Land comprehensive improvement quality evaluation method and system based on big data

Abstract

The invention relates to the technical field of land resource management, and discloses a big data-based land comprehensive improvement quality evaluation method and system, and the method comprises the steps: obtaining the initial land data of an improvement region through a satellite remote sensing technology and a GIS, and carrying out the preprocessing of the initial land data, and obtaining the preprocessed land data; determining an evaluation index system of the renovation area and determining an index weight corresponding to each index by adopting an analytic hierarchy process; performing data analysis on the preprocessed land data to extract feature data related to the evaluation index system, and evaluating the feature data by adopting a land comprehensive improvement quality evaluation model to obtain an evaluation result; the evaluation result is displayed in a visual mode, and feedback suggestions and suggestions are provided for comprehensive land improvement according to the evaluation result; according to the invention, the automatic evaluation of the land comprehensive improvement quality is realized, and the evaluation efficiency and objectivity are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of land resource management, and in particular to a method and system for evaluating the quality of comprehensive land improvement based on big data. Background Art

[0002] Comprehensive land improvement is a systematic project that promotes the comprehensive improvement of fields, water, roads, forests, and villages within a certain area, in accordance with the goals and uses determined by the overall land use plan, with land consolidation, reclamation, development, and the increase and decrease of urban and rural construction land as a platform, improves rural production, living conditions and ecological environment, promotes large-scale agricultural operations, concentrated population residence, industrial agglomeration development, and promotes the process of urban and rural integration. However, traditional land improvement quality assessment methods often rely on manual field surveys and limited data samples, and the assessment results are highly subjective, inefficient, and inaccurate. With the development of big data technology, new ideas and methods have been provided for the quality assessment of comprehensive land improvement. Summary of the invention

[0003] The purpose of the present invention is to solve the above problems and to design a land comprehensive improvement quality assessment method and system based on big data.

[0004] The first aspect of the present invention provides a method for evaluating the quality of comprehensive land improvement based on big data.

[0005] Optionally, in a first implementation of the first aspect of the present invention, the method for evaluating the quality of comprehensive land improvement based on big data comprises the following steps:

[0006] Using satellite remote sensing technology and GIS to obtain initial land data of the remediation area, preprocessing the initial land data to obtain preprocessed land data;

[0007] Determine the evaluation index system of the remediation area and use the analytic hierarchy process to determine the index weight corresponding to each index, wherein the evaluation index system at least includes land leveling, irrigation and drainage conditions, soil fertility, ecological environment quality and economic benefits;

[0008] Performing data analysis on the pre-processed land data to extract characteristic data related to the evaluation index system, and using a land comprehensive improvement quality evaluation model to evaluate the characteristic data to obtain an evaluation result;

[0009] The evaluation results will be displayed in a visual form, and feedback and suggestions will be provided for comprehensive land improvement based on the evaluation results.

[0010] Optionally, in a second implementation of the first aspect of the present invention, the using of satellite remote sensing technology and GIS to obtain initial land data of the remediation area, and preprocessing the initial land data to obtain preprocessed land data includes:

[0011] Obtain image data of the remediation area through satellite remote sensing equipment, obtain basic data of the remediation area from GIS, integrate the image data and the basic data to obtain initial land data, wherein the image data at least includes the topography, land use type and vegetation coverage of the area, and the basic data at least includes the topography, soil type and slope;

[0012] removing noise data from the initial land data to obtain first land data, and performing standardization processing on the first land data to obtain second land data;

[0013] The missing data in the second land data are interpolated by using regression interpolation method, and the interpolation effect is evaluated by root mean square error. After the missing values ​​are processed, the preprocessed land data are obtained.

[0014] Optionally, in a third implementation of the first aspect of the present invention, determining the evaluation index system of the remediation area and using the analytic hierarchy process to determine the index weight corresponding to each index includes:

[0015] Determine the evaluation index system of the remediation area and set X = {X1, X2, …, X n} is the indicator set of the evaluation indicator system of the remediation area, and a judgment matrix is ​​constructed:

[0016]

[0017] In the formula, a ij Represents X i Relative to X j The importance of ij =1 / a, a ij >0;

[0018] Calculate the product M of each row of the judgment matrix i , and calculate M i The nth root of Perform vector normalization:

[0019]

[0020] Where W i Indicates the weight of each indicator;

[0021] The maximum characteristic root of the judgment matrix is ​​calculated, and the consistency index CI, random consistency RI and consistency ratio CR are calculated based on the maximum characteristic root:

[0022] CR=CI / RI

[0023] If CR < 0.1, the consistency is passed, otherwise, the judgment matrix is ​​reconstructed and the consistency check is performed again until the consistency requirements are met.

[0024] Optionally, in a fourth implementation of the first aspect of the present invention, the step of performing data analysis on the pre-processed land data to extract characteristic data related to the evaluation index system includes:

[0025] The LOF algorithm is used to calculate the independent aggregation of each feature in the preprocessed land data, and the centroid method is used to calculate the independent separability of each feature. The scores of all features are calculated, and the first m features with the highest scores are selected to form the initial feature subset T0;

[0026] Select the i-th feature in descending order, and form a new feature subset T with the initial feature subset. Use the LOF algorithm to calculate the aggregation of the new feature subset, and then use the centroid method to calculate the separability of the new feature subset and calculate the score of the new feature subset.

[0027] Determine whether the score of the new feature subset is less than the score of the initial feature subset. If so, i=i+1 and continue to iterate. Otherwise, T0=T, output the selected feature subset and obtain feature data.

[0028] Optionally, in a fifth implementation of the first aspect of the present invention, the use of a land comprehensive improvement quality assessment model to assess the characteristic data to obtain an assessment result includes:

[0029] The random forest algorithm is used to build a land comprehensive improvement quality assessment model, randomly generate k sample training sets containing land data, and generate a decision tree at the same time;

[0030] Select m variables in the sample training set and perform node growth, calculate the error between the predicted value of a single tree and the actual value, combine the prediction results of all decision trees, and calculate the prediction value of the random forest;

[0031] The SSA algorithm is used to iteratively optimize the model hyperparameters of the land comprehensive improvement quality assessment model to obtain the current optimal model hyperparameters;

[0032] The first-order eigenvalue of the input is used as the input of the next order, and features are extracted in the network in a recursive manner. The data input into the land comprehensive improvement quality assessment model is classified, and the hidden layer activation value of the last layer of the network is used as the input of the classifier, and then the assessment result is output.

[0033] Optionally, in a sixth implementation of the first aspect of the present invention, the iterative optimization of the model hyperparameters of the land comprehensive improvement quality assessment model using the SSA algorithm to obtain the current optimal model hyperparameters includes:

[0034] Assume that the number of sparrows in population X is n, and the sparrows are divided into discoverers, trackers, and caretakers according to their different functions;

[0035] The discoverer is responsible for searching the area where the food is located. When a natural enemy is found, all individuals must leave the foraging area and update their location.

[0036] The tracker is a parallel search individual of the discoverer. The tracker uses half the number of individuals as the update trigger condition. When the tracker is greater than n / 2, it means that the fitness is low and it goes to another area to search for food. Otherwise, it joins the current area to search for food together and updates the position.

[0037] The caretaker can be used for warning, using the fitness value as the trigger condition for position update, calculating and sorting the fitness, and the algorithm stops running after the maximum number of iterations is met.

[0038] The second aspect of the present invention provides a land comprehensive improvement quality assessment system based on big data, the land comprehensive improvement quality assessment system based on big data comprises a data preprocessing module, an indicator determination module, a quality assessment module and a visualization module, wherein:

[0039] A data preprocessing module is used to obtain initial land data of the remediation area using satellite remote sensing technology and GIS, and preprocess the initial land data to obtain preprocessed land data;

[0040] An indicator determination module is used to determine the evaluation indicator system of the remediation area and use the hierarchical analysis method to determine the indicator weight corresponding to each indicator, wherein the evaluation indicator system at least includes land flatness, irrigation and drainage conditions, soil fertility, ecological environment quality and economic benefits;

[0041] A quality assessment module is used to analyze the pre-processed land data to extract characteristic data related to the assessment index system, and use a comprehensive land improvement quality assessment model to assess the characteristic data to obtain an assessment result;

[0042] The visualization module is used to display the evaluation results in a visual form and provide feedback and suggestions for comprehensive land management based on the evaluation results.

[0043] Optionally, in a first implementation of the second aspect of the present invention, the data preprocessing module includes a data integration submodule, a standardization processing submodule and a missing value processing submodule, wherein:

[0044] The data integration submodule is used to obtain image data of the remediation area through satellite remote sensing equipment, obtain basic data of the remediation area from GIS, and integrate the image data and the basic data to obtain initial land data, wherein the image data at least includes the topography, land use type and vegetation coverage of the area, and the basic data at least includes the topography, soil type and slope;

[0045] A standardization processing submodule, used for removing noise data in the initial land data to obtain first land data, and performing standardization processing on the first land data to obtain second land data;

[0046] The missing value processing submodule is used to interpolate the missing data in the second land data by using the regression interpolation method, and evaluate the interpolation effect by the root mean square error. After the missing value processing, the preprocessed land data is obtained.

[0047] Optionally, in a second implementation of the second aspect of the present invention, the quality assessment module includes a construction submodule, a calculation submodule, an iteration submodule and an output submodule, wherein:

[0048] Construct a submodule for constructing a land comprehensive improvement quality assessment model using a random forest algorithm, randomly generating k sample training sets containing land data, and generating a decision tree at the same time;

[0049] The calculation submodule is used to select m variables in the sample training set and perform node growth, calculate the error between the predicted value of a single tree and the actual value, integrate the prediction results of all decision trees, and calculate the prediction value of the random forest;

[0050] An iterative submodule is used to iteratively optimize the model hyperparameters of the land comprehensive improvement quality assessment model using an SSA algorithm to obtain the current optimal model hyperparameters;

[0051] The output submodule is used to take the input first-order eigenvalue as the input of the next order, and extract features in the network in a recursive manner, classify the data input into the land comprehensive improvement quality assessment model, take the hidden layer activation value of the last layer of the network as the input of the classifier, and then output the assessment result.

[0052] In the technical solution provided by the present invention, satellite remote sensing technology and GIS are used to obtain initial land data of the remediation area, the initial land data are preprocessed to obtain preprocessed land data; an evaluation index system for the remediation area is determined and a hierarchical analysis method is used to determine the index weight corresponding to each index; data analysis is performed on the preprocessed land data to extract feature data related to the evaluation index system, and a land comprehensive remediation quality evaluation model is used to evaluate the feature data to obtain an evaluation result; the evaluation result is displayed in a visual form, and feedback and suggestions are provided for the land comprehensive remediation based on the evaluation result; the present invention utilizes big data technology to collect and process a large amount of multi-source data, thereby improving the accuracy and comprehensiveness of the evaluation, and by establishing an evaluation index system and adopting a machine learning algorithm, an automated evaluation of the quality of land comprehensive remediation is realized, the evaluation efficiency and objectivity are improved, and the result visualization and feedback functions provide strong support for the management and decision-making of land comprehensive remediation projects. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the following detailed description of the preferred embodiment.The drawings are only for the purpose of illustrating the preferred embodiments and are not to be construed as limiting the invention.

[0054] Figure 1 A schematic diagram of an embodiment of a land comprehensive improvement quality assessment method based on big data provided by an embodiment of the present invention;

[0055] Figure 2 A schematic diagram of the structure of a land comprehensive improvement quality assessment system based on big data provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0056] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, device, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0057] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1The embodiment of the present invention provides a schematic diagram of a method for evaluating the quality of comprehensive land improvement based on big data, which specifically includes the following steps:

[0058] Step 101: using satellite remote sensing technology and GIS to obtain initial land data of the remediation area, and preprocessing the initial land data to obtain preprocessed land data;

[0059] Step 102, determine the evaluation index system of the remediation area and use the analytic hierarchy process to determine the index weight corresponding to each index;

[0060] In this embodiment, the evaluation index system includes at least land leveling, irrigation and drainage conditions, soil fertility, ecological environment quality and economic benefits;

[0061] Step 103: Analyze the pre-processed land data to extract characteristic data related to the evaluation index system, and use the land comprehensive improvement quality evaluation model to evaluate the characteristic data to obtain the evaluation results;

[0062] Step 104: Display the evaluation results in a visual form, and provide feedback and suggestions for comprehensive land management based on the evaluation results.

[0063] In this embodiment, the evaluation results are displayed in the form of graphs, charts or other intuitive visualizations to facilitate a clearer understanding and analysis of the effects and problems of land consolidation. In this way, complex data and information can be converted into easy-to-understand visual elements, such as bar charts, line charts, map markers, etc., to help decision makers and relevant stakeholders quickly grasp the overall situation and key indicators of land consolidation; after completing the visualization of the evaluation results, based on these results, a series of targeted feedback and suggestions can be provided for comprehensive land consolidation. These feedback and suggestions will be based on the advantages and disadvantages found in the evaluation, and aim to guide the further optimization and improvement of land consolidation work. For example, if the evaluation results show that the land utilization rate in certain areas is low, it can be recommended to take corresponding measures such as land reclamation, improving soil quality or adjusting land use planning. If it is found that there is a waste of resources or environmental impact problems in the process of land consolidation, a more environmentally friendly and sustainable consolidation plan may be proposed, as well as suggestions for strengthening supervision and evaluation mechanisms. In short, these feedback and suggestions will be committed to improving the efficiency and effectiveness of land consolidation and ensuring the rational use and sustainable development of land resources.

[0064] In this embodiment, satellite remote sensing technology and GIS are used to obtain initial land data of the remediation area, and the initial land data is preprocessed to obtain preprocessed land data, including:

[0065] Obtain image data of the remediation area through satellite remote sensing equipment, obtain basic data of the remediation area from GIS, integrate the image data and basic data to obtain initial land data, where the image data at least includes the topography, land use type and vegetation coverage of the area, and the basic data at least includes the topography, soil type and slope;

[0066] removing noise data from the initial land data to obtain first land data, and performing standardization processing on the first land data to obtain second land data;

[0067] The regression interpolation method is used to interpolate the missing data in the second land data, and the interpolation effect is evaluated by the root mean square error. After the missing values ​​are processed, the preprocessed land data are obtained.

[0068] In this embodiment, an evaluation index system for the remediation area is determined and the index weight corresponding to each index is determined by using the hierarchical analysis method, including:

[0069] Determine the evaluation index system of the remediation area and set X = {X1, X2, …, X n} is the indicator set of the evaluation indicator system of the remediation area, and a judgment matrix is ​​constructed:

[0070]

[0071] In the formula, a ij Represents X i Relative to X j The importance of ij =1 / a, a ij >0;

[0072] Calculate the product M of each row of the judgment matrix i , and calculate M i The nth root of Perform vector normalization:

[0073]

[0074] Where W i Indicates the weight of each indicator;

[0075] Calculate the maximum eigenvalue of the judgment matrix, and calculate the consistency index CI, random consistency RI, and consistency ratio CR based on the maximum eigenvalue:

[0076] CR=CI / RI

[0077] If CR < 0.1, the consistency is passed, otherwise, the judgment matrix is ​​reconstructed and the consistency check is performed again until the consistency requirements are met.

[0078] In this embodiment, data analysis is performed on the pre-processed land data to extract characteristic data related to the evaluation index system, including:

[0079] The LOF algorithm is used to calculate the independent aggregation of each feature in the preprocessed land data, and the centroid method is used to calculate the independent separability of each feature. The scores of all features are calculated, and the first m features with the highest scores are selected to form the initial feature subset T0;

[0080] Select the i-th feature in descending order, and form a new feature subset T with the initial feature subset. Use the LOF algorithm to calculate the aggregation of the new feature subset, and then use the centroid method to calculate the separability of the new feature subset and calculate the score of the new feature subset.

[0081] Determine whether the score of the new feature subset is less than the score of the initial feature subset. If so, i=i+1 and continue to iterate. Otherwise, T0=T, output the selected feature subset and obtain feature data.

[0082] In this embodiment, the land comprehensive improvement quality assessment model is used to assess the characteristic data to obtain the assessment results, including:

[0083] The random forest algorithm is used to build a land comprehensive improvement quality assessment model, randomly generate k sample training sets containing land data, and generate a decision tree at the same time;

[0084] Select m variables in the sample training set and perform node growth, calculate the error between the predicted value of a single tree and the actual value, combine the prediction results of all decision trees, and calculate the prediction value of the random forest;

[0085] The SSA algorithm is used to iteratively optimize the model hyperparameters of the land comprehensive improvement quality assessment model to obtain the current optimal model hyperparameters;

[0086] The first-order eigenvalue of the input is used as the input of the next order, and features are extracted in the network in a recursive manner. The data input into the land comprehensive improvement quality assessment model is classified, and the hidden layer activation value of the last layer of the network is used as the input of the classifier to output the assessment result.

[0087] In this embodiment, the SSA algorithm is used to iteratively optimize the model hyperparameters of the land comprehensive improvement quality assessment model to obtain the current optimal model hyperparameters, including:

[0088] Assume that the number of sparrows in population X is n, and the sparrows are divided into discoverers, trackers, and caretakers according to their different functions;

[0089] The discoverer is responsible for searching the area where the food is located. When a natural enemy is found, all individuals must leave the foraging area and update their location.

[0090] The tracker is a parallel search individual of the discoverer. The tracker uses half the number of individuals as the update trigger condition. When the tracker is greater than n / 2, it means that the fitness is low and it goes to another area to search for food. Otherwise, it joins the current area to search for food together and updates the position.

[0091] The caretaker can be used for warning, using the fitness value as the trigger condition for position update, calculating and sorting the fitness, and the algorithm stops running after the maximum number of iterations is met.

[0092] See also Figure 2 , a schematic diagram of the structure of a land comprehensive improvement quality assessment system based on big data provided by an embodiment of the present invention, the system system includes a data preprocessing module, an indicator determination module, a quality assessment module and a visualization module, wherein,

[0093] The data preprocessing module is used to obtain the initial land data of the remediation area by using satellite remote sensing technology and GIS, and preprocess the initial land data to obtain the preprocessed land data;

[0094] The indicator determination module is used to determine the evaluation indicator system of the remediation area and use the hierarchical analysis method to determine the indicator weight corresponding to each indicator, where the evaluation indicator system at least includes land flatness, irrigation and drainage conditions, soil fertility, ecological environment quality and economic benefits;

[0095] The quality assessment module is used to analyze the pre-processed land data to extract characteristic data related to the assessment index system, and use the land comprehensive improvement quality assessment model to assess the characteristic data to obtain the assessment results;

[0096] The visualization module is used to display the evaluation results in a visual form and provide feedback and suggestions for comprehensive land management based on the evaluation results.

[0097] In this embodiment, the data preprocessing module includes a data integration submodule, a standardization processing submodule and a missing value processing submodule, wherein:

[0098] The data integration submodule is used to obtain the image data of the remediation area through satellite remote sensing equipment, obtain the basic data of the remediation area from GIS, integrate the image data and the basic data to obtain the initial land data, wherein the image data at least includes the topography, land use type and vegetation coverage of the area, and the basic data at least includes the topography, soil type and slope;

[0099] A standardization processing submodule is used to remove noise data in the initial land data to obtain first land data, and perform standardization processing on the first land data to obtain second land data;

[0100] The missing value processing submodule is used to interpolate the missing data in the second land data by using the regression interpolation method, and evaluate the interpolation effect by the root mean square error. After the missing value processing, the preprocessed land data is obtained.

[0101] In this embodiment, the quality assessment module includes a construction submodule, a calculation submodule, an iteration submodule and an output submodule, wherein:

[0102] Construct a submodule for constructing a land comprehensive improvement quality assessment model using a random forest algorithm, randomly generating k sample training sets containing land data, and generating a decision tree at the same time;

[0103] The calculation submodule is used to select m variables in the sample training set and perform node growth, calculate the error between the predicted value of a single tree and the actual value, integrate the prediction results of all decision trees, and calculate the prediction value of the random forest;

[0104] The iterative submodule is used to iteratively optimize the model hyperparameters of the land comprehensive improvement quality assessment model using the SSA algorithm to obtain the current optimal model hyperparameters;

[0105] The output submodule is used to take the first-order eigenvalues ​​of the input as the input of the next order, and to extract features in the network in a recursive manner, to classify the data input into the land comprehensive improvement quality assessment model, and to take the hidden layer activation value of the last layer of the network as the input of the classifier, and then output the assessment results.

[0106] Through the implementation of the above scheme,

[0107] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A land comprehensive improvement quality assessment method based on big data, characterized in that: The land comprehensive improvement quality assessment method based on big data includes the following steps: Using satellite remote sensing technology and GIS to obtain initial land data of the remediation area, preprocessing the initial land data to obtain preprocessed land data; Determine the evaluation index system of the remediation area and use the analytic hierarchy process to determine the index weight corresponding to each index, wherein the evaluation index system at least includes land leveling, irrigation and drainage conditions, soil fertility, ecological environment quality and economic benefits; Performing data analysis on the pre-processed land data to extract characteristic data related to the evaluation index system, and using a land comprehensive improvement quality evaluation model to evaluate the characteristic data to obtain an evaluation result; The evaluation results will be displayed in a visual form, and feedback and suggestions will be provided for comprehensive land improvement based on the evaluation results.

2. The land comprehensive improvement quality assessment method based on big data as claimed in claim 1 is characterized in that: The method of using satellite remote sensing technology and GIS to obtain initial land data of the remediation area and preprocessing the initial land data to obtain preprocessed land data includes: Obtain image data of the remediation area through satellite remote sensing equipment, obtain basic data of the remediation area from GIS, integrate the image data and the basic data to obtain initial land data, wherein the image data at least includes the topography, land use type and vegetation coverage of the area, and the basic data at least includes the topography, soil type and slope; removing noise data from the initial land data to obtain first land data, and performing standardization processing on the first land data to obtain second land data; The missing data in the second land data are interpolated by using regression interpolation method, and the interpolation effect is evaluated by root mean square error. After the missing values ​​are processed, the preprocessed land data are obtained.

3. The land comprehensive improvement quality assessment method based on big data as claimed in claim 1 is characterized in that: The evaluation index system for determining the remediation area and using the analytic hierarchy process to determine the index weight corresponding to each index include: Determine the evaluation index system of the remediation area and set X = {X1, X2, …, X n } is the indicator set of the evaluation indicator system of the remediation area, and a judgment matrix is ​​constructed: In the formula, a ij Represents X i Relative to X j The importance of ij =1 / a, a ij >0; Calculate the product M of each row of the judgment matrix i , and calculate M i The nth root of Perform vector normalization: Where W i Indicates the weight of each indicator; The maximum characteristic root of the judgment matrix is ​​calculated, and the consistency index CI, random consistency RI and consistency ratio CR are calculated based on the maximum characteristic root: CR=CI / RI If CR < 0.1, the consistency is passed, otherwise, the judgment matrix is ​​reconstructed and the consistency check is performed again until the consistency requirements are met.

4. The land comprehensive improvement quality assessment method based on big data as claimed in claim 1 is characterized in that: The data analysis of the pre-processed land data to extract characteristic data related to the evaluation index system includes: The LOF algorithm is used to calculate the independent aggregation of each feature in the preprocessed land data, and the centroid method is used to calculate the independent separability of each feature. The scores of all features are calculated, and the first m features with the highest scores are selected to form the initial feature subset T0; Select the i-th feature in descending order, and form a new feature subset T with the initial feature subset. Use the LOF algorithm to calculate the aggregation of the new feature subset, and then use the centroid method to calculate the separability of the new feature subset and calculate the score of the new feature subset. Determine whether the score of the new feature subset is less than the score of the initial feature subset. If so, i=i+1 and continue to iterate. Otherwise, T0=T, output the selected feature subset and obtain feature data.

5. The land comprehensive improvement quality assessment method based on big data as claimed in claim 1 is characterized in that: The characteristic data is evaluated by using the land comprehensive improvement quality evaluation model to obtain evaluation results, including: The random forest algorithm is used to build a land comprehensive improvement quality assessment model, randomly generate k sample training sets containing land data, and generate a decision tree at the same time; Select m variables in the sample training set and perform node growth, calculate the error between the predicted value of a single tree and the actual value, combine the prediction results of all decision trees, and calculate the prediction value of the random forest; The SSA algorithm is used to iteratively optimize the model hyperparameters of the land comprehensive improvement quality assessment model to obtain the current optimal model hyperparameters; The first-order eigenvalue of the input is used as the input of the next order, and features are extracted in the network in a recursive manner. The data input into the land comprehensive improvement quality assessment model is classified, and the hidden layer activation value of the last layer of the network is used as the input of the classifier, and then the assessment result is output.

6. The method for evaluating land comprehensive improvement quality based on big data according to claim 5, characterized in that: The SSA algorithm is used to iteratively optimize the model hyperparameters of the land comprehensive improvement quality assessment model to obtain the current optimal model hyperparameters, including: Assume that the number of sparrows in population X is n, and the sparrows are divided into discoverers, trackers, and caretakers according to their different functions; The discoverer is responsible for searching the area where the food is located. When a natural enemy is found, all individuals must leave the foraging area and update their location. The tracker is a parallel search individual of the discoverer. The tracker uses half the number of individuals as the update trigger condition. When the tracker is greater than n / 2, it means that the fitness is low and it goes to another area to search for food. Otherwise, it joins the current area to search for food together and updates the position. The caretaker can be used for warning, using the fitness value as the trigger condition for position update, calculating and sorting the fitness, and the algorithm stops running after the maximum number of iterations is met.

7. The land comprehensive improvement quality assessment system based on big data is characterized by: The land comprehensive improvement quality assessment system based on big data includes a data preprocessing module, an indicator determination module, a quality assessment module and a visualization module, wherein: A data preprocessing module is used to obtain initial land data of the remediation area using satellite remote sensing technology and GIS, and preprocess the initial land data to obtain preprocessed land data; An indicator determination module is used to determine the evaluation indicator system of the remediation area and use the hierarchical analysis method to determine the indicator weight corresponding to each indicator, wherein the evaluation indicator system at least includes land flatness, irrigation and drainage conditions, soil fertility, ecological environment quality and economic benefits; A quality assessment module is used to analyze the pre-processed land data to extract characteristic data related to the assessment index system, and use a comprehensive land improvement quality assessment model to assess the characteristic data to obtain an assessment result; The visualization module is used to display the evaluation results in a visual form and provide feedback and suggestions for comprehensive land management based on the evaluation results.

8. The land comprehensive improvement quality assessment system based on big data as claimed in claim 7 is characterized in that: The data preprocessing module includes a data integration submodule, a standardization processing submodule and a missing value processing submodule, wherein: The data integration submodule is used to obtain image data of the remediation area through satellite remote sensing equipment, obtain basic data of the remediation area from GIS, and integrate the image data and the basic data to obtain initial land data, wherein the image data at least includes the topography, land use type and vegetation coverage of the area, and the basic data at least includes the topography, soil type and slope; A standardization processing submodule, used for removing noise data in the initial land data to obtain first land data, and performing standardization processing on the first land data to obtain second land data; The missing value processing submodule is used to interpolate the missing data in the second land data by using the regression interpolation method, and evaluate the interpolation effect by the root mean square error. After the missing value processing, the preprocessed land data is obtained.

9. The land comprehensive improvement quality assessment system based on big data as claimed in claim 7, characterized in that: The quality assessment module includes a construction submodule, a calculation submodule, an iteration submodule and an output submodule, wherein: Construct a submodule for constructing a land comprehensive improvement quality assessment model using a random forest algorithm, randomly generating k sample training sets containing land data, and generating a decision tree at the same time; The calculation submodule is used to select m variables in the sample training set and perform node growth, calculate the error between the predicted value of a single tree and the actual value, integrate the prediction results of all decision trees, and calculate the prediction value of the random forest; An iterative submodule is used to iteratively optimize the model hyperparameters of the land comprehensive improvement quality assessment model using an SSA algorithm to obtain the current optimal model hyperparameters; The output submodule is used to take the input first-order eigenvalue as the input of the next order, and extract features in the network in a recursive manner, classify the data input into the land comprehensive improvement quality assessment model, take the hidden layer activation value of the last layer of the network as the input of the classifier, and then output the assessment result.

Images