Toughness evaluation method for ecological maintenance system of corrosive saline soil site

Abstract

The invention relates to the technical field of ecological toughness analysis, and discloses a toughness evaluation method for an ecological maintenance system of a corrosive saline soil site, which comprises the following steps: dividing a to-be-detected region into a plurality of initial to-be-detected units with the same area, and acquiring a first detection data set including soil characteristics, vegetation growth, microorganisms and environmental climate data of the initial to-be-detected units; comparing and analyzing the adjacent units, classifying the adjacent units into a plurality of to-be-detected unit sets, selecting representative to-be-detected units, and taking data of the representative to-be-detected units as a first detection data set of the to-be-detected units; and preprocessing the first detection data set to obtain a second detection data set, and calculating a toughness evaluation value of each to-be-detected unit set. Performing rationality judgment on the evaluation value, and if the evaluation value is reasonable, outputting the evaluation value and a corresponding range; and if not, recalculation is carried out, and if not, an alarm signal is sent out. According to the method, the ecological toughness of the corrosive saline soil site can be evaluated more accurately and stably, and a scientific basis is provided for ecological management and restoration.

Description

Technical Field

[0001] The present invention relates to the technical field of ecological resilience analysis, and more specifically, to a method for evaluating the resilience of an ecological maintenance system for corrosive saline soil sites. Background Art

[0002] Corrosive saline soil sites are widely distributed in coastal, arid, and industrial pollution areas. Their soil environment with high salinity and strong corrosiveness poses a serious threat to the stability of the ecosystem and artificial engineering facilities. In recent years, certain progress has been made in ecological restoration and maintenance technologies for such sites, such as salt-tolerant vegetation planting and salt leaching projects. However, how to scientifically evaluate the resilience of the ecological system remains a core problem.

[0003] Currently, existing methods for evaluating the resilience of ecological systems in the context of saline soil sites mostly rely on static detection indicators for evaluation. For example, the corrosiveness index is obtained based on the method for measuring the total salt content in soil in the GB 15618-2018 standard, and then the resilience of the ecological system is judged according to the corrosiveness index. However, the corrosiveness of saline soil sites changes dynamically based on salt concentration, and static indicators cannot quantify the impact of salt dynamic fluctuations on vegetation survival rate and microbial activity, etc., thus resulting in relatively low accuracy and reliability in the evaluation of the resilience of the ecological system.

[0004] Therefore, it is particularly important to construct a dynamic and multi-dimensional coupled resilience evaluation method to accurately quantify the adaptability and recovery potential of the ecological maintenance system of corrosive saline soil under salt dynamic fluctuations and compound disturbances. Summary of the Invention

[0005] In view of this, the present invention proposes a method for evaluating the resilience of an ecological maintenance system for corrosive saline soil sites, aiming to solve the problem of relatively low accuracy and reliability in the evaluation of the resilience of the ecological system in the current technology.

[0006] The method for evaluating the resilience of an ecological maintenance system for corrosive saline soil sites proposed by the present invention includes: Dividing the area to be measured into several initial measurement units with equal areas, and obtaining the first detection data set of each of the initial measurement units; the first detection data set includes: soil characteristic data, vegetation growth data, microbial data, and environmental climate data obtained through periodic detection; Performing non-repetitive comparative analysis on the first detection data sets of each of the initial measurement units and the adjacent initial measurement units, classifying each initial measurement unit into several sets, each set being denoted as a measurement unit set, selecting an initial measurement unit as a representative measurement unit in each of the measurement unit sets, and using the first detection data set of the representative measurement unit as the first detection data set of the measurement unit set; Preprocess the first detection data set of the unit set to be measured to obtain a second detection data set, and calculate the toughness evaluation value of each unit set to be measured based on the second detection data set; At the same time stamp, judge the rationality of the toughness evaluation value of each unit set to be measured. If the judgment is reasonable, output the toughness evaluation value and the corresponding range of the unit set to be measured. If the judgment is unreasonable, recalculate the toughness evaluation value according to the second detection data and perform the rationality judgment again; if the judgment is reasonable, output, if the judgment is unreasonable, send an alarm signal.

[0007] Further, the soil characteristic data includes: EC value, pH value, moisture content and CEC value of the soil; The vegetation growth data includes: NDVI index, LAI index and vegetation survival rate; The microbial data includes: Shannon index, halotolerant microbial abundance and soil respiration metabolic rate; The environmental climate data includes: precipitation, temperature, wind speed and relative humidity.

[0008] Further, the non-repetitive comparative analysis includes: When it is judged that the similarity of various first detection data sets of two adjacent initial units to be measured is greater than or equal to a preset similarity threshold, the two initial units to be measured are classified into the same unit set to be measured. Finally, the initial units to be measured that are not classified into the unit set to be measured are independently classified into different unit sets to be measured, and the number of initial units to be measured in all unit sets to be measured is counted and marked.

[0009] Further, when calculating the similarity, obtain the soil characteristic data, vegetation growth data, microbial data and environmental climate data for normalization processing, and then calculate the Euclidean distance of various data in the normalized soil characteristic data, vegetation growth data, microbial data and environmental climate data, and calculate the similarity according to the weighted sum of the Euclidean distance. The similarity satisfies the following relationship: ; Among them, represents the similarity; , , and respectively represent the soil importance weight, vegetation importance, microbial importance weight and environmental climate importance weight, + + + = 1; , , and They represent the weighted normalized distances of soil property data, vegetation growth data, microbial data and environmental climate data respectively.

[0010] Furthermore, when obtaining the soil importance weight, vegetation importance, microbial importance weight and environmental climate importance weight, it includes: The first detection data set is obtained, and the first detection data set is processed by Min-Max normalization. The processed first detection data set is used as an input variable, and the historical ecological resilience assessment value is used as a target variable to input into the XGBoost model for training, and the contribution of each category of the first detection data set is obtained, and the contribution is converted into importance weights. The importance weights satisfy the following relationship: ; Where i=1,2,3,4, to They correspond to the importance weight of soil, the importance weight of vegetation, the importance weight of microorganisms and the importance weight of environmental climate respectively; is the i-th contribution; is the first contribution, indicating the contribution of soil characteristic data; is the second contribution, indicating the contribution of vegetation growth data; is the third contribution, indicating the contribution of microbial data; It is the fourth contribution, indicating the contribution of environmental climate data.

[0011] Furthermore, the rationality judgment includes: Taking any of the unit sets to be tested as the center of the circle, selecting the toughness evaluation values ​​of all the unit sets to be tested within a preset radius according to the spatial position of each unit set to be tested, selecting the toughness evaluation values ​​of the unit sets to be tested passing under any diameter of the circle, fitting the selected toughness evaluation values ​​with a sine function, calculating the fitting degree based on the residual sum of squares, and judging it to be reasonable when any of the following conditions is met, otherwise judging it to be unreasonable; Condition 1: The degree of fit is greater than or equal to 0.9; Condition 2: When the degree of fit is less than 0.9, the numerical difference between any two adjacent toughness evaluation values ​​passing through the unit set to be tested at the current diameter is less than or equal to the maximum limit value.

[0012] Furthermore, the first detection data set of the unit set to be tested is preprocessed, including: setting a data threshold, judging data exceeding the data threshold as an outlier and deleting it; supplementing missing values ​​according to the mean filling method; and performing normalization according to the Min-Max normalization method.

[0013] Further, when the toughness evaluation value of each unit set to be tested is calculated based on the second detection data set, it includes: The toughness evaluation value satisfies the following relationship: ; ; wherein, is the toughness evaluation value of the m-th unit set to be measured; is the adjustment coefficient, and 0 < < 1; is the weighted normalized distance of the m-th unit set to be measured; is the weighted Euclidean distance of the soil property data of the m-th unit set to be measured; is the weighted Euclidean distance of the vegetation growth data of the m-th unit set to be measured, is the weighted Euclidean distance of the microbial data of the m-th unit set to be measured, is the weighted Euclidean distance of the environmental climate data of the m-th unit set to be measured.

[0014] Furthermore, , , and are determined respectively through the following relationships: ; ; ; ; wherein, , , and are the soil property data, vegetation growth data, microbial data and environmental climate data of the j-th unit set to be measured at the same time stamp; , , and are respectively the standard values of the soil property data, vegetation growth data, microbial data and environmental climate data of the j-th unit set to be measured in the state of ecological health; , , and are respectively the characteristic quantities of the soil property data, vegetation growth data, microbial data and environmental climate data.

[0015] Furthermore, when the rationality judgment is performed again and an alarm is given for an unreasonable judgment, it further includes: Calculate the difference between the numerical gap between any two adjacent toughness evaluation values passing through the unit set to be measured at the current diameter and the maximum limit value, and perform hierarchical early warning according to the difference, and the difference is proportional to the level during hierarchical early warning.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: By dividing the initial unit to be measured and merging similar units, redundant data is reduced, the representativeness of the evaluation is improved, and the evaluation result is more in line with the actual ecological situation.

[0017] XGBoost is used to calculate the contribution degrees of soil, vegetation, microorganisms, and environmental climate to resilience, making the weight calculation of various types of data more scientific and reasonable. Data preprocessing methods such as outlier removal, mean filling, and Min-Max normalization are adopted to ensure data quality and improve the stability and reliability of the evaluation.

[0018] After normalization processing, weighted Euclidean distance calculation is performed to avoid evaluation distortion caused by different data dimensions. By calculating the weighted normalized distance, the health degrees of different ecological elements are reasonably measured, and a regulation coefficient is introduced to make the resilience evaluation value more flexible. Using the standard value of the ecological health state as a reference improves the ecological significance of the evaluation result.

[0019] By fitting with a sine function and calculating the sum of squared residuals for the goodness of fit, it is judged whether the resilience evaluation value is reasonable, improving the credibility of the evaluation result. Comparing the resilience evaluation values of adjacent unit sets ensures spatial coherence and reduces the influence of abnormal data on the evaluation result.

[0020] When the evaluation value is unreasonable, the system will recalculate and judge. If it is still unreasonable, an alarm signal will be issued to facilitate the timely discovery of ecological degradation problems. The hierarchical early warning mechanism can be managed hierarchically according to the degree of abnormality of the evaluation value, providing response measures at different levels to facilitate decision-makers to formulate targeted ecological restoration plans.

[0021] Machine learning is used to optimize the calculation of the importance weights of ecological elements to improve adaptability, and it can be applied to ecological monitoring in different regions and different soil environments. It is applicable to long-term ecological monitoring, can conduct dynamic evaluation by combining multi-temporal and multi-spatial data, and helps in the sustainable management of corrosive saline soil sites Description of the Drawings By reading the following detailed description of the preferred embodiments, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present invention. Moreover, throughout the drawings, the same reference numerals are used to represent the same components. In the drawings: Figure 1 It is a flowchart of the resilience evaluation method for the ecological maintenance system of corrosive saline soil sites provided by an embodiment of the present invention. Detailed Embodiments

[0022] Exemplary embodiments disclosed in the present application will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be fully conveyed to those skilled in the art. It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the drawings and in conjunction with the embodiments.

[0023] Referring to Figure 1 as shown, an embodiment of the present invention provides a method for evaluating the resilience of an ecological maintenance system for corrosive saline soil sites, including: S1: Divide the area to be measured into a number of initial measurement units with equal areas, and obtain the first detection data set for each initial measurement unit; the first detection data set includes: soil property data, vegetation growth data, microbial data, and environmental climate data obtained through periodic detection; S2: Compare and analyze the first detection data sets of each initial measurement unit with those of adjacent initial measurement units without repetition, classify each initial measurement unit into a number of sets, each set is denoted as a measurement unit set, select an initial measurement unit as a representative measurement unit in each measurement unit set, and use the first detection data set of the representative measurement unit as the first detection data set of the measurement unit set; S3: Preprocess the first detection data set of the measurement unit set to obtain a second detection data set, and calculate the resilience evaluation value for each measurement unit set based on the second detection data set; S4: At the same time stamp, judge the rationality of the resilience evaluation value of each measurement unit set. If the judgment is reasonable, output the resilience evaluation value and the corresponding range of the measurement unit set. If the judgment is unreasonable, recalculate the resilience evaluation value according to the second detection data and perform the rationality judgment again; if the judgment is reasonable, output, and if the judgment is unreasonable, send an alarm signal.

[0024] It should be noted that the initial measurement unit including the boundary of the area to be measured is an irregular figure with at least one straight side, and the initial measurement units inside each area to be measured (i.e., the measurement units not including the boundary of the area to be measured) are preferably rectangles. The number of initial measurement units is the total area of the area to be measured delimited / the preset area of the measurement unit. Among them, the preset area of the measurement unit can be adjusted according to the actual situation such as the number, cost, or terrain of the available data collection devices. As for the specific division scheme during the specific division, it belongs to the prior art and will not be elaborated here.

[0025] The non-repetitive comparative analysis specifically means that, for example, a to-be-measured area contains an initial to-be-measured unit A and adjacent initial to-be-measured units B, C, D, E, F, G. When conducting a comparative analysis with the initial to-be-measured unit A as the target, it is necessary to separately compare and analyze the first detection data of the initial to-be-measured unit A with the initial to-be-measured units B - G. When taking the initial to-be-measured unit B as the target, it is determined that the initial to-be-measured unit B has already conducted a comparative analysis with the initial to-be-measured unit A, and at this time, the initial to-be-measured unit B no longer conducts a comparative analysis with the initial to-be-measured unit A.

[0026] It can be understood that by dividing the to-be-measured area into multiple initial to-be-measured units and conducting non-repetitive comparative analysis based on the data of adjacent units, the ecological changes in different regions can be accurately captured, ensuring the broad representativeness and regional adaptability of the evaluation results. The preprocessing method is used to clean, fill, and normalize the data, eliminating the influence of data noise and outliers, ensuring a higher quality of the second detection data set, and further enhancing the stability and reliability of the resilience evaluation value. After calculating the resilience evaluation value, a rationality judgment is carried out. By comparing with historical data and spatial data, it is ensured that the evaluation results conform to ecological laws, avoiding misjudgment caused by abnormal data, and enhancing the scientific nature and operability of the evaluation. Through the output and alarm mechanism after the rationality judgment, when the evaluation results are unreasonable, it can recalculate in time and send out an alarm signal, effectively avoiding wrong judgments, providing timely feedback for ecological restoration, and reducing ecological management risks. This method is applicable to large-scale and complex corrosive saline soil sites, can process multi-dimensional and multi-temporal data, has strong adaptability, and provides strong support for the dynamic monitoring and management of soil and ecological environments. Providing real-time and reliable resilience evaluation can provide a scientific basis for ecological maintenance, help managers make timely and accurate repair and protection decisions based on the evaluation results, and improve the efficiency of ecological management.

[0027] In some embodiments of the present application, the soil property data includes: the EC value, pH value, moisture content, and CEC value of the soil; the vegetation growth data includes: the NDVI index, LAI index, and vegetation survival rate; the microbial data includes: the Shannon index, halotolerant microbial abundance, and soil respiration metabolic rate; the environmental climate data includes: precipitation, temperature, wind speed, and relative humidity.

[0028] It should be noted that for the soil property data: the EC value (electrical conductivity), the electrical conductivity of the soil solution is measured by a conductivity meter (EC meter). In this embodiment, the soil sample is mixed with deionized water, and the conductivity of the mixed solution is measured. The EC value is related to the salt concentration of the soil.

[0029] The pH value is measured using a pH meter to measure the acidity and alkalinity of the soil sample. After the soil sample is mixed with deionized water or a buffer solution, the pH value is directly measured using an electrode-type pH meter.

[0030] Moisture content. In this embodiment, the moisture content in the soil is directly measured by a soil moisture meter (such as a time domain reflectometer (TDR) or a resistive moisture sensor). Additionally, the drying method can also be used to estimate the moisture content by measuring the mass change of soil samples.

[0031] CEC value (cation exchange capacity), which is determined by chemical methods. The ammonia exchange method or the sodium chloride method is used. The soil sample reacts with a specific solution, and the number of exchanged cations is proportional to the CEC value of the soil.

[0032] Vegetation growth data: NDVI (Normalized Difference Vegetation Index), which is obtained through remote sensing technology. Reflectance data in the red (Red) and near-infrared (NIR) bands of vegetation are captured using multispectral sensors carried by satellites or drones, and the NDVI value is calculated.

[0033] LAI (Leaf Area Index), which is measured using an instrument method (LAI-2200A instrument) to measure the leaf area of the plant canopy. Additionally, the LAI value can also be inversely calculated from remote sensing data based on the spectral reflectance characteristics of vegetation.

[0034] Vegetation survival rate, which is evaluated by on-site monitoring or using remote sensing images to assess vegetation coverage and health. It can be obtained through sampling and surveys, or the survival status of vegetation can be analyzed in combination with remote sensing data.

[0035] Microbial data: Shannon index, which is detected by gas chromatography for volatile organic compounds (VOCs) in the soil.

[0036] Halotolerant microbial abundance, which is determined by cultivation methods and molecular biology techniques (PCR amplification of specific microbial genes) to measure the number or abundance of halotolerant microorganisms. Selective media can be used to culture microorganisms in a specific salinity environment, and then the abundance can be further calculated.

[0037] Soil respiration metabolic rate, which is measured using a soil respirometer (IRGA infrared gas analyzer) to measure the carbon dioxide release amount from the soil, and then the soil respiration rate is estimated. This measurement usually simulates the metabolic activities of soil microorganisms under natural conditions by controlling environmental conditions such as temperature and humidity.

[0038] Precipitation, temperature, wind speed, and relative humidity in environmental climate data are common data, and the detection and acquisition methods of their data will not be elaborated here.

[0039] In some embodiments of this application, non-repetitive comparative analysis includes: When it is determined that the similarity of various first detection data sets of two adjacent initial units to be measured is greater than or equal to a preset similarity threshold, the two initial units to be measured are classified into the same unit set to be measured. Finally, the initial units to be measured that are not classified into the unit set to be measured are independently classified into different unit sets to be measured, and the number of initial units to be measured in all unit sets to be measured is counted and marked.

[0040] It should be noted that in this embodiment, the similarity threshold is obtained through the following method: Establish classification labels: Collect data of units to be measured that are known to be "similar" or "dissimilar"; Set the known "reasonable" classification situation.

[0041] Train the model (such as logistic regression, XGBoost): Use the similarity S as the input and the correct classification situation as the label for training; Adjust the decision boundary to find the optimal threshold.

[0042] It can be understood that by performing similarity analysis on the first detection data sets of adjacent initial units to be measured and classifying the units with similarity greater than the set threshold into the same unit set to be measured, it helps to accurately identify similar regions and reduce unnecessary classification errors. This method can ensure the effective integration of similar regions, making subsequent analysis more representative and consistent. Classifying similar initial units to be measured into the same set can effectively reduce the waste of computing resources. Compared with processing each unit separately, the classified unit set to be measured can perform toughness evaluation and subsequent analysis more efficiently. Through non-repetitive comparative analysis, the repeated processing of the same feature data is reduced, which helps to improve the efficiency of data processing and avoid redundant computing and storage requirements. This classification method based on similarity analysis can improve the scientific nature of the evaluation results and ensure that the toughness evaluation values of different units to be measured reflect the real ecological differences between regions. If some initial units to be measured cannot be classified into any unit set to be measured, it indicates that these units may have characteristics significantly different from the surrounding environment. This abnormal situation can be quickly discovered and further analyzed or an alarm signal can be issued, thus ensuring the stability and accuracy of the evaluation system.

[0043] In some embodiments of the present application, when calculating the similarity, soil property data, vegetation growth data, microbial data, and environmental climate data are obtained for normalization processing, and then the Euclidean distances of various types of data in the normalized soil property data, vegetation growth data, microbial data, and environmental climate data are calculated, and the similarity is calculated by weighted summation according to the Euclidean distance. The similarity satisfies the following relationship: ; Wherein, represents the similarity; , , and respectively represent the soil importance weight, vegetation importance, microbial importance weight, and environmental climate importance weight, + + + = 1; 、 、 and respectively represent the weighted normalized distances of soil characteristic data, vegetation growth data, microbial data, and environmental climate data.

[0044] It can be understood that this method comprehensively considers multi-dimensional data such as soil characteristic data, vegetation growth data, microbial data, and environmental climate data, and calculates the similarity through weighted summation. This comprehensive method can more accurately evaluate the similarity of the unit to be measured, and improve the scientificity and comprehensiveness of the ecosystem resilience assessment. By introducing different data weights to represent the importance of various types of data, the evaluation process becomes more meticulous and can better reflect the actual impact of each factor on the ecosystem resilience. For example, the weight of soil characteristic data can be set according to its influence degree on plant growth and microbial activities. This method allows adjusting the weights of different data sources according to the actual situation, so as to achieve personalized evaluation. Under different regional or environmental conditions, the relative importance of various types of data can be flexibly adjusted to better meet different ecological maintenance needs. Through the distance-based similarity calculation, different regions or units can be more accurately clustered, which helps to better understand the overall health status of the regional ecology, and thus provides strong support for subsequent environmental management and improvement measures. The data normalization processing method makes various types of data comparable, avoids the deviation between different data magnitudes, and further improves the accuracy of the similarity calculation.

[0045] In some embodiments of the present application, when obtaining the soil importance weight, vegetation importance, microbial importance weight, and environmental climate importance weight, it includes: Obtain the first detection data set, process the first detection data set using Min-Max normalization, use the processed first detection data set as the input variable, and the historical ecological resilience evaluation value as the target variable to input into the XGBoost model for training to obtain the contribution degrees of various types of the first detection data set, and convert the contribution degrees into importance weights. The importance weights satisfy the following relationship: ; where i = 1, 2, 3, 4, to respectively correspond to the soil importance weight, vegetation importance weight, microbial importance weight, and environmental climate importance weight; is the i-th contribution degree; is the first contribution degree, representing the contribution degree of soil characteristic data; is the second contribution degree, representing the contribution degree of vegetation growth data; is the third contribution degree, representing the contribution degree of microorganism data; is the fourth contribution degree, representing the contribution degree of environmental climate data.

[0046] It should be noted that the various types refer to the soil characteristic data, vegetation growth data, microorganism data, and environmental climate data in the first detection dataset, a total of four types of data.

[0047] It can be understood that by using the XGBoost model for training and based on the importance of various types of the dataset, the weights of soil, vegetation, microorganism, and environmental climate data are automatically calculated. This method can optimize the weight allocation according to the actual performance of the data, improving the accuracy of model evaluation. Using Min-Max normalization processing to transform the first detection dataset into a unified range ensures that different types of data can be fairly compared. In this way, the impact of different data units and ranges on the model can be reduced, ensuring that various types of data are processed under the same conditions. According to the feedback of the actual data, the weights of different data types are dynamically adjusted. In this way, factors such as soil, vegetation, microorganism, and environmental climate can automatically obtain corresponding weights according to their importance in ecological assessment, thus providing an assessment result that is more in line with the actual ecological environment. Through this weight adjustment and data processing mechanism, the interaction relationship between various factors in the ecosystem can be more accurately reflected, making the ecological resilience assessment result more scientific and reasonable, and thus providing more effective support for ecological protection and governance. Since the XGBoost model can process various types of data and optimize predictions through training, this method can be flexibly applied in different ecological environments to meet the assessment needs in various scenarios.

[0048] In some embodiments of the present application, the rationality judgment includes: Taking any set of units to be measured as the center, selecting the resilience assessment values of all sets of units to be measured within a preset radius range according to the spatial position of each set of units to be measured, selecting the resilience assessment values of the sets of units to be measured passing through under any diameter of the circle, performing sine function fitting on the selected resilience assessment values, calculating the goodness of fit based on the residual sum of squares, and judging as reasonable when any of the following conditions is met, and judging as unreasonable otherwise; Condition 1: The goodness of fit is greater than or equal to 0.9; Condition 2: When the goodness of fit is less than 0.9, the numerical difference between any two adjacent resilience assessment values of the sets of units to be measured passing through under the current diameter is less than or equal to the maximum limit value.

[0049] It should be noted that in this embodiment, the toughness evaluation values of all the unit sets to be measured within a preset radius range selected according to the spatial positions of each unit set to be measured mean that, taking the area to be measured as a plane, when making a rationality judgment, one unit set to be measured is selected as the center point, and with it as the center, the toughness evaluation values of all the unit sets to be measured within the range of the circle formed under the preset radius are selected. If any unit set to be measured is partially within the range of the circle and partially outside the range of the circle (that is, this unit set to be measured contains the boundary of the circle), this unit set to be measured is also regarded as being within the range of the circle and is selected. The range of the maximum limit value is 0-20%, and the size of the maximum limit value is in a proportional relationship with the area of the average unit set to be measured. When the area of the unit set to be measured is larger, the distance between the center points of two adjacent unit sets to be measured is also larger. If the spatial distance is larger, the fluctuation magnitude and fluctuation probability of the toughness evaluation value will normally increase. Therefore, it is necessary to increase the value of the maximum limit value.

[0050] It can be understood that through the rationality judgment based on the spatial position, this method takes into account the spatial distribution among the unit sets to be measured and can effectively reflect the relationship of ecological toughness between different regions. By selecting the unit sets to be measured within the range of the center point and the radius, it can ensure that when conducting the toughness evaluation, the influence of the spatial position on the ecological toughness is fully considered, thus avoiding the evaluation error caused by spatial discontinuity or dislocation. Through the sine function fitting and the calculation of the goodness of fit, the variation law of the toughness evaluation value in the spatial distribution can be more accurately quantified. When the goodness of fit is greater than 0.9, it means that the distribution of the toughness evaluation values of the unit sets to be measured is relatively stable and meets the expectations, and can provide a more accurate evaluation result. In the case where the goodness of fit is less than 0.9, by setting the maximum limit value (0-20%), the system is allowed to tolerate a certain error when the toughness evaluation value fluctuates greatly. This provides the evaluation system with fault tolerance ability, can cope with the situation where the toughness evaluation values between the unit sets to be measured fluctuate greatly in some special cases, and at the same time ensure the rationality of the evaluation result. The adjustment of the maximum limit value is in a proportional relationship with the area of the unit set to be measured, and can be flexibly adjusted according to the characteristics of different regions or environments, so that the evaluation system can adapt to the areas to be measured of different scales and characteristics. This flexibility ensures the wide applicability of this method in different scenarios. Through means such as the reasonable selection of the spatial position, the sine function fitting, and the adjustment of the maximum limit value, this method can effectively reduce the influence of factors such as data fluctuation, error, or uneven spatial distribution on the evaluation result, and improve the accuracy and reliability of the evaluation.

[0051] In some embodiments of the present application, preprocessing is performed on the first detection data set of the unit set to be measured, including: setting a data threshold, determining the data exceeding the data threshold as outliers and deleting them; supplementing the missing values according to the mean filling method; and performing normalization processing according to the Min-Max normalization method.

[0052] It is understandable that by setting data thresholds and removing outliers beyond the thresholds, the quality of the data can be ensured. Outliers may be caused by external interference or measurement errors. If not handled, they may have an inaccurate impact on subsequent analysis and evaluation. Removing outliers helps improve the accuracy and reliability of the data. Using the mean filling method to supplement missing values is a simple and commonly used data preprocessing method. By filling the missing data points with the mean of the variable, the information loss caused by missing values can be reduced. For cases where complete data cannot be obtained, the mean filling method can ensure the integrity of the data, thus guaranteeing the smooth progress of subsequent analysis. Using the Min-Max normalization method for data normalization can transform the data to a unified scale. Normalization eliminates the differences between different data dimensions, making each feature have the same scale, avoiding the problem that some data features have too much or too little impact on the model, and helping to improve the accuracy of the analysis model. Through these data preprocessing methods, inaccurate, missing, or inconsistent-scale parts can be removed, thereby ensuring that the data is cleaner and more standardized, and improving the accuracy and credibility of the data analysis results. By preprocessing the data, the data can be made more suitable for subsequent statistical analysis or machine learning algorithms. For example, many machine learning algorithms are sensitive to the quality and scale of the data, and the preprocessing steps can effectively improve the performance and convergence speed of the algorithms.

[0053] In some embodiments of the present application, when calculating the toughness evaluation value of each unit set to be measured based on the second detection data set, it includes: The toughness evaluation value satisfies the following relationship: ; ; Wherein, is the toughness evaluation value of the m-th unit set to be measured; is the adjustment coefficient, 0 < < 1; is the weighted normalized distance of the m-th unit set to be measured; is the weighted Euclidean distance of the soil characteristic data of the m-th unit set to be measured; is the weighted Euclidean distance of the vegetation growth data of the m-th unit set to be measured, is the weighted Euclidean distance of the microbial data of the m-th unit set to be measured, is the weighted Euclidean distance of the environmental climate data of the m-th unit set to be measured.

[0054] It is understandable that by combining the Euclidean distances of multiple factors such as soil, vegetation, microorganisms, and climate in a weighted manner, the impacts of various environmental factors on ecosystem resilience can be comprehensively evaluated. This comprehensive evaluation method helps to more accurately understand the health status and development trends of the ecosystem. By calculating the resilience evaluation value of each unit to be measured, a complex ecological maintenance problem can be quantified, enabling managers to make more scientific and effective decisions based on data. By using the weighted Euclidean distance and the exponential decay model to calculate the resilience evaluation value, the differences and their mutual relationships between the units to be measured can be more accurately reflected. In particular, considering the different contributions of different factors (such as soil, vegetation, etc.) to system resilience makes the evaluation results more accurate. The rationality judgment through sine function fitting and goodness-of-fit judgment ensures the rationality of the evaluation results. This can effectively avoid the interference of errors and abnormal data, ensuring that the finally output resilience evaluation value is reliable.

[0055] In some embodiments of the present application, 、 、 and are respectively determined by the following relationships: ; ; ; ; where 、 、 and are the soil property data, vegetation growth data, microorganism data, and environmental climate data of the jth unit set to be measured at the same time stamp; 、 、 and are respectively the standard values of the soil property data, vegetation growth data, microorganism data, and environmental climate data of the jth unit set to be measured in the ecological healthy state; 、 、 and are respectively the characteristic quantities of the soil property data, vegetation growth data, microorganism data, and environmental climate data.

[0056] It should be noted that the characteristic data means that, for example, the soil property data includes: the EC value, pH value, moisture content, and CEC value of the soil, a total of four data. At this time, the characteristic quantity is 4.

[0057] It is understandable that by comparing the soil, vegetation, microorganism, and climate data of different unit sets to be measured with their standard values, the scale differences between different units can be eliminated, enabling each feature to be evaluated under the same standard and ensuring the consistency and fairness of the evaluation process.

[0058] Incorporating multiple ecological factors such as soil properties, vegetation growth, microorganism data, and environmental climate data into the evaluation can more comprehensively reflect the health status of the ecosystem. Data from different dimensions can provide multi-angle analysis support, improving the accuracy of the overall evaluation.

[0059] Through the standardization of the number of features, the resilience evaluation value of the unit set to be measured can be dynamically adjusted and optimized according to the standard value under the ecological health state. This dynamic evaluation mechanism helps to monitor the changes in the ecosystem in real time and make timely adjustments.

[0060] Using the relationship between the standard value and the number of features can effectively perform unified quantification processing on different types of data, simplifying the data processing process and avoiding biases caused by excessive data differences.

[0061] By comparing the data of different units to be measured with the standard value, the scientific nature of the evaluation method can be improved, ensuring that the evaluation results in different regions or environments can be compared under the same standard. This can effectively avoid errors caused by data deviation or environmental differences, making the evaluation results more reliable.

[0062] By using the setting of the number of features, the data fusion process is ensured to be simplified without losing key information. Even in the face of complex and diverse environments, this method can ensure the simplicity and effectiveness of data processing.

[0063] In some embodiments of the present application, when making a rationality judgment again and giving an alarm when the judgment is unreasonable, it further includes: Calculating the difference between the numerical gap between any two adjacent resilience evaluation values passing through the unit set to be measured at the current diameter and the maximum limit value, and performing hierarchical early warning according to the difference. The difference is proportional to the level during hierarchical early warning.

[0064] It is understandable that through this hierarchical setting of the difference and the warning level, a more accurate, flexible, and real-time alarm mechanism can be achieved, thereby improving the system's response speed and processing ability to abnormal situations, which plays an important role in the health management and risk control of the ecological environment.

[0065] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications or equivalent replacements can still be made to the specific implementation manners of the present invention, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A resilience assessment method for an ecological maintenance system for corrosive saline soil sites, characterized in that: include: Divide the area to be tested into a number of initial units to be tested with equal areas, and obtain a first detection data set for each of the initial units to be tested; The first detection data set includes: soil property data, vegetation growth data, microbial data and environmental climate data obtained through periodic detection; Performing a non-repetitive comparative analysis on each of the initial units to be tested and the first detection data sets of the adjacent initial units to be tested, classifying each initial unit to be tested into a plurality of sets, each set being recorded as a unit to be tested set, selecting an initial unit to be tested as a representative unit to be tested in each unit to be tested set, and using the first detection data set of the representative unit to be tested as the first detection data set of the unit to be tested set; Preprocessing the first test data set of the unit set to be tested to obtain a second test data set, and calculating the toughness evaluation value of each unit set to be tested based on the second test data set; At the same timestamp, the rationality of the toughness assessment value of each unit set to be tested is judged. If the judgment is reasonable, the toughness assessment value and the corresponding range of the unit set to be tested are output; if the judgment is unreasonable, the toughness assessment value is recalculated based on the second detection data and the rationality is judged again; if the judgment is reasonable, it is output; if the judgment is unreasonable, an alarm signal is issued.

2. The resilience assessment method for the ecological maintenance system of corrosive saline soil sites according to claim 1 is characterized in that: The soil characteristic data include: EC value, pH value, moisture content and CEC value of the soil; The vegetation growth data include: NDVI index, LAI index and vegetation survival rate; The microbial data include: Shannon index, salt-tolerant microbial abundance and soil respiration metabolic rate; The environmental climate data include: precipitation, temperature, wind speed and relative humidity.

3. The method for evaluating the resilience of the ecological maintenance system for corrosive saline soil sites according to claim 2, characterized in that: The non-repetitive comparative analysis includes: When it is determined that the similarity of various types of first detection data sets of two adjacent initial units to be tested is greater than or equal to a preset similarity threshold, the two initial units to be tested are classified into the same unit set to be tested, and finally the initial units to be tested that are not classified into the unit set to be tested are independently classified into different unit sets to be tested, and the number of initial units to be tested in all unit sets to be tested is counted and marked.

4. The method for evaluating the resilience of an ecological maintenance system for corrosive saline soil sites according to claim 3, characterized in that: When calculating the similarity, soil property data, vegetation growth data, microbial data and environmental climate data are obtained for normalization, and then the Euclidean distance of each type of data in the normalized soil property data, vegetation growth data, microbial data and environmental climate data is calculated, and the similarity is calculated by weighted summation based on the Euclidean distance. The similarity satisfies the following relationship: ； in, Indicates similarity; , , and They represent the importance weight of soil, vegetation, microorganisms and environmental climate, respectively. + + + =1; , , and They represent the weighted normalized distances of soil property data, vegetation growth data, microbial data and environmental climate data respectively.

5. The method for evaluating the resilience of the ecological maintenance system for corrosive saline soil sites according to claim 4, characterized in that: When obtaining the soil importance weight, vegetation importance, microbial importance weight and environmental climate importance weight, it includes: The first detection data set is obtained, and the first detection data set is processed by Min-Max normalization. The processed first detection data set is used as an input variable, and the historical ecological resilience assessment value is used as a target variable to input into the XGBoost model for training, and the contribution of each category of the first detection data set is obtained, and the contribution is converted into importance weights. The importance weights satisfy the following relationship: ； Among them, i=1,2,3,4, to They correspond to the importance weight of soil, the importance weight of vegetation, the importance weight of microorganisms and the importance weight of environmental climate respectively; is the i-th contribution; is the first contribution, indicating the contribution of soil characteristic data; is the second contribution, indicating the contribution of vegetation growth data; is the third contribution, indicating the contribution of microbial data; It is the fourth contribution, indicating the contribution of environmental climate data.

6. The method for evaluating the resilience of the ecological maintenance system for corrosive saline soil sites according to claim 5, characterized in that: The reasonableness judgment includes: Taking any of the unit sets to be tested as the center of the circle, selecting the toughness evaluation values ​​of all the unit sets to be tested within a preset radius according to the spatial position of each unit set to be tested, selecting the toughness evaluation values ​​of the unit sets to be tested passing under any diameter of the circle, fitting the selected toughness evaluation values ​​with a sine function, calculating the fitting degree based on the residual sum of squares, and judging it to be reasonable when any of the following conditions is met, otherwise judging it to be unreasonable; Condition 1: The degree of fit is greater than or equal to 0.9; Condition 2: When the degree of fit is less than 0.9, the numerical difference between any two adjacent toughness evaluation values ​​passing through the unit set to be tested at the current diameter is less than or equal to the maximum limit value.

7. The method for evaluating the resilience of an ecological maintenance system for corrosive saline soil sites according to claim 6, characterized in that: When the first detection data set of the unit set to be tested is preprocessed, it includes: A data threshold is set, and data exceeding the data threshold is judged as an outlier and deleted; missing values ​​are supplemented according to the mean filling method; and normalization is performed according to the Min-Max normalization method.

8. The method for evaluating resilience of an ecological maintenance system for corrosive saline soil sites according to claim 7, characterized in that: When the toughness evaluation value of each unit set to be tested is calculated based on the second detection data set, it includes: The toughness evaluation value satisfies the following relationship: ； ； in, is the toughness evaluation value of the mth unit set to be tested; is the adjustment coefficient, 0< <1; is the weighted normalized distance of the mth unit set to be tested; is the weighted Euclidean distance of the soil property data of the mth unit set to be tested; is the weighted Euclidean distance of the vegetation growth data of the mth unit set to be tested, is the weighted Euclidean distance of the microbial data of the mth unit set to be tested, is the weighted Euclidean distance of the environmental climate data of the mth unit set to be tested.

9. The method for evaluating resilience of an ecological maintenance system for corrosive saline soil sites according to claim 8, characterized in that: , , and They are determined by the following relationships: ； ； ； ； in, , , and is the soil property data, vegetation growth data, microbial data and environmental climate data of the jth unit set to be tested at the same timestamp; , , and are the standard values ​​of soil property data, vegetation growth data, microbial data and environmental climate data of the jth unit set to be tested under ecological health conditions; , , and They are the characteristic quantities of soil property data, vegetation growth data, microbial data and environmental climate data.

10. The method for evaluating resilience of an ecological maintenance system for corrosive saline soil sites according to claim 9, characterized in that: When the rationality judgment is made again and an alarm is issued if the judgment is unreasonable, it also includes: The difference between the numerical difference between any two adjacent toughness evaluation values ​​passing through the unit set to be tested at the current diameter and the maximum limit value is calculated, and a graded warning is performed according to the difference, and the difference is proportional to the level of the graded warning.

Images