as:

[0111]

[0112] In the formula, α i T is the treatment intensity coefficient for the i-th bare land unit, ranging from 0 to 1, representing the treatment intensity ratio; i Let be the amount of dust transported from the i-th bare land unit to the target point P; n is the number of observation points.

[0113] In this model, the objective function reflects the relationship between the dust contribution and the intensity of treatment for all bare land units. This is achieved by adjusting α... i The value is used to reduce the pollution load at the target receiving point.

[0114] As an alternative, given the limited resources available for actual governance, resource constraints are typically set to avoid over-concentration of governance resources. Specific constraints are as follows:

[0115] Overall governance resource constraints:

[0116] In the formula, c i Let C be the governance cost of the i-th unit. max This represents the total available governance resources.

[0117] This constraint ensures that resources are allocated reasonably during the governance process and do not exceed the upper limit of available resources.

[0118] In addition, boundary conditions for the unit governance intensity need to be set to ensure that the governance intensity coefficient α of each bare land unit is within the specified range. i The value is in the range [0,1].

[0119] Unit governance boundary constraint: 0 ≤ α i ≤1.

[0120] In some embodiments, to enhance the flexibility of the model, the governance strength coefficient α i Weighted adjustments may be made based on the characteristics of the bare land unit (such as area, land type, dust source intensity, etc.). For example, some bare land units may be allocated higher treatment intensity because they are easier to treat or have better pollution reduction effects after treatment.

[0121] Specifically, assuming a bare land unit has a low governance cost c i and higher pollution contribution T i Then the unit may be assigned a higher governance intensity coefficient α. i This is to prioritize the allocation of limited governance resources to this unit, thereby achieving maximum pollution reduction.

[0122] In further optimization, mathematical tools such as linear programming or integer programming can be used to solve the objective function. Specifically, by calculating different α... i The optimal governance intensity allocation strategy is obtained by combining the following methods to satisfy the total governance resource constraint C. max And minimize the total dust concentration at the target receiving point.

[0123] In one possible implementation, the optimization process gradually adjusts α using numerical methods (such as gradient descent or simulated annealing). iTo obtain the optimal solution, an initial governance intensity coefficient value can be set (such as all zeros or uniform distribution), and then iteratively adjusted until the optimization objective and constraints are met.

[0124] In practical applications, if subsequent monitoring reveals that the dust concentration at the target receiving point fails to meet expectations, or that resources are not used efficiently, the parameters of the governance optimization model can be adjusted or the governance intensity coefficient can be re-solved based on real-time data feedback to achieve dynamic optimization and continuous improvement.

[0125] S5. Based on the governance intensity coefficient of each bare land unit, generate a corresponding governance operation instruction dataset, and periodically receive the measured concentration data collected by the dust monitoring equipment deployed in the target area. Feed the measured concentration data back to the governance optimization model to update relevant parameters and solve iteratively.

[0126] The generation of the remediation operation instruction dataset begins with the remediation intensity coefficients obtained through the aforementioned steps. These coefficients, as key parameters for measuring the remediation intensity of each bare land unit, will be specifically reflected in subsequent remediation operation instructions. Generally, the generation of remediation operation instructions requires the formulation of different remediation measures based on the remediation intensity coefficient of each bare land unit. Remediation measures include, but are not limited to, soil improvement, vegetation restoration, water application, and mulch application. These instructions will specify the operation type, operation intensity, and required resources for each bare land unit.

[0127] Specifically, for each bare land unit, its remediation intensity coefficient will directly affect the intensity and scale of remediation operations. For example, for bare land units with a higher remediation intensity coefficient, the directives may require greater resource input, such as increasing the frequency of water spraying or increasing vegetation planting density. Conversely, for bare land units with a lower remediation intensity coefficient, the intensity of remediation measures will be relatively lower. These directives will guide the implementation of remediation measures in practice.

[0128] Furthermore, with the deployment of dust monitoring equipment and the collection of real-time data, the pollution control optimization model will periodically receive measured concentration data within the target area. Generally, dust monitoring equipment collects dust concentration data from the target area through sensors and feeds this data back to the pollution control optimization system. This measured data mainly includes environmental factors such as dust concentration, wind speed, humidity, and temperature, and is compared with pre-set model parameters.

[0129] In one possible implementation, the governance optimization model compares the measured concentration data with the predicted concentration data to identify errors. In this way, the system can correct for factors such as source intensity and transmission path calculated in the model, thereby optimizing the governance strategy. If the feedback data indicates that the current governance intensity coefficient is insufficient to achieve the expected governance effect, the system will automatically adjust the governance intensity coefficient and rerun the governance optimization model.

[0130] In some embodiments, the feedback data update process is performed through rolling optimization. Specifically, the rolling optimization method gradually adjusts the treatment intensity coefficient of each bare land unit based on the real-time received monitoring data to reduce dust concentration and optimize treatment effectiveness. Each optimization update updates the relevant parameters in the model based on the latest measured data, thereby enabling the treatment scheme to continuously adapt to environmental changes and improve treatment efficiency.

[0131] In pollution control optimization models, there are generally two main optimization objectives: first, to minimize the dust concentration at the target receiving point; and second, to allocate a pollution control intensity coefficient to each bare land unit based on total resource constraints. Based on feedback from measured concentration data, the model continuously adjusts the pollution control intensity coefficient for each bare land unit until the optimal effect is achieved. The pollution control intensity coefficient typically ranges from 0 to 1, representing the intensity of pollution control measures. By continuously adjusting these coefficients, the system can ensure the optimal configuration of pollution control measures and achieve continuous improvement in regional pollution control effectiveness.

[0132] By combining real-time monitoring data feedback with iterative optimization model solutions, an adaptive and dynamically adjustable dust control solution is provided. By generating an operational instruction dataset based on a control intensity coefficient and periodically receiving and feeding back measured concentration data, continuous optimization during dust control can be ensured, thereby achieving more efficient and precise control results.

[0133] The dust source control device based on high-resolution satellite described below and the dust source control method based on high-resolution satellite described above can be referred to in correspondence.

[0134] Please see the appendix Figure 2 The present invention also provides a dust source control device based on high-resolution satellites, comprising:

[0135] The remote sensing data processing module is used to acquire remote sensing image data from high-resolution satellites, perform radiometric correction, atmospheric correction and geometric correction on the remote sensing image data, extract spectral features related to bare land, and generate a spatial distribution map containing bare land type labels.

[0136] The meteorological data acquisition and processing module is used to acquire meteorological data of the spatial location of the target area in real time, and to perform rasterization processing on the meteorological data to construct meteorological grid data consistent with the remote sensing data;

[0137] The dust dispersion calculation module calculates the amount of dust transmitted from each bare land unit to the target receiving point based on the remote sensing image data and the meteorological grid data, and evaluates the contribution of each bare land unit to dust pollution based on the amount of dust transmitted.

[0138] The governance optimization and decision-making module calculates and outputs the optimal dust control scheme based on the calculated dust pollution contribution and total amount constraints, and generates a governance intensity coefficient.

[0139] The governance implementation and feedback module is used to implement governance measures according to the governance intensity coefficient scheme, and to receive the measured concentration data of dust monitoring equipment in the target area in real time, and to feed the monitoring data back to the governance optimization model to update the governance strategy.

[0140] The device in this embodiment can be used to execute the above method embodiments, and its principle and technical effects are similar, so they will not be described again here.

[0141] The dust source control system based on high-resolution satellites described below and the dust source control method based on high-resolution satellites described above can be referred to and correspond to each other.

[0142] Please see the appendix Figure 3 The present invention also provides a dust source control system based on high-resolution satellites, comprising:

[0143] The remote sensing data acquisition unit is used to acquire remote sensing image data from high-resolution satellites and preprocess the remote sensing image data to extract spectral features related to bare land areas.

[0144] The bare land identification and classification unit classifies remote sensing images based on the spectral features, identifies various types of bare land within the target area, and generates a spatial distribution map.

[0145] The meteorological data processing unit is used to acquire real-time meteorological data of the target area and perform rasterization processing on the meteorological data to construct a spatial distribution of the bare land. Figure 1 Meteorological data grid;

[0146] The dust diffusion prediction unit is used to construct a dust source diffusion model based on the bare land type and meteorological data grid, and to predict the contribution of each bare land unit to dust pollution in the target area.

[0147] The optimization decision-making unit is used to optimize dust source control schemes and output corresponding control measures based on dust source diffusion models and total emission constraints.

[0148] The system in this embodiment can be used to execute the above method embodiments, and its principle and technical effect are similar, so they will not be described again here.

[0149] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A dust source management method based on high-resolution satellite, characterized in that, The method comprises the following steps: S1, obtaining high-resolution satellite remote sensing image data covering a target area, sequentially performing radiation correction, atmospheric correction and geometric correction on the satellite remote sensing image data, extracting spectral features related to bare land areas, and inputting the extracted spectral features into a preset land class classification model to generate a spatial distribution map containing labels of various types of bare land; S2, collecting meteorological observation data corresponding to the time of obtaining the satellite remote sensing image data and the spatial position of the target area, the meteorological observation data including wind speed, wind direction, temperature and relative humidity, and performing gridding processing on the meteorological observation data by a unified spatial and temporal resolution interpolation method to construct a meteorological observation data grid consistent with the spatial distribution map; S3, spatially registering the spatial distribution map of various types of bare land labels and the meteorological observation data grid, and constructing a dust source diffusion model based on the spatial registration, the dust source diffusion model taking the spatial position, type information of the bare land unit and the meteorological parameters of the corresponding grid as input variables to calculate the dust transmission amount of each bare land unit to a target receiving point; S4, based on the calculated dust transmission amount of each bare land unit to the target receiving point, combining the treatability parameters of each bare land unit and the total amount constraint condition of treatment resources, constructing a treatment optimization model, the treatment optimization model being used to distribute the treatment intensity coefficient of each bare land unit; S5, based on the treatment intensity coefficient of each bare land unit, generating a corresponding treatment operation instruction data set, and periodically receiving the measured concentration data collected by the dust monitoring equipment arranged in the target area, and feeding back the measured concentration data to the treatment optimization model to update the related parameters and iteratively solve; In step S3, the calculation of the dust transmission amount of each bare land unit to the target receiving point comprises: inputting the position, type and corresponding meteorological parameters of the bare land unit into the dust source diffusion model, the dust source diffusion model being constructed based on the principles of mass conservation and convection diffusion; the dust transmission amount is expressed by the following integral: ； wherein is the dust emission rate of the i-th bare land unit; is the dust emission rate of the i-th bare land unit; is the dust emission rate of the i-th bare land unit; is the wind speed vector of the i-th bare land unit; is the transmission path distance of the i-th bare land unit to the target point P; is the wind speed direction and the path function.

2. The method according to claim 1, wherein, In step S1, sequentially performing radiation correction, atmospheric correction and geometric correction on the satellite remote sensing image data comprises: performing radiation correction on the original image data by using an absolute calibration coefficient to obtain apparent radiance brightness; performing atmospheric correction by using an atmospheric transmission model to eliminate atmospheric aerosols and water vapor to obtain a ground reflectance image; performing geometric correction based on ground control points and a digital terrain model to make the satellite remote sensing image data meet the spatial accuracy requirements.

3. The method according to claim 1, wherein, In step S1, the generation of the spatial distribution map containing labels of various types of bare land comprises: extracting a typical band combination of the remote sensing image as a classification feature; inputting the classification feature into a pre-trained supervised classification model, the supervised classification model being constructed based on a support vector machine; performing pixel-by-pixel classification on the spatial distribution map to output a bare land type labeling result, the bare land type labeling result containing bare land type and geographic location encoding information.

4. The method according to claim 1, wherein, In step S2, the gridding processing of the meteorological observation data comprises: converting discrete meteorological observation point data into a continuous spatial meteorological field based on inverse distance weighting interpolation; The inverse distance weighted interpolation result is aligned with the spatial distribution map at a uniform spatial resolution, forming a variable multi-channel meteorological grid data set containing wind speed, wind direction, temperature and relative humidity.

5. The method according to claim 1, wherein, In step S3, the spatial registration of the spatial distribution map of each type of bare land and the meteorological observation data grid comprises: applying a uniform spatial projection coordinate system to the spatial distribution map and the meteorological observation data grid and resampling to a consistent spatial resolution; matching each bare land unit to its corresponding meteorological grid unit based on the spatial projection coordinates to achieve a one-to-one spatial correspondence.

6. The method according to claim 1, wherein, In step S4, the construction of the governance optimization model comprises: setting a target function to minimize the total dust concentration at the target receiving point: ； In the formula, is the governance intensity coefficient of the first bare land unit, and the value range is 0 to 1, indicating the governance intensity proportion; is the dust transmission amount of the first bare land unit to the target point P; is the number of observation points; and setting the following constraint conditions: Total governance resource constraints: ; Cell governance boundary constraints: ; In the formula, is the first unit governance cost, is the total available governance resources.

7. The method according to claim 1, wherein the method further comprises: determining the location of the dust source based on the satellite image; and determining the location of the dust source based on the satellite image. In step S5, the updating of the related parameters and the iterative solution comprises: error calculation of the measured concentration data of the target receiving point collected by the dust monitoring equipment and the model predicted value; correction of the dust source intensity estimate value based on the error calculation; re-running the governance optimization model using a rolling optimization method and solving the updated optimal governance intensity coefficient.

8. A high-resolution satellite-based dust source management device according to any one of claims 1-7, characterized in that, It comprises: a remote sensing data processing module for acquiring high-resolution satellite remote sensing image data, and performing radiation correction, atmospheric correction and geometric correction on the remote sensing image data, extracting spectral features related to bare land, and generating a spatial distribution map containing bare land type labeling; a meteorological data acquisition and processing module for real-time acquisition of meteorological observation data at the spatial position of the target area, and gridding processing of the meteorological observation data to construct meteorological grid data consistent with the remote sensing data; a dust source diffusion calculation module for calculating the dust transmission amount of each bare land unit to the target receiving point based on the remote sensing image data and the meteorological grid data, and evaluating the contribution of each bare land unit to dust pollution according to the dust transmission amount; a governance optimization and decision-making module for calculating and outputting the optimal dust control scheme based on the calculated dust pollution contribution and total amount constraint condition, and generating a governance intensity coefficient; a governance implementation and feedback module for implementing governance measures according to the governance intensity coefficient scheme, and real-time receiving of measured concentration data of dust monitoring equipment in the target area, and feeding back the monitoring data to the governance optimization model to update the governance strategy.

9. A high-resolution satellite-based dust source management system, according to any one of claims 1-7, wherein, It comprises: a remote sensing data acquisition unit for acquiring high-resolution satellite remote sensing image data, and preprocessing the remote sensing image data to extract spectral features related to bare land regions; a bare land identification and classification unit for classifying remote sensing images based on the spectral features, identifying various types of bare land in the target area and generating a spatial distribution map; a meteorological data processing unit for acquiring real-time meteorological observation data of the target area and gridding processing the meteorological observation data to construct a meteorological observation data grid consistent with the bare land spatial distribution map; a dust source diffusion prediction unit for constructing a dust source diffusion model based on the bare land type and the meteorological observation data grid, and predicting the transmission amount of each bare land unit to dust pollution in the target area; an optimization decision unit for optimizing the dust source control scheme and outputting the corresponding governance measures according to the dust source diffusion model and the total amount constraint condition.

Images