Soil quality evaluation method based on deviation fine tuning model

Abstract

The invention provides a soil quality evaluation method based on a deviation fine tuning model, and relates to the technical field of quality evaluation.The method comprises the steps that multi-dimensional indexes of soil are collected and comprehensively evaluated, important indexes are screened through an expert evaluation method, data are refined through methods such as principal component analysis, and the soil quality is scientifically evaluated; accurate and meticulous improvement suggestions are provided; a deviation fine tuning model and an analytic hierarchy process are introduced, so that the accuracy and flexibility of an evaluation result are enhanced, and the actual quality of specific soil types such as black land is reflected; by calculating the soil quality division index S and performing quality grade division and deviation fine adjustment, quantitative expression and high-precision evaluation of results are realized, agricultural managers can quickly understand soil conditions, targeted improvement measures and bases are provided for the agricultural managers, continuous improvement of soil quality is promoted, and agricultural production efficiency is optimized.

Description

[0001] This application is a divisional application of the application with the application number 202410742232.8, the application date of June 11, 2024, and the invention name of "A Soil Quality Evaluation Method and a Soil Information Collection Device" at the time of application. Technical Field

[0002] The present invention relates to the technical field of quality evaluation, and specifically to a soil quality evaluation method based on a deviation fine-tuning model. Background Technique

[0003] With the growing demand for global food security and sustainable agricultural development, the accurate evaluation and management of soil quality have become particularly important. In this context, soil quality evaluation technology has experienced a leapfrog development from initial simple physical and chemical analysis to large-scale soil monitoring using remote sensing technology and geographic information system (GIS); especially the application of the global navigation satellite system (GNSS) in soil quality evaluation has greatly improved the accuracy and efficiency of soil sample positioning; at the same time, by integrating soil physical, chemical, environmental, and biological indicators, researchers and engineers have been able to more comprehensively evaluate soil health status; however, despite continuous technological progress, current soil quality evaluation methods still face many challenges;

[0004] Existing deficiencies: Existing soil quality evaluation technologies often focus on the measurement of specific types of indicators, lacking a comprehensive evaluation framework and being unable to fully reflect the actual situation of the soil; in addition, traditional methods are usually too simplistic in data processing and evaluation models, ignoring the complexity and dynamic variability of soil quality evaluation; for example, few methods can conduct dynamic analysis by combining historical data or adjust the evaluation model according to specific soil types and regional characteristics, resulting in evaluation results often being unable to accurately guide the implementation of soil management and improvement measures.

[0005] The above information disclosed in the background technique section is only used to strengthen the understanding of the background of the present disclosure, and thus it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0006] The purpose of the present invention is to provide a soil quality evaluation method based on a deviation fine-tuning model to solve the problems raised in the above background technique.

[0007] To achieve the above purpose, the present invention provides the following technical solutions:

[0008] A soil quality evaluation method, the specific steps include:

[0009] Step S1: Obtain the GNSS signal of the GNSS receiver to determine the location of the sampling point, and obtain the corresponding historical soil quality evaluation data based on the location of the sampling point;

[0010] Step S2: Collect the soil physical index, soil chemical index, soil environmental index and soil biological index of the sampling point, and each index contains several collection parameters. Screen the several collection parameters included in each index through the expert judgment method to generate the first preset index;

[0011] Step S3: Obtain the first preset index, perform standardization processing on the original data of the first preset index, and perform principal component analysis on the processed data to generate the second preset index;

[0012] Step S4: Obtain the historical second preset index included in the historical soil quality evaluation data. After comparing and analyzing the historical second preset index with the second preset index in Step S3, generate a deviation fine-tuning model; Represent the deviation fine-tuning model as a deviation value D;

[0013] Step S5: After obtaining the second preset index, perform analytic hierarchy process analysis. The analytic hierarchy process includes an objective layer, a criterion layer and an index layer, and measure the index values of the index layer to generate the third preset index;

[0014] Step S6: After obtaining the third preset index, perform analysis and processing to generate a soil quality division index S, and divide the soil quality grade based on the soil quality division index S. The formula of the soil quality division index S is as follows;

[0015]

[0016] Among them, S is the soil quality division index, i represents the i-th index in the third preset index, f i is the score of the soil sample in the i-th index of the third preset index, W i represents the relative weight of the i-th index in the third preset index;

[0017] Step S7: After obtaining the deviation value D corresponding to the deviation fine-tuning model, divide the range of the deviation value D into grade intervals. Each divided grade interval represents the corresponding adjustment amplitude, and fine-tune the output result of the soil quality division index S according to this adjustment amplitude.

[0018] A soil information collection device, which is used to execute the soil quality evaluation method described above. The collection device includes a soil sampling robot, which is composed of a soil sampling, storing and sampling module, a solar power supply module, a walking mechanism based on GNSS positioning, a sensor detection module, and a wireless network transmission module. The soil sampling robot also includes the following black soil quality evaluation system:

[0019] Data acquisition module: It is used to obtain the GNSS signal of the GNSS receiver to determine the position of the sampling point, and obtain the corresponding historical soil quality evaluation data based on the position of the sampling point;

[0020] First index generation module: It is used to collect the soil physical index, soil chemical index, soil environmental index and soil biological index of the sampling point, and each index contains several collection parameters. The expert judgment method is used to screen the several collection parameters included in each index to generate the first preset index;

[0021] Second index generation module: It is used to obtain the first preset index, perform standardization processing on the original data of the first preset index, and perform principal component analysis on the processed data to generate the second preset index;

[0022] Model construction module: It is used to obtain the historical second preset index included in the historical soil quality evaluation data. After comparing and analyzing the historical second preset index with the second preset index in step S3, a deviation fine-tuning model is generated; The deviation fine-tuning model is represented as a deviation value D;

[0023] Third index generation module: After obtaining the second preset index, perform analytic hierarchy process analysis. The analytic hierarchy process includes a target layer, a criterion layer, and an index layer, and measure the index values of the index layer to generate the third preset index;

[0024] Index generation module: It is used to perform analysis and processing after obtaining the third preset index, generate a soil quality division index S, and divide the soil quality grade based on the soil quality division index S. The formula of the soil quality division index S is as follows;

[0025]

[0026] Among them, S is the soil quality division index, i represents the i-th index in the third preset index, f i is the score of the soil sample in the i-th index of the third preset index, W i represents the relative weight of the i-th index in the third preset index;

[0027] Fine-tuning building block: After obtaining the deviation value D corresponding to the deviation fine-tuning model, the value range of the deviation value D is divided into hierarchical intervals, and each hierarchical interval represents the corresponding adjustment amplitude, and the output result of the soil quality division index S is fine-tuned according to this adjustment amplitude.

[0028] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0029] 1. By collecting multi-dimensional indicators of the soil and conducting comprehensive evaluation, including screening important indicators using the expert judgment method and refining data using methods such as principal component analysis, this step makes the soil quality assessment more scientific and comprehensive; 2. By comprehensively considering physical, chemical, environmental, and biological factors, it is possible to analyze the soil quality in all aspects, thereby providing more accurate and detailed soil improvement suggestions. The expert judgment method and principal component analysis are introduced to optimize the index selection and data processing process, aiming to improve the accuracy and reliability of the evaluation; in particular, the introduced deviation fine-tuning model and analytic hierarchy process enhance the precision of the evaluation results and also allow for flexible adjustment according to the specific conditions of the soil, ensuring that the evaluation results can better reflect the actual quality of specific soil types such as black soil; 3. By calculating the soil quality division index S from the results of the comprehensive analysis and based on this index, the quality grade is divided and deviation fine-tuning is performed. This step not only realizes the quantitative expression of the soil quality assessment results, but also ensures the high precision and practicality of the evaluation results through the fine-tuning mechanism; in addition, the division of the quality grade is convenient for agricultural managers to quickly understand the soil conditions and provides a basis for targeted soil improvement measures, ultimately achieving the beneficial effects of promoting the continuous improvement of soil quality and optimizing agricultural production efficiency. Description of the Drawings

[0030] Figure 1 It is a schematic diagram of the overall method flow of the present invention; Figure 2 It is a schematic diagram of the weights of each index;

[0031] Figure 3 It is a schematic diagram of the evaluation layer, criterion layer, and index layer for hierarchical processing of the third preset index. Detailed Embodiments

[0032] In order to make the purpose, technical solutions, and advantages of the present invention clearer, the following further details the present invention in conjunction with specific embodiments.

[0033] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should have the ordinary meanings understood by those with ordinary skills in the field to which the present invention belongs. The "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. Words such as "comprising" or "including" mean that the elements or objects appearing before this word cover the elements or objects listed after this word and their equivalents, without excluding other elements or objects. Words such as "connected" or "linked" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right", etc. are only used to indicate relative position relationships. When the absolute position of the object being described changes, the relative position relationship may also change accordingly.

[0034] Please refer to Figure 1 、 Figure 2 and Figure 3 , the present invention provides a technical solution:

[0035] Example 1:

[0036] A soil quality evaluation method, the specific steps include: Step S1, obtaining the GNSS signal of the GNSS receiver to determine the position of the sampling point, and obtaining the corresponding historical soil quality evaluation data based on the position of the sampling point; the specific steps for determining the position of the sampling point are as follows:

[0037] Configuration and initialization of the GNSS receiver: Before starting, ensure that the GNSS receiver has been correctly configured and initialized, which includes setting the correct time, date and geographical location parameters of the receiver; ensure that the receiver has a clear antenna view so that it can receive signals from multiple satellites; Signal acquisition and positioning: The GNSS receiver captures signals transmitted from global satellite systems such as GPS, GLONASS, and BeiDou; the receiver uses the information contained in these signals to calculate its precise position on the earth. Each satellite sends information about its position and the transmission time. The receiver calculates the distance from each satellite through these data; through the data of at least three satellites, the receiver can accurately calculate the longitude and latitude of its geographical location. If more satellites are added, the positioning accuracy will be further improved; Ensuring positioning accuracy: Using differential GNSS (DGNSS) technology to further improve the position accuracy. This technology corrects errors by comparing the data of two receivers, one is a base station with a fixed position and the other is a mobile station; Recording and marking of sampling points: Once the receiver determines its accurate position, this position is recorded as the sampling point of the soil sample; Record and mark these positions in a geographic information system (GIS) or other mapping tools for subsequent soil collection and analysis;

[0038] Step S2: Collect the soil physical indicators, soil chemical indicators, soil environmental indicators and soil biological indicators at the sampling point. Each indicator contains several collection parameters. Screen the several collection parameters included in each indicator through the expert evaluation method to generate the first preset indicators.

[0039] Step S3: Obtain the first preset indicators, perform standardization processing on the original data of the first preset indicators, and perform principal component analysis on the processed data to generate the second preset indicators.

[0040] Step S4: Obtain the historical second preset indicators included in the historical soil quality evaluation data. After comparing and analyzing the historical second preset indicators with the second preset indicators in Step S3, generate a deviation fine-tuning model. Represent the deviation fine-tuning model as a deviation value D.

[0041] Step S5: After obtaining the second preset indicators, perform the analytic hierarchy process analysis. The analytic hierarchy process includes an objective layer, a criterion layer and an indicator layer, and measure the indicator values of the indicator layer to generate the third preset indicators.

[0042] Step S6: After obtaining the third preset indicators, perform analysis and processing to generate a soil quality division index S, and divide the soil quality grade based on the soil quality division index S. The formula for the soil quality division index S is as follows;

[0043]

[0044] Among them, S is the soil quality division index, i represents the i-th indicator in the third preset indicators, f i is the score of the soil sample for the i-th indicator in the third preset indicators, and W i represents the relative weight of the i-th indicator in the third preset indicators.

[0045] Step S7: After obtaining the deviation value D corresponding to the deviation fine-tuning model, divide the value range of the deviation value D into grade intervals. Each divided grade interval represents the corresponding adjustment range, and fine-tune the output result of the soil quality division index S according to this adjustment range.

[0046] Example 2:

[0047] The historical soil quality evaluation data is expressed as the evaluation data of the previous N times at the sampling point location. The sampling point locations corresponding to the N times of evaluation data are all limited to the same piece of farmland. The same piece of farmland means that the season and month and the farmland label number of the current sampling point and the historical sampling points are the same, so as to ensure that the fluctuation ranges of the soil physical indicators, soil chemical indicators, soil environmental indicators and soil biological indicators are within the limited thresholds, thereby reducing the impact on the evaluation results.

[0048] The method of screening a number of collected parameters included in each index through the expert evaluation method to generate the first preset index specifically includes the following; the steps of the expert evaluation method are successively the formation of a diversity expert group, preliminary index selection, preliminary evaluation, expert scoring, summary and screening, and final screening;

[0049] The said number of parameters includes the following 43 kinds of index parameter contents:

[0050] Organic matter, total nitrogen, available phosphorus, available potassium, water content, cation exchange capacity, humus content, iron and manganese, black soil layer thickness, soil layer thickness, gully erosion area, pH value, vegetation coverage rate, heavy metal concentration, land degradation (salt content), microbial content, and enzyme activity, and also includes the following remaining parameters;

[0051] Soil density: Affects the air permeability and water retention capacity of the soil; capillary porosity: Reflects the water conduction capacity of the soil; non-capillary porosity: Affects the circulation of soil air; total porosity: Affects the soil structure and root growth; soil permeability: Determines the movement rate of water and nutrients in the soil; soil particle composition, such as the proportion of sand, silt, and clay: Affects the texture and other physical properties of the soil; soil structure type: Affects the air permeability and water retention capacity of the soil; soil saturation: Reflects the maximum water retention capacity in the soil; soil temperature: Affects plant growth and microbial activity; soil evaporation: Reflects the loss rate of soil water; soil aggregation: Affects the structural stability and erosion resistance of the soil; soil wind erosion rate: Reflects the soil's resistance to wind erosion; soil organic carbon: An important indicator for measuring soil organic matter and affects soil fertility; carbon-nitrogen ratio: Reflects the characteristics of soil organic matter decomposition and nutrient cycling; exchangeable potassium: Measures the potassium content available for plants to absorb in the soil; exchangeable calcium: Affects the chemical properties and structure of the soil; exchangeable magnesium: An important nutrient element that affects plant growth; soil conductivity: Reflects the salt status of the soil; sulfur content: One of the important nutrient elements in the soil; silicon content: Is important for the growth of certain crops, especially gramineous plants; soil respiration: Reflects the intensity of soil microbial activity; rhizosphere microbial diversity: Reflects the soil biological activity and health status; pathogen quantity: Affects crop growth and yield; biomass carbon: Reflects the conversion ability of soil microorganisms to organic matter;

[0052] Soil animal diversity, such as nematodes and soil insects: Affects soil structure and organic matter decomposition;

[0053] Each of the above-collected parameters is subjected to a normalization dimension operation to make it quantifiable;

[0054] Expert team formation: Form an expert team: This step involves selecting a group of domain experts with extensive soil science knowledge and experience; these experts come from different fields, such as soil physics, chemistry, biology, and environmental science, to ensure the comprehensiveness of the assessment;

[0055] Diversity of experts: Selecting experts with different backgrounds and expertise can ensure that different viewpoints and expertise are taken into account during the assessment process;

[0056] Initial selection of indicators:

[0057] Provide an indicator list: First, a list of 43 soil indicators covering physical, chemical, environmental, and biological aspects needs to be provided;

[0058] Preliminary assessment: The expert team will conduct a preliminary assessment of these indicators, examining the importance, applicability, and relevance of each indicator to the assessment of black soil quality;

[0059] Expert scoring:

[0060] Scoring method: Experts score each indicator using a quantitative method on a scale of 1 - 10;

[0061] Scoring criteria: The scoring focuses on the contribution and sensitivity of the indicators in evaluating black soil quality; this requires experts to comprehensively consider the scientific basis, practical applicability, and evaluation effect of the indicators;

[0062] Summarization and screening: Summarize the scores: Summarize the scores of all experts and calculate the average score of each indicator for easy comparison and analysis;

[0063] Discussion on low - score indicators: Conduct in - depth discussions on the indicators with low scores, and experts can provide detailed explanations and viewpoints from their professional perspectives;

[0064] Final screening:

[0065] Determine the screening results: Based on the experts' scores and discussions, determine which indicators are the most important and applicable for the quality evaluation of the current black soil sampling site location;

[0066] Through this process, the most representative and practical soil indicators can be effectively screened out, thus accurately and comprehensively evaluating the quality of black soil.

[0067] Example 3:

[0068] On the basis of Example 2, it is further illustrated that the acquisition steps of the second preset indicators include:

[0069] Determine the original index of the first preset index. The original index is the type of index parameter selected by the first preset index among the 43 soil indexes described in Example 2. Perform data standardization processing, and set the determined original index so that the mean is 0 and the standard deviation is 1. The specific formula is:

[0070]

[0071] where X is the original data, μ is the mean, σ is the standard deviation, and Z is the standardized data;

[0072] Use the principal component analysis method PCA on the standardized data set to calculate the eigenvalue and eigenvector of each index. The eigenvalue reflects the contribution degree of each principal component in the total data variation. The eigenvector indicates the direction of each principal component;

[0073] Set conditions to select principal components:

[0074] The eigenvalue is greater than the first design value: The eigenvalue represents the amount of data variance explained by the principal component. Only when the eigenvalue of the principal component is greater than a preset threshold, this principal component is considered important. The first design value represented by this threshold needs to be preset according to the specific analysis purpose and data set characteristics;

[0075] The cumulative contribution rate of the main factors is greater than the first design value: The selected principal components should be able to explain most of the total data variation. Usually, a threshold for the cumulative contribution rate is set, such as 90%, which means that the selected principal components should be able to explain at least this proportion of the total variation;

[0076] Quantify with mathematical expressions:

[0077] Suppose there are n original indexes and m samples, and the eigenvalue of each principal component j obtained by principal component analysis is λ j ;

[0078] Eigenvalue condition: λ j > M, where M represents the first design value. When the eigenvalue λ of the principal component j is greater than M, it can be used as the pre-selected index parameter of the second preset index;

[0079] The formula for the cumulative contribution rate condition is:

[0080]

[0081] where k is the number of selected principal components, and Ml is the cumulative contribution rate of the first design value. When the pre-selected index parameter meets the cumulative contribution rate condition formula, the screened index parameter is used as the second preset index;

[0082] By applying the above steps and conditions, the most representative several principal components can be effectively screened out from a large number of indicators, simplifying the subsequent analysis process while retaining the main information in the dataset.

[0083] Example 4:

[0084] Based on Example 3 for further illustration, the generation of the deviation fine-tuning model specifically includes the following content;

[0085] Represent the deviation fine-tuning model as the deviation value D, and the calculation formula for the deviation value D is as follows:

[0086]

[0087] where w j is the importance weight of the second preset indicator based on the PCA result, obtained by normalizing through the eigenvalue λ j The formula is as follows:

[0088]

[0089] m represents the number of samples;

[0090] is the k-th evaluation data of the historical second preset indicator for the j-th principal component;

[0091] y j is the evaluation data of the current second preset indicator for the j-th principal component;

[0092] N is the number of historical evaluations;

[0093] By calculating the difference between the historical data and the current data, and weighting according to the importance of each principal component, a total deviation value D is finally obtained for fine-tuning the current soil quality evaluation; such a design enables the change of each parameter to affect the deviation value D, thus being reflected in the fine-tuning of the soil quality evaluation;

[0094] Range interpretation:

[0095] The range of this formula depends on the change range of the actual soil quality evaluation data, but theoretically the value range of D should be limited. Set the value range of the deviation value D as [-R, R], where R is the maximum positive deviation value obtained based on the historical data change amplitude. Set R1 as the positive value within the value range [-R, R], and R1 is the fine-tuning start threshold close to the positive value; when D ≥ R1, it means that the current soil quality assessment is too low and needs to be adjusted upward by one grade unit; when D ≤ -R1, it means that the soil quality assessment is too high and needs to be adjusted downward by one grade unit; when -R1 < D < R1, it means that the soil quality assessment result meets the conditions and no adjustment is required.

[0096] Example 5:

[0097] On the basis of Example 4, it is further explained. Refer to the appendix Figure 3 ; The acquisition of the third preset index specifically includes the following content; Set the target layer as the black soil quality evaluation,

[0098] The criterion layer is black soil nutrients, black soil environment, black soil layer thickness, and black soil ecological indicators, and corresponds to the soil chemical indicators, soil environmental indicators, soil physical indicators, and soil biological indicators in step S2 in sequence. The criterion layer is a subset of various indicators collected in step S2;

[0099] The index layer is organic matter, total nitrogen, available phosphorus, available potassium, water content, cation exchange capacity, humus content, iron, manganese, black soil layer thickness, soil layer thickness, gully erosion area, pH value, vegetation coverage rate, heavy metal concentration, land salinization, microbial content, and enzyme activity index, which constitute the third preset index; The above data is an example after screening, and does not limit the type of fixed index parameters, and is adjusted and replaced according to actual screening;

[0100] Land salinization represents the salt content of land degradation, that is, land degradation (salt content);

[0101] First, judge the criterion layer, and arrange the priority order of the criterion layer to obtain: black soil nutrients > black soil layer thickness > black soil environment > black soil ecological indicators; Calculate the relative weight of each item, construct a 4×4 judgment matrix T1, and then use the pairwise comparison method to construct and compare the importance degree;

[0102] The definition of the judgment matrix scale is as follows:

[0103] Scale Meaning 1 When comparing two factors, they are equally important 2 Take the middle value between 1 and 3 3 When comparing two factors, the former is more important than the latter 4 Take the middle value between 3 and 5 5 When comparing two factors, the former is much more important than the latter 6 Take the middle value between 5 and 7 7 When comparing two factors, the former is strongly more important than the latter 8 Take the middle value between 7 and 9 9 When comparing two factors, the former is extremely more important than the latter Reciprocal When comparing two factors, the latter is the reciprocal of the former compared to the former compared to the latter

[0104] According to the scale, construct the judgment matrix T1 for the criterion layer, and calculate its relative weight by the square root method. The construction data of the judgment matrix T1 is as follows:

[0105] Nutrient A1 of black soil Soil layer thickness A2 of black soil Environment A3 of black soil Ecological index A4 of black soil Nutrient A1 of black soil 1 2 3 4 Soil layer thickness A2 of black soil 1 / 2 1 2 3 Environment A3 of black soil 1 / 3 1 / 2 1 2 Ecological index A4 of black soil 1 / 4 1 / 3 1 / 2 1

[0106] Therefore, the ratio of black soil nutrient A1 to black soil nutrient A1 is 1, the ratio of black soil layer thickness A2 to black soil nutrient A1 is 1 / 2, the ratio of black soil environment A3 to black soil nutrient A1 is 1 / 3, the ratio of black soil ecological index A4 to black soil nutrient A1 is 1 / 4, the ratio of black soil nutrient A1 to black soil layer thickness A2 is 2, the ratio of black soil layer thickness A2 to black soil layer thickness A2 is 1, the ratio of black soil environment A3 to black soil layer thickness A2 is 1 / 2, the ratio of black soil ecological index A4 to black soil layer thickness A2 is 1 / 3, the ratio of black soil nutrient A1 to black soil environment A3 is 3, the ratio of black soil layer thickness A2 to black soil environment A3 is 2, the ratio of black soil environment A3 to black soil environment A3 is 1, the ratio of black soil ecological index A4 to black soil environment A3 is 1 / 2, the ratio of black soil nutrient A1 to black soil ecological index A4 is 4, the ratio of black soil layer thickness A2 to black soil ecological index A4 is 3, the ratio of black soil environment A3 to black soil ecological index A4 is 2, and the ratio of black soil ecological index A4 to black soil ecological index A4 is 1;

[0107] The judgment matrix T1 is processed as follows;

[0108]

[0109] Multiply the elements of the judgment matrix T1 row by row to obtain a new vector. Take the nth root of each component of the new vector to get the following content, where n is 4:

[0110]

[0111] And perform normalization on M1, M2, M3, and M4 so that the sum of all elements is equal to 1. The following is the weight vector set W1:

[0112]

[0113] In the Analytic Hierarchy Process (AHP), CI represents the Consistency Index; the consistency index is an indicator used to test the degree of consistency of the judgments made by decision-makers when pairwise comparing different factors or alternative solutions; in AHP, the calculation of the consistency index CI involves the use of the Random Index (RI); RI is a set of randomly determined consistency indices based on experience for a given dimension of the judgment matrix, i.e., the order of the judgment matrix; based on the RI value, the Consistency Ratio (CR) can be calculated to evaluate whether the decision-maker's judgment is within an acceptable range. If the CR value is less than a pre-set threshold of 0.1, the consistency of the judgment matrix is considered acceptable; otherwise, the judgment matrix needs to be adjusted to improve consistency;

[0114] Perform a consistency test and define the test formula as follows:

[0115]

[0116] The average random consistency index is as follows:

[0117] n 1 2 3 4 5 6 7 8 9 10 11 12 13 14 RI 0 0 0.52 0.89 1.12 1.26 1.36 1.41 1.46 1.49 1.52 1.54 1.56 1.58

[0118] Calculate the consistency ratio as follows:

[0119]

[0120] where λ max is the largest eigenvalue of the judgment matrix T1; W i is the sum of the weights of the elements in the i-th row of the judgment matrix, and W 1i represents the weight corresponding to the i-th criterion. The prefix 1 in 1i is an index number, and the following explanations are the same and will not be repeated; the larger the value of CI, the greater the degree to which the judgment matrix deviates from perfect consistency, and the smaller the value of CI, the better the consistency of the judgment matrix;

[0121] Look up the corresponding average random consistency index RI in a table. RI is the average of the consistency indices of random judgment matrices of the same order. Introducing it can, to a certain extent, overcome the drawback that the consistency judgment index increases significantly as n increases;

[0122] When CR < 0.1, the judgment matrix is considered to have satisfactory consistency. When CR ≥ 0.1, the judgment matrix is considered inconsistent, and the matrix needs to be adjusted to satisfy CR < 0.1;

[0123] Therefore, in the criterion layer, the weight of black soil nutrients is 0.47, the weight of black soil layer thickness is 0.27, the weight of black soil environment is 0.16, and the weight of black soil ecological indicators is 0.10;

[0124] In calculating the weights of the index layer, first calculate the weight of the black soil layer thickness and the soil layer thickness in the black soil layer thickness. The importance of the black soil layer thickness > the soil layer thickness;

[0125] The construction data of the judgment matrix T2 are as follows;

[0126]

[0127] The ratio of the black soil layer thickness B1 to the black soil layer thickness B1 is 1, the ratio of the soil layer thickness B2 to the black soil layer thickness B1 is 1 / 5, the ratio of the black soil layer thickness B1 to the soil layer thickness B2 is 5, and the ratio of the soil layer thickness B2 to the soil layer thickness B2 is 1. And establish the following 2×2 judgment matrix T2;

[0128]

[0129] Multiply the elements of the judgment matrix T2 row by row to get a new vector, and take the nth root of each component of the new vector;

[0130]

[0131] And perform normalization processing on M5 and M6 so that the sum of all elements is equal to 1. The following is the weight vector set

[0132] W2;

[0133]

[0134] Perform consistency check:

[0135]

[0136] The average random consistency index is as follows;

[0137] n 1 2 3 4 5 6 7 8 9 10 11 12 13 14 RI 0 0 0.52 0.89 1.12 1.26 1.36 1.41 1.46 1.49 1.52 1.54 1.56 1.58

[0138] Because when n = 2, RI = 0, so no consistency check is required. In the black soil layer thickness, the weight of the black soil layer thickness index is 0.83, and the weight of the soil layer thickness is 0.17. Similarly, in the black soil ecological indicators, because the scale ratio of the microbial content to the enzyme activity is equal to the scale ratio of the black soil layer thickness to the soil layer thickness in the black soil layer thickness, and the importance of the microbial content > the importance of the enzyme activity, so in the black soil ecological indicators, the weight of the microbial content index is 0.83, and the weight of the enzyme activity is 0.17;

[0139] Calculate the area of gully erosion, pH, vegetation coverage rate, heavy metal concentration, and the weight of land degradation (salt content) in the black soil environment, and the importance of vegetation coverage rate > the importance of heavy metal concentration > the importance of pH > the importance of land salinization > the importance of gully erosion area;

[0140] The construction data of the judgment matrix T3 is as follows:

[0141]

[0142] The ratio of vegetation coverage rate C1 to vegetation coverage rate C1 is 1, the ratio of heavy metal concentration C2 to vegetation coverage rate C1 is 1 / 2, the ratio of pH C3 to vegetation coverage rate C1 is 1 / 3, the ratio of land degradation C4 to vegetation coverage rate C1 is 1 / 4, the ratio of gully erosion area C5 to vegetation coverage rate is 1 / 5, the ratio of vegetation coverage rate C1 to heavy metal concentration C2 is 2, the ratio of heavy metal concentration C2 to heavy metal concentration C2 is 1, the ratio of pH C3 to heavy metal concentration C2 is 1 / 2, the ratio of land degradation C4 to heavy metal concentration C2 is 1 / 3, the ratio of gully erosion area C5 to heavy metal concentration C2 is 1 / 4, the ratio of vegetation coverage rate C1 to pH C3 is 3, the ratio of heavy metal concentration C2 to pH C3 is 2, the ratio of pH C3 to pH C3 is 1, the ratio of land degradation to pH C3 is 1 / 2, the ratio of gully erosion area to pH C3 is 1 / 3, the ratio of vegetation coverage rate C1 to land degradation C4 is 4, the ratio of heavy metal concentration C2 to land degradation C4 is 3, the ratio of pH C3 to land degradation C4 is 2, the ratio of land degradation C4 to land degradation C4 is 1, the ratio of gully erosion area C5 to land degradation C4 is 1 / 2, the ratio of vegetation coverage rate C1 to gully erosion area C5 is 5, the ratio of heavy metal concentration C2 to gully erosion area C5 is 4, the ratio of pH C3 to gully erosion area C5 is 3, the ratio of land degradation to gully erosion area C5 is 2,

[0143] The ratio of gully erosion area C5 to gully erosion area C5 is 1; and establish the following judgment matrix T3:

[0144]

[0145] Multiply the elements of the judgment matrix T3 row by row to obtain a new vector, and take the nth root of each component of the new vector;

[0146]

[0147] And perform normalization processing on E1, E2, E3, E4, and E5 so that the sum of all elements is equal to 1. The following is the weight vector set W3;

[0148]

[0149] Perform consistency test:

[0150]

[0151]

[0152] Average random consistency index

[0153] n 1 2 3 4 5 6 7 8 9 10 11 12 13 14 RI 0 0 0.52 0.89 1.12 1.26 1.36 1.41 1.46 1.49 1.52 1.54 1.56 1.58

[0154] Calculate the consistency ratio as follows:

[0155]

[0156] In the black soil environment, the weight of the index vegetation coverage rate is 0.42, the weight of the heavy metal concentration is 0.26, the weight of pH is 0.16, the weight of land degradation (salt content) is 0.1, and the weight of the gully erosion area is 0.06;

[0157] Calculate the weights of the indexes organic matter, total nitrogen, available phosphorus, available potassium, water content, cation exchange capacity, humus layer, iron, and manganese in the nutrients of black soil;

[0158] The construction data of the judgment matrix T4 are as follows:

[0159]

[0160] And establish the following judgment matrix T4:

[0161]

[0162] Multiply the elements of the judgment matrix T4 row by row to obtain a new vector, and take the nth root of each component of the new vector,

[0163]

[0164] And perform normalization processing on F1, F2, F3, F4, F5, F6, F7, F8, F9 so that the sum of all elements is equal to 1. The following is the weight vector set W4;

[0165]

[0166]

[0167] Perform consistency test:

[0168]

[0169] Average random consistency index:

[0170] n 1 2 3 4 5 6 7 8 9 10 11 12 13 14 RI 0 0 0.52 0.89 1.12 1.26 1.36 1.41 1.46 1.49 1.52 1.54 1.56 1.58

[0171] Calculate the consistency ratio:

[0172]

[0173] In the nutrients of black soil, the weight of organic matter in the indicators is 0.254, the weight of total nitrogen is 0.171, the weight of available phosphorus is 0.18, the weight of available potassium is 0.148, the weight of water content is 0.1, the weight of cation exchange capacity is 0.062, the weight of humus layer is 0.019, the weight of iron is 0.03, and the weight of manganese is 0.036;

[0174] Score the index values of the third preset index: According to the index values of the third preset index, the corresponding membership degree of the index, and membership degree × 100 is the score of this index. The following is the corresponding relationship between the index values and membership degrees of the index;

[0175] Membership degree of pH value:

[0176]

[0177] For organic matter, total nitrogen, available phosphorus, available potassium, cation exchange capacity, humus layer in the nutrients of black soil, black soil layer thickness in the black soil layer thickness, vegetation coverage rate in the black soil environment, and for index microbial content and enzyme activity in the black soil ecological index, substitute them into the f(x) formula:

[0178]

[0179] Among them, X1 and X2 are the preset critical values for each index. The characteristics of the f(x) membership function applicable are: the score will increase with the increase of the index content. For iron and manganese in the nutrients of black soil, heavy metal concentration in the black soil environment, land degradation (salt content), and gully erosion area;

[0180]

[0181] X1 and X2 are the preset critical values for each index. The characteristics of the f2(x) membership function applicable are: for some indexes, they are beneficial within a certain range, and exceeding a certain index will cause harm;

[0182] X1, X2 critical value table:

[0183]

[0184]

[0185] The process of black soil quality scoring: First, for the index layer, sampling is carried out first, and then the corresponding membership degree is obtained according to the index value of the index. The membership degree × 100 is the evaluation score of the index. According to each criterion layer corresponding to the index, the scores are summed to obtain the comprehensive score of each item in the criterion layer. Then, according to the weight of each item in the criterion layer, multiply by the comprehensive score of each item to finally obtain the comprehensive score of black soil quality;

[0186] The criterion layer indicators are divided into grades;

[0187]

[0188] The black soil score grade is divided as follows:

[0189] Grade Score First level 100-85 Second level 85-70 Third level 70-60 Fourth level ＜60

[0190] Example 6:

[0191] On the basis of Example 5, it is further explained that after obtaining the deviation value D corresponding to the deviation fine-tuning model, the value range of the deviation value D is divided into grade intervals, and each divided grade interval represents the corresponding adjustment amplitude, and the output result of the soil quality division index S is fine-tuned according to this adjustment amplitude, which specifically includes the following content;

[0192] The value range [-R, R] of the deviation value D is divided into five intervals, and each interval represents a specific adjustment amplitude;

[0193] Furthermore, R2 is set as the fine-tuning sensitivity parameter to adjust the fineness of the fine-tuning amplitude;

[0194] Based on the most recent time period, assuming this time period is one year, R1 is dynamically calculated according to the volatility of the soil quality evaluation data within one year to ensure that the fine-tuning model can adapt to the data variability in different time periods; if the data in the most recent year changes greatly, R1 is correspondingly reduced to improve the sensitivity of fine-tuning;

[0195] For the fine-tuning sensitivity R2, an evaluation based on the soil quality importance index is introduced. For soils with a high contribution degree to ecosystem services, the value of R2 is increased, so as to increase its change amplitude during fine-tuning to reflect the importance of soil quality to the ecosystem;

[0196] The fine-tuned soil quality division index is defined as S', and the fine-tuned soil quality division index S' is expressed by the following expression:

[0197] S′ = S + ΔS

[0198] Among them, ΔS is the adjustment value determined according to the interval where the value of the deviation fine-tuning model D is located;

[0199] When D ≥ R1, ΔS = min(D × P1, T max - S);

[0200] When D ≤ R1, ΔS = max(D × P2, T min - S);

[0201] Among them, P1 and P2 are adjustment factors, selected according to the current grade of the soil and the magnitude of the deviation; T max and T min are the upper and lower limit scores of the current grade; such a design ensures that the fine-tuned score remains within a reasonable range and prevents over-adjustment;

[0202] When D ≥ R1, a non-linear adjustment strategy is adopted to increase the percentage of S, which is ΔS, not only based on the interval where D is located, but also considering the value of R2; the percentage increase in S depends not only on the specific value of D, but also on R2 to achieve a more refined adjustment;

[0203] When D ≤ -R1, the same non-linear adjustment strategy is adopted to reduce the percentage of S, considering the interval where D is located and the value of R2;

[0204] When R1 > D > -R1, considering that the small fluctuations in the data do not represent substantial changes in soil quality, so S is not adjusted within this interval;

[0205] Adjustment range for first-level soil quality: When (85 ≤ S < 100),

[0206] If D ≥ R1, that is, the score is underestimated, ΔS = min(D × P1, T max - S);

[0207] If D ≤ -R1, that is, the score is overestimated, ΔS = max(D × P2, T min - S);

[0208] Specifically, when D ≥ R1, that is, when the score is underestimated, the calculation method of the adjustment value ΔS is to ensure that the adjustment made can reflect the actual situation of the score and will not exceed the upper limit of this grade; the calculation expression here ΔS = min(D × P1, T max - S) means that the adjustment range ΔS is determined by the smaller value of two parts:

[0209] (D × P1) represents the product based on the deviation D and a preset proportional factor P1; P1 represents the adjustment ratio corresponding to each unit of deviation D, that is, for each unit of the underestimated score amplitude, how many scores should be increased to adjust S; this product directly reflects the positive adjustment amount required due to the underestimated score;

[0210] (T max-S) represents the gap between the current score S and the upper limit T of its level; this gap is the maximum amplitude by which the score S can increase without any adjustment to ensure that the score remains within the range of the current level after adjustment; max Between; this gap is the maximum amplitude by which the score S can increase without any adjustment to ensure that the score remains within the range of the current level after adjustment;

[0211] By taking the smaller of these two values as ΔS, two important principles are ensured:

[0212] The precision of the adjustment means that the adjustment amount calculated by (D × P1) ensures that the amplitude of the score adjustment matches the actual situation where the score is underestimated, thus ensuring the rationality of the adjustment and the accuracy of the goal;

[0213] The limit of the adjustment means that by comparing with (T max -S) and taking the smaller value, the adjustment amplitude is limited so that the score will not exceed the upper limit of its current level, avoiding over - adjustment;

[0214] The specific ΔS is determined through the following steps:

[0215] First, based on the degree D by which the actual score is underestimated and the proportional factor P1, a preliminary adjustment amplitude is calculated;

[0216] Secondly, the gap between the current score and the upper limit of the level is determined, which represents the maximum adjustable space not exceeding the upper limit of the level;

[0217] Compare and select the smaller value: Finally, compare the above two values and take the smaller one as the final adjustment amplitude ΔS;

[0218] In this way, the determination of ΔS takes into account both the need for score adjustment and avoids adjusting the score to an unreasonable range, ensuring the rationality and feasibility of the adjustment;

[0219] For the sake of easy expression, assume R = 10. The five intervals are [-10, -6], (-6, -2], (-2, 2), [2, 6] and (6, 10]. The adjustment amplitudes of the soil quality division index S corresponding to each interval are as follows:

[0220] When the value range of the interval D is [-10, -6], S decreases by (1% + 0.1% × R2);

[0221] When the value range of the interval D is (-6, -2], S decreases by (2% + 0.2% × R2);

[0222] In these expressions, [1% + 0.1% × R2] means that for each unit of deviation D, S will decrease by 1% plus one - tenth of R2 as a percentage, and [2% + 0.2% × R2] similarly means that for each unit of deviation D, S will decrease by 2% plus one - fifth of R2 as a percentage;

[0223] When the value range of interval D is (-2, 2), S remains unchanged;

[0224] When the value range of interval D is [2, 6], S increases by (3% + 0.3% × R2);

[0225] When the value range of interval D is (6, 10], S increases by (5% + 0.5% × R2);

[0226] Secondary soil quality adjustment range:

[0227] The scoring range of secondary soil is (70 ≤ S < 85):

[0228] If D ≥ R1, that is, the score is underestimated, ΔS = min(D × P1, T max - S); where T max = 85;

[0229] If D ≤ -R1, that is, the score is overestimated, ΔS = max(D × P2, T min - S); where T min = 70;

[0230] Tertiary soil quality adjustment range:

[0231] The scoring range of tertiary soil is (60 ≤ S < 70):

[0232] If D ≥ R1, that is, the score is underestimated, ΔS = min(D × P1, T max - S); where T max = 70;

[0233] If D ≤ -R1, that is, the score is overestimated, ΔS = max(D × P2, T min - S); where T min = 60;

[0234] Quaternary soil quality adjustment range:

[0235] The scoring range of quaternary soil is (S < 60):

[0236] If D ≥ R1, that is, the score is underestimated, ΔS = min(D × P1, T max - S); where T max = 55;

[0237] If D ≤ -R1, that is, the score is overestimated, ΔS = max(D × P2, T min - S); where T min = 40;

[0238] In the above adjustment strategy, P1 and P2 are adjustment ratio factors for different soil grades and can be set according to actual situations;

[0239] The determination of P1 and P2 is specifically referred to the following methods:

[0240] First, collect historical soil quality evaluation data and related experimental results;

[0241] Analyze these data, especially paying attention to the deviation D between the soil quality score S and the actual soil quality;

[0242] Based on the analysis results, determine the adjustment ratio factors, P1 and P2, which are determined according to the magnitude of the deviation value and the current soil grade to achieve reasonable adjustment of the score;

[0243] The adjustment ratio factor P1 is used to increase S when the score is underestimated, while P2 is used to decrease S when the score is overestimated;

[0244] Verify the determined P1 and P2, and confirm the rationality and effectiveness of these factors through simulation experiments of historical data or adjustments of expert opinions.

[0245] Example Seven:

[0246] On the basis of Example Six, further explain that before designing this example, first determine the purpose of the experiment: verify that the refinement of the fine-tuning strategy can effectively adjust the soil quality score to make it more accurately reflect the actual situation of the soil. The experimental design focuses on the refinement of the fine-tuning strategy, and proves the effectiveness and innovation of the strategy by comparing the score changes of different soil quality grades before and after applying the new strategy;

[0247] The experiment is divided into three stages: preparation, execution and analysis. In the preparation stage, four representative soil samples are selected, belonging to the first-level, second-level, third-level and fourth-level soil quality respectively; the preliminary score of each sample is obtained based on the traditional evaluation method, and the fine-tuning strategy is not applied at this time; meanwhile, set R = 10, P1 = 0.8, P2 = 0.6, which are adjustment ratio factors respectively, and these parameters will be used in the subsequent adjustment process;

[0248] In the execution stage, record the D value of each sample in detail, that is, the difference between the pre-score and the actual score, and calculate the adjustment range ΔS according to the positive or negative and magnitude of the D value according to the rules of the refinement of the fine-tuning strategy; this process requires precise adjustment of the score of each sample to ensure that each adjustment is within the acceptable error range and finally make the score closer to the actual soil quality;

[0249] To prove the effectiveness of the fine-tuning strategy, the changes in soil quality scores before and after implementing the fine-tuning strategy were compared. The preliminary scores, adjusted scores, D values, and adjustment amplitudes ΔS of each sample were carefully recorded and used for subsequent data analysis;

[0250] Excel spreadsheet data:

[0251] The following are the experimental data, presented in the format of an Excel spreadsheet:

[0252]

[0253] Analysis of the spreadsheet data:

[0254] From the above table, it can be observed that after applying the refined fine-tuning strategy, the scores of each soil sample are closer to their actual quality status; for the first-class soil samples, the score increased by 4 points after adjustment, indicating that the strategy can effectively identify and correct underestimated soil quality scores; similarly, the scores of the second-class and fourth-class soil samples decreased after adjustment, accurately reflecting the overestimated scores; while the score adjustment of the third-class soil samples further verified the applicability of this strategy for medium-quality soil, with the adjustment amplitude meeting expectations and making the scores more precise;

[0255] Through in-depth analysis of the experimental data, a detailed description is provided for the preliminary scores, adjustment amplitudes, and adjusted scores of each soil sample;

[0256] First, consider the first-class soil samples. Their preliminary score is 87 points, within the score range of 85 ≤ S < 100 for first-class soil quality; the D value of this sample is 5, indicating that the preliminary score is lower than the actual situation; after applying the fine-tuning strategy, the adjustment amplitude ΔS is 4 points, which is calculated based on the D value and the adjustment ratio factor P1 = 0.8; the adjusted score is 91 points, closer to the actual high-quality state of this soil sample; this adjustment result proves the effectiveness of the fine-tuning strategy for underestimated high-quality soil samples, capable of appropriately increasing the score to reflect their true soil quality;

[0257] For the second-class soil samples, the preliminary score is 73 points, within the score range of 70 ≤ S < 85 for second-class soil quality; its D value is -4, indicating that the preliminary score is higher than the actual situation; according to the fine-tuning strategy, the adjustment amplitude ΔS is -2.4 points. Considering P2 = 0.6, the adjusted score is reduced to 70.6 points, more accurately representing the actual soil quality of this sample; this result shows the corrective effect of the fine-tuning strategy on overestimated scores, precisely matching the soil quality grade by reducing the score;

[0258] The situation of the third-level soil sample is also very illustrative; its preliminary score is 65 points, and according to its D value of 3, it indicates that the score is slightly lower than the actual; after adjustment with ΔS = 2.4, the final score is 67.4 points, which more accurately reflects the medium quality of this soil; this process demonstrates the applicability of the fine-tuning strategy in the scoring of medium-quality soils, by appropriately increasing the score to correct the slightly underestimated situation;

[0259] Finally, the preliminary score of the fourth-level soil sample is 50 points, and the D value is -5, suggesting that the score is overestimated; after adjustment with ΔS = -3, the score drops to 47 points, which more appropriately represents the actual situation of this soil sample, that is, a lower soil quality; this result further confirms the effectiveness of the fine-tuning strategy in the scoring of soil quality at all levels, especially in downward-adjusting the score to reflect soil samples with lower actual quality;

[0260] In summary, through the refinement of the introduced fine-tuning strategy, not only can the soil quality score be effectively adjusted to more accurately reflect the actual situation of the soil, but this process also shows good adaptability and precision in different soil quality grades; each adjustment is based on the actual scoring deviation, ensuring that the scoring adjustment is both reasonable and targeted, and ultimately achieving a closer consistency between the score and the actual quality of the soil.

[0261] Example Eight:

[0262] A soil information collection device, which is used to execute the soil quality evaluation method described above. The collection device includes a soil sampling robot, which is composed of a soil sampling, storing and sampling module, a solar power supply module, a walking mechanism based on GNSS positioning, a sensor detection module, and a wireless network transmission module, and includes a control module for control. Among them, the sampling module breaks the ground of the black soil in different seasons, the sensor detection device acquires data on the black soil after breaking the ground, the GNSS and path planning algorithms are used to divide the sampling plot into grids and plan the paths between sampling points, and machine vision and wireless transmission network devices are used to divide the evaluation units of the black soil and upload the soil detection results of the corresponding sampling points to the cloud;

[0263] In the control process of the control module for the traveling mechanism, the control device receives GNSS signals through a GNSS receiver to determine the position of the device, and performs grid division according to the remote sensing map of the field where the device is located to determine the positions of the sampling points; and plans the traveling route according to the algorithm and controls the traveling modes such as forward, turning, and backward of the traveling device based on the position of the device; the control module controls the sampling device according to whether it judges that the device has reached the sampling point.

[0264] The soil sampling robot also includes the following black soil quality evaluation system:

[0265] Data acquisition module: used to obtain the GNSS signals of the GNSS receiver to determine the location of the sampling point, and obtain the corresponding historical soil quality evaluation data based on the location of the sampling point;

[0266] First index generation module: used to collect the soil physical index, soil chemical index, soil environmental index and soil biological index of the sampling point, and each index contains several collection parameters. The expert judgment method is used to screen the several collection parameters included in each index to generate the first preset index;

[0267] Second index generation module: used to obtain the first preset index, perform standardization processing on the original data of the first preset index, and perform principal component analysis on the processed data to generate the second preset index;

[0268] Model construction module: used to obtain the historical second preset index included in the historical soil quality evaluation data, generate a deviation fine-tuning model after comparing and analyzing the historical second preset index with the second preset index in step S3; represent the deviation fine-tuning model as a deviation value D;

[0269] Third index generation module: perform analytic hierarchy process analysis after obtaining the second preset index. The analytic hierarchy process includes a target layer, a criterion layer and an index layer, and measure the index values of the index layer to generate the third preset index;

[0270] Index generation module: used to perform analysis and processing after obtaining the third preset index, generate a soil quality division index S, and divide the soil quality grade based on the soil quality division index S. The formula for the soil quality division index S is as follows;

[0271]

[0272] Among them, S is the soil quality division index, i represents the i-th index in the third preset index, f i is the score of the soil sample for the i-th index in the third preset index, and W i represents the relative weight of the i-th index in the third preset index;

[0273] Fine-tuning construction module: used to divide the range of the deviation value D corresponding to the deviation fine-tuning model into grade intervals after obtaining the deviation value D, and each divided grade interval represents the corresponding adjustment amplitude, and fine-tune the output result of the soil quality division index S according to the adjustment amplitude.

[0274] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to get a formula closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0275] The specific values of δ, ε, ∈, σ, etc. in the formula are generally determined by those skilled in the art according to the actual situation. The essence of this formula in this application is weighted summation for comprehensive analysis. Those skilled in the art collect multiple groups of sample data and set corresponding preset proportional coefficients for each group of sample data; substitute the set preset proportional coefficients and the collected sample data into the formula. Any four formulas form a system of linear equations with four variables. Screen and take the average of the calculated coefficients to obtain the values of δ, ε, ∈, σ, etc.

[0276] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those skilled in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in this article can be implemented through email, or a combination of computer software and electronic hardware. Whether these functions are executed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0277] The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units. They can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0278] The above is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in this application, and all should be covered by the protection scope of this application.

Claims

1. A soil quality evaluation method based on a deviation fine-tuning model, characterized in that: The specific steps include: Step S1, obtaining a GNSS signal from a GNSS receiver to determine the location of a sampling point, and obtaining corresponding historical soil quality evaluation data based on the location of the sampling point; The specific steps for determining the location of the sampling point are as follows: (1) Configuration and initialization of the GNSS receiver: Before starting, ensure that the GNSS receiver has been correctly configured and initialized; (2) Signal acquisition and positioning: The GNSS receiver captures the signal transmitted from the global satellite system; (3) Ensuring positioning accuracy: Using differential GNSS technology to further improve the accuracy of the position, the error is corrected by comparing the data of two receivers, one is a fixed-position base station and the other is a mobile station; (4) Recording and marking of sampling points: Once the receiver determines its exact location, this location is recorded as the sampling point of the soil sample; Step S2, collecting soil physical indicators, soil chemical indicators, soil environmental indicators and soil biological indicators of the sampling point, and each indicator includes a number of collection parameters, and screening the several collection parameters included in each indicator through an expert judgment method to generate a first preset indicator; Step S3, obtaining the first preset index, standardizing the original data of the first preset index, and performing principal component analysis on the processed data to generate the second preset index; Step S4, obtaining the historical second preset index contained in the historical soil quality evaluation data, comparing and analyzing the historical second preset index with the second preset index of step S3, and generating a deviation fine-tuning model; The deviation fine-tuning model is represented as the deviation value D, and the deviation value D is calculated as follows: Among them, w j It is the importance weight of the second preset index based on the PCA result of the principal component analysis method, through the eigenvalue λ of each principal component j j Normalized, the formula is as follows: m represents the sample size; It is the kth evaluation data of the second historical preset indicator for the jth principal component; y j is the evaluation data of the current second preset indicator for the jth principal component; N is the number of historical evaluations; By calculating the difference between historical data and current data and weighting them according to the importance of each principal component, a total deviation value D is finally obtained to fine-tune the current soil quality evaluation; the range of the deviation value D is set to [-R, R], where R is the maximum positive deviation value based on the change amplitude of historical data, R1 is set to a positive value within the range [-R, R], and R1 is the fine-tuning start threshold close to a positive value; When D ≥ R1, it means that the current soil quality assessment is too low and needs to be adjusted upward; When D≤-R1, it means that the soil quality assessment is too high and needs to be adjusted downward; When R1>D>-R1, it means that the soil quality assessment results meet the conditions and no adjustment is required; Step S5, after obtaining the second preset index, perform a hierarchical analysis method, the hierarchical analysis method includes a target layer, a criterion layer and an index layer, and measure the index value of the index layer to generate a third preset index; Step S6, after obtaining the third preset index, analyze and process it to generate a soil quality classification index S, and classify the soil quality grades based on the soil quality classification index S, and the formula of the soil quality classification index S is as follows; Among them, S is the soil quality classification index, i represents the i-th index in the third preset index, and f i Score the soil sample for the i-th indicator in the third preset indicator, W i It represents the relative weight of the ith indicator in the third preset indicator; Step S7, after obtaining the deviation value D corresponding to the deviation fine-tuning model, the value range of the deviation value D is divided into level intervals, each level interval represents a corresponding adjustment range, and the output result of the soil quality classification index S is fine-tuned according to the adjustment range; Divide the deviation value D value range [-R, R] into five intervals, each interval represents a specific adjustment range; Set R2 as the fine-tuning sensitivity parameter to adjust the degree of fine-tuning; Based on the most recent time period, assuming that the time period is one year, R1 is dynamically calculated according to the volatility of soil quality evaluation data within one year to ensure that the fine-tuning model can adapt to the variability of data in different time periods; if the data in the most recent year changes greatly, R1 is reduced accordingly to improve the sensitivity of fine-tuning; For the fine-tuning sensitivity R2, an evaluation based on soil quality importance indicators is introduced. For soils with high contribution to ecosystem services, the R2 value is increased, thereby increasing its variation range during fine-tuning to reflect the importance of soil quality to the ecosystem; The fine-tuned soil quality partition index is defined as S', and the fine-tuned soil quality partition index S' is expressed by the following expression: S′=S+ΔS Among them, ΔS is the adjustment value determined according to the interval of the value of the deviation fine-tuning model D; When D≥R1, ΔS=min(D×P1,T max -S); When D≤R1, ΔS=max(D×P2,T min -S); Among them, P1 and P2 are adjustment factors, which are selected according to the current grade and deviation of the soil; T max and T min are the upper and lower scores of the current level.

2. The soil quality evaluation method based on the deviation fine-tuning model according to claim 1 is characterized in that: The historical soil quality evaluation data is represented by the evaluation data of the sampling point location N times before, and the sampling point locations corresponding to the N times of evaluation data are all limited to the same field; The steps of the expert evaluation method are: forming a diverse expert group, preliminary selection of indicators, preliminary evaluation, expert scoring, summary and screening, and final screening.

3. The soil quality evaluation method based on the deviation fine-tuning model according to claim 2 is characterized in that: The second step of obtaining the preset index includes: Determine the original indicators of the first preset indicators, and perform data standardization on the original indicators. Set the determined original indicators to have an average value of 0 and a standard deviation of 1. The specific formula is: Among them, X is the original data, μ is the mean, σ is the standard deviation, and Z is the standardized data set; The principal component analysis (PCA) method is used on the standardized data set to calculate the eigenvalue and eigenvector of each indicator; Suppose there are n original indicators and m samples, and the eigenvalue of each principal component j obtained by principal component analysis is λ j ; Setting λ j >M, where M represents the first design value, when the eigenvalue λ of the principal component j Only when it is greater than M, it can be used as the pre-selected indicator parameter for the second preset indicator; The conditional formula for defining the cumulative contribution rate is: Among them, k is the number of principal components selected, Ml is the cumulative contribution rate of the first design value, and when the pre-selected indicator parameters meet the cumulative contribution rate condition formula, the screened indicator parameters are used as the second preset indicators.

4. The soil quality evaluation method based on the deviation fine-tuning model according to claim 3 is characterized in that: The acquisition of the third preset indicators specifically includes the following contents: The analytic hierarchy process is used for the second preset index; the target layer is set as the black soil quality evaluation, The criterion layer is black soil nutrients, black soil environment, black soil layer thickness and black soil ecological indicators, which correspond to the soil chemical indicators, soil environmental indicators, soil physical indicators and soil biological indicators in step S2 in sequence. The criterion layer is a subset of various indicators collected in step S2; The indicator layers are organic matter, total nitrogen, available phosphorus, available potassium, water content, cation exchange capacity, humus content, iron, manganese, black soil layer thickness, soil layer thickness, gully erosion area, pH value, vegetation coverage, heavy metal concentration, land salinization, microbial content and enzyme activity indicators, which constitute the third preset indicators.

5. A soil information collection device, characterized in that: The device is used to execute a soil quality evaluation method based on a deviation fine-tuning model as described in any one of claims 1 to 4. The collection device includes a soil sampling robot, which is composed of a soil sampling and storage module, a solar power supply module, a walking mechanism based on GNSS positioning, a sensor detection module, and a wireless network transmission module, and includes a control module for control. The soil sampling robot also includes the following black soil quality evaluation system: Data acquisition module: used to obtain the GNSS signal of the GNSS receiver to determine the location of the sampling point, and obtain the corresponding historical soil quality evaluation data based on the sampling point location; The first indicator generation module is used to collect soil physical indicators, soil chemical indicators, soil environmental indicators and soil biological indicators of the sampling point, and each indicator includes a number of collection parameters. The several collection parameters included in each indicator are screened by expert judgment method to generate the first preset indicators; The second indicator generation module is used to obtain the first preset indicator, standardize the original data of the first preset indicator, and perform principal component analysis on the processed data to generate the second preset indicator; Model building module: used to obtain the historical second preset index contained in the historical soil quality evaluation data, compare and analyze the historical second preset index with the second preset index of step S3, and generate a deviation fine-tuning model; the deviation fine-tuning model is represented as a deviation value D; The third indicator generation module: after obtaining the second preset indicator, perform a hierarchical analysis method, the hierarchical analysis method includes a target layer, a criterion layer and an indicator layer, and measure the indicator value of the indicator layer to generate the third preset indicator; Index generation module: used to obtain the third preset index and then analyze and process it, generate the soil quality classification index S, and classify the soil quality grades based on the soil quality classification index S. The formula of the soil quality classification index S is as follows; Among them, S is the soil quality classification index, i represents the i-th index in the third preset index, and f i Score the soil sample for the i-th indicator in the third preset indicator, W i It represents the relative weight of the ith indicator in the third preset indicator; Fine-tuning construction module: After obtaining the deviation value D corresponding to the deviation fine-tuning model, the value range of the deviation value D is divided into grade intervals. Each grade interval represents the corresponding adjustment range, and the output result of the soil quality classification index S is fine-tuned according to the adjustment range.

Images