Multi-objective land utilization optimization method based on spatial data

Abstract

The invention provides a multi-objective land utilization optimization method based on spatial data, which integrates land surface settlement data as a key constraint variable into an optimization model, and quantifies land surface settlement by integrating land utilization, social economy and ecological data and utilizing sentry SAR data, so that the method has the advantages that the land utilization efficiency is improved, and the land utilization efficiency is improved. A multi-objective function which pursues economic benefit maximization and ecological benefit maximization at the same time is constructed; and then, carrying out multi-objective optimization by adopting an improved non-dominated genetic algorithm (NSGA-II), and realizing spatial refined distribution of the optimized land utilization structure by applying a PLUS model. According to the method, the defect that the dynamic change of the geological environment is neglected in the existing method is overcome, and the optimization result can truly reflect the ground surface settlement risk, so that a scientific decision basis is provided for making a land utilization plan considering safety, development and protection in rapid development and geological sensitive areas.

Description

Technical Field

[0001] This invention relates to the field of land use optimization and ecosystem service value assessment, and more specifically, to a multi-objective land use optimization method based on spatial data. Background Technology

[0002] The rational use of land resources is directly related to the sustainability of the ecological environment and the healthy development of the economy. With the acceleration of urbanization, land use problems are becoming increasingly prominent, threatening ecosystem services and triggering a series of problems such as resource depletion and environmental degradation. From a global perspective, the services provided by ecosystems, including climate regulation, water supply, soil protection, and biodiversity maintenance, are crucial and have irreplaceable economic and environmental value.

[0003] Numerous studies have shown that the economic value of ecosystem services can be quantified not only through their direct products but also systematically assessed using value-per-unit-area equivalent factors. This assessment method can provide clear economic indicators for services across different ecosystem types, thus offering quantitative support for land-use decisions. Nevertheless, traditional land-use decisions often prioritize maximizing economic benefits, neglecting the sustainability and long-term benefits of the ecological environment. This can damage ecosystem health and reduce its potential for sustainable development. For example, in some resource-depleted cities, overdevelopment coupled with neglect of ecological protection has ultimately led to a sharp decline in municipal conditions, impacting overall social benefits.

[0004] Against this backdrop, multi-objective optimization methods have gradually gained attention. They can comprehensively consider factors at multiple levels, including economic, social, and ecological aspects, thereby maximizing ecological and environmental protection while achieving economic growth. However, existing research has particularly lacked consideration of land subsidence, a crucial influencing factor. This invention innovatively incorporates land subsidence data into the multi-objective land use optimization process as an important variable, thereby enhancing the scientific rigor and practical applicability of the model.

[0005] By integrating land subsidence data, we can comprehensively assess the impact of land use decisions, providing policymakers with effective data-driven support and helping to achieve a win-win situation for economic development and ecological protection. Therefore, we propose a multi-objective land use optimization method based on spatial data. Summary of the Invention

[0006] This invention provides a multi-objective land use optimization method based on spatial data. Its core innovation lies in the first-time integration of land subsidence data as a key constraint variable into the optimization model. This method integrates land use, socioeconomic, and ecological data, and uses Sentinel SAR data to quantify land subsidence, constructing a multi-objective function that simultaneously maximizes economic and ecological benefits. Then, an improved Non-Dominated Genetic Algorithm (NSGA-II) is employed for multi-objective optimization, and the PLUS model is used to achieve a refined spatial distribution of the optimized land use structure. This invention overcomes the shortcomings of existing methods that neglect dynamic changes in the geological environment, enabling the optimization results to accurately reflect land subsidence risks. This provides a scientific basis for decision-making in rapidly developing and geologically sensitive areas, balancing safety, development, and protection in land use planning.

[0007] This invention provides the following technical solution: a multi-objective land use optimization method based on spatial data, comprising the following steps: Step 1: Construct a dataset of economic and ecological benefit functions and constraints for the northern Anhui region; Step 2: Calculate surface subsidence data for the northern Anhui region; Step 3: Use the grey prediction model (GM(1,1)) to calculate the future economic benefit index and ecological benefit index, and collect land use data and set up multiple scenarios; Step 4: Construct a multi-objective optimization model and solve the model using the Non-Dominated Sorting Genetic Algorithm (NSGA-II); Step 5: Simulate and evolve the spatial pattern based on the PLUS model.

[0008] As a preferred technical solution of the present invention, step 1 specifically includes: Step 1.1: Collect statistical yearbook data for cities in northern Anhui; Obtain statistical yearbook data from the statistical bureaus of various cities in northern Anhui Province between 2010 and 2020, and add the data for the corresponding years to obtain the overall statistical data for northern Anhui Province. Step 1.2: Preliminary construction of function values ​​under the economic development scenario of northern Anhui region; The economic development scenario sets maximizing economic benefits as its core objective, emphasizing increasing the proportion of high-economic-value land use types. The core of this scenario lies in fully releasing the economic potential of land resources through optimal allocation, thereby enhancing the region's economic output capacity. The corresponding objective function can be expressed as: In the formula: f ( X) This indicates the economic value corresponding to land use in northern Anhui. X iFor the first i Area of ​​land use type; i =1,2,......,6 correspond to six land use types: cultivated land, forest land, grassland, water area, construction land, and unused land; No For the first i Economic output coefficients for different land use types. This study uses the economic value created per unit area by agriculture, forestry, animal husbandry, fisheries, and the secondary and tertiary industries as the economic value coefficients for arable land, forest land, grassland, water areas, and construction land. The economic value coefficient for unused land is 0. Therefore, statistical yearbook data is consulted, and the corresponding economic benefit coefficients for each year are calculated. E i .

[0009] As a preferred technical solution of the present invention, step 1.3 involves initially constructing function values ​​under the ecological protection scenario in northern Anhui. The ecological protection scenario aims to maximize ecological benefits, with its core objective focusing on enhancing ecosystem service functions to promote carbon storage growth and improve the ecological environment. Under this scenario, priority is given to improving the regional ecosystem service value through rational optimization of land use structure, thereby maximizing the ecosystem service value of Henan Province. The corresponding objective function can be expressed as follows: In the formula, h ( X This represents the total value of ecosystem services for land use types in northern Anhui. X i For the first i The area of ​​land use type, of which i =1,2,......,6 correspond to six land use types: cultivated land, forest land, grassland, water area, construction land, and unused land; G i Then it is the first i Ecological value coefficients for different land use types. The study adopted the ecosystem service value equivalent table proposed by Xie Gaodi et al. Considering the economic development level of northern Anhui, one standard unit of ecosystem service value equivalent was defined as 1 / 7 of the economic value of the average annual grain yield of one hectare of farmland. The study used the average price, average yield, and average planted area of ​​cash crops such as rice, wheat, and corn in northern Anhui from 2010 to 2020 as basic data. The calculation formula is as follows: In the formula, E a This indicates the economic value of the food production services provided by a unit area of ​​farmland ecosystem. i As a type of crop; pi for i Average price of crops (yuan / ton); q i for i Yield per unit area of ​​crops (tons / hmm 2 ); m i for i Planting area of ​​crops ( hmm 2 ); M For all crop planting areas ( hmm 2 Therefore, a preliminary ecological benefit index for northern Anhui Province from 2010 to 2020 was constructed. Step 1.4: Data collection of constraints for simulating future spatial states; The latest administrative boundary map and elevation DEM map of the northern Anhui region were obtained from Bigmap GIS; data on primary and secondary roads, population density, annual precipitation, and aerosol density were obtained from the resource and environmental science data platform.

[0010] As a preferred technical solution of the present invention, step 2 specifically includes: Step 2.1, Impact of surface subsidence values; Land subsidence signifies far more than just a loss of ground elevation; it represents a chain reaction leading from fundamental degradation of land functions to structural failure of spatial planning. Firstly, it directly causes the decline of land's core function as a production and carrying capacity factor: in agriculture, this manifests as reduced productivity due to waterlogging and salinization of arable land; in construction, it manifests as depreciation in safety levels and asset value due to infrastructure damage. This functional degradation is essentially a continuous depletion of land's natural capital and economic value. Furthermore, by permanently altering the fundamental geographical parameter of surface elevation, subsidence undermines the physical basis of land use control, causing systemic risks of discrepancies between maps and reality to emerge from ecological protection red lines, flood control and drainage systems, and urban-rural layouts delineated based on the original topography. Ultimately, this process clearly reveals that human land use intensity has exceeded the critical point of geological carrying capacity and natural recovery, becoming a key negative indicator and an irreversible risk warning for measuring the sustainability of regional development models and the effectiveness of land use governance systems.

[0011] As a preferred technical solution of the present invention, step 2.2, surface subsidence calculation; The ultimate goal of the entire InSAR processing is to decompose and calculate the raw phase difference observed by radar into surface deformation; this is based on a fundamental physical relationship: In the formula The total interferometric phase is the phase difference between corresponding pixels in two SAR images, which is the sum of all contributions. , It's radar phase, the satellite is... t 1 and t 2 The phase of the radar echo signal received from the ground unified target; The wavelength of the electromagnetic waves emitted by satellite radar sensors; This refers to the displacement of the ground target along the satellite's line of sight during the two imaging sessions; The phase component contributing to the actual surface deformation; Phase contribution caused by surface undulations; The delay caused by the atmospheric effect on the speed of radar signal propagation; The random phase is caused by factors such as sensor noise, temporal decoherence, and spatial decoherence; after eliminating the terrain, atmospheric, and noise phase through data processing, a clean deformation phase is obtained. Therefore, the deformation along the line of sight is calculated. Then convert the viewing deformation into vertical deformation. The surface subsidence value was calculated.

[0012] As a preferred technical solution of the present invention, step 3 specifically includes: Step 3.1: Land use data collection; The land use data used in this study comes from the China Land Cover Dataset (CLCD) released by a research team from Wuhan University. This dataset has a spatial resolution of 30 meters. Based on the national standard "Classification of Current Land Use" (GB / T 21010-2017), we reclassified the original data into six types: cultivated land, forest land, grassland, water area, construction land, and unused land. Finally, we extracted and constructed a land use classification dataset covering northern Anhui Province. Step 3.2, Setting the Scene; By setting up differentiated scenario frameworks to simulate and predict regional development paths, the core of this study lies in using multiple scenarios to systematically extrapolate future development trends. This helps provide a scientific basis for formulating policies related to economic development and ecological protection, and optimizes the decision-making process. This study sets up eight scenario structures: The natural development scenario, where land use trends are mainly extrapolated from historical data without additional economic or policy intervention; the economic development scenario, which aims to maximize economic benefits and maximize the economic potential of land resources; the ecological protection scenario, which prioritizes enhancing ecosystem service functions to maximize ecological benefits; and the comprehensive development scenario, which balances economic and ecological benefits by coordinating the proportions of different land use types to seek a balance between economic development and ecological protection. Additionally, eight scenario structures are set up, including the natural development scenario, economic development scenario, ecological protection scenario, and comprehensive development scenario, all taking into account land subsidence data.

[0013] As a preferred technical solution of the present invention, step 4 specifically includes: Step 4.1: Construct a multi-objective optimization model; Multi-objective optimization models provide a mathematical framework for solving optimization problems with multiple conflicting objectives, expressed as follows: In the formula, F ( x () represents the objective function, containing m objective functions. For decision variables, including n There are 1 variable; the constraint condition is defined as: In the formula and Let these represent equality constraints and inequality constraints, respectively. k , J These are the number of equality constraints and the number of inequality constraints, respectively.

[0014] As a preferred technical solution of the present invention, step 4.2 involves selecting a non-dominated sorting genetic algorithm; The algorithm (Non-dominated Sorting Genetic Algorithm II, NSGA-II) is based on fast non-dominated sorting, crowding distance calculation and elite retention strategy. It has the characteristics of high computational efficiency and strong convergence. Its specific steps are as follows: (1) Randomly generate the initial population. ,in It is a vector of decision variables that needs to satisfy initial conditions and constraints, and then the objective function value of each solution is calculated. (2) Define the population into multiple non-dominant levels based on the dominance relationship. And the condition for a solution to dominate a solution is: and The sorting results must satisfy: F 1 The solution contains no other Pareto front solutions. F 2 :quilt F 1Dominate but not dominate F 1 The set of solutions, and so on; let the solution... X i The number of dominant solutions is Its dominating set is (3) To maintain population diversity, crowding distance is calculated at each level. The formula is In the formula and In the target upper solution X i The objective value of the adjacent solutions, and The goal The maximum and minimum values; (4) Using binary tournament selection, priority is given to solutions with lower non-dominated levels. If the levels are the same, the crowding distance is compared, i.e., if Then choose X i ;like p = q Then choose the solution with the larger crowding distance, i.e. Time Selection X i (5) For crossover and mutation operations, simulated binary crossover (SBX) and polynomial mutation are used to generate offspring populations. When performing crossover operations, it is assumed that there are two parent solutions. and To generate child solutions Y 1 , Y 2 : in In the formula It is a cross-distribution index Related random numbers, It is a random number belonging to the interval [0,1]; in the mutation operation, for , after mutation Calculated as , It can be expressed by the following formula: In the formula It is the distribution index of variation. It is a random number within the interval [0,1]; (6) Merge parent populations and offspring population Generate a joint population ,right R t Perform non-dominated sorting and crowding distance calculation, and select the top... N Each solution constitutes the next generation population. The joint population selection formula is: If the last level Exceed N Select examples in descending order; If the maximum number of iterations is reached If the population objective function converges, the algorithm terminates and returns all non-dominated solutions in the population as the final Pareto front solution set. The economic and ecological value coefficients of land use types in 2030 are predicted by GM(1,1), and an NSGA-II optimization model is constructed based on this.

[0015] As a preferred technical solution of the present invention, step 5 specifically includes: Step 5.1, Land Expansion Analysis Strategy; The LEAS model, as one of the core modules of the PLUS model, extracts and samples the expansion areas of various land use types from the land use data of the study area, and then applies the random forest algorithm to quantify the development probability of different land use types. The core of this module is the random forest algorithm, whose formula is: , In the formula Indicates the first i Each unit has different land use types k The probability of growth; d The value is 0 or 1. A value of 1 indicates that other land use types have been converted to land use type k, and a value of 0 indicates other conversions. x A vector consisting of multiple driving factors; I ( ) is the indicator function of the decision tree; h ( x ) is the first n The prediction types of each decision tree; M This represents the total number of decision trees.

[0016] As a preferred technical solution of the present invention, step 5.2, CA model of multiple types of random patch seeds; CARS (CA Model Based on Multiple Random Patch Seeds): The PLUS model uses a threshold-decreasing degree-type random patch seed generation mechanism to simulate the spatial evolution of land use types. It generates dynamically changing "seeds" on the land use type development probability surface output by the LEAS module, using the following formula: In the formula, Indicates the first i Land use type of each cell k The probability of growth, Future land use types k The impact of demand, Represents a cell i Neighborhood effect r A random number between 0 and 1. Land use type k The threshold for the generation of new land patches; if a new land use type c If a player wins a round of competition, a decreasing threshold is applied. The evaluation is based on the following formula: , In the formula and yes t-1 generation and first t During iteration, land use type c The difference between the current quantity and future demand. Decreasing threshold The attenuation factor ranges from 0 to 1. rl The values ​​are normally distributed random values ​​with a mean of 1. l It is a decay series. It defines whether land use types are permitted. k Convert to c The transformation matrix; By simultaneously defining constraints and objective functions, and combining calculated surface subsidence data, various development scenarios are established to optimize both the quantity and spatial aspects of land use types. First, a land use quantity optimization scheme is established using the NSGA-II model. Then, the PLUS model is used to predict the development probability and expansion impact of various land use types under different future development scenarios, establishing a land use spatial optimization scheme. Finally, the land use status of the northern Anhui region under multiple scenarios is simulated.

[0017] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. By integrating land subsidence data into the model, the actual state of the natural environment can be more accurately reflected in the land use decision-making process. This improves the scientific level of land use decision-making and reduces potential environmental risks.

[0018] 2. By combining the NSGA-II and PLUS models, a balance is effectively struck between different development goals, promoting both economic efficiency and ecological environmental protection, which aligns with the concept of sustainable development. This not only improves the efficiency of land resource utilization but also maximizes ecological benefits.

[0019] 3. Eight development scenarios are set up, including four scenarios that do not consider land subsidence (such as natural development, economic development priority, ecological protection priority, and economic-ecological balance) and four scenarios that do consider land subsidence. This rich array of scenarios enables policymakers to conduct in-depth analysis and selection under different environmental and social conditions, flexibly respond to various real-world challenges, and effectively support long-term planning.

[0020] 4. Since rapid urbanization often leads to irrational resource allocation, this research provides a scientific basis for addressing these challenges, especially in optimizing land use in rapidly developing and water-scarce cities, and helps to reconcile the contradiction between economic development and ecological protection. Attached Figure Description

[0021] Figure 1 The method flowchart provided by the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention.

[0023] Therefore, the following detailed description of the embodiments of the present invention is not intended to limit the scope of the claimed invention, but merely illustrates some embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention. It should be noted that, in the absence of conflict, the embodiments and features and technical solutions in the embodiments of the present invention can be combined with each other. It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0024] Example 1: A multi-objective land use optimization method based on spatial data, comprising the following steps: Step 1: Collect and construct a dataset of economic and ecological benefit functions and constraint conditions in northern Anhui. Step 1.1: Collect statistical yearbook data for cities in northern Anhui; Obtain statistical yearbook data from the statistical bureaus of various cities in northern Anhui Province between 2010 and 2020, and add the data for the corresponding years to obtain the overall statistical data for northern Anhui Province. Step 1.2: Preliminary construction of function values ​​under the economic development scenario of northern Anhui region; The economic development scenario sets maximizing economic benefits as its core objective, emphasizing increasing the proportion of high-economic-value land use types. The core of this scenario lies in fully releasing the economic potential of land resources through optimal allocation, thereby enhancing the region's economic output capacity. The corresponding objective function can be expressed as: In the formula: f ( X) This indicates the economic value corresponding to land use in northern Anhui. X i For the first i Area of ​​land use type; i =1,2,......,6 correspond to six land use types: cultivated land, forest land, grassland, water area, construction land, and unused land; No For the first i Economic output coefficients for different land use types. This study uses the economic value created per unit area by agriculture, forestry, animal husbandry, fisheries, and the secondary and tertiary industries as the economic value coefficients for arable land, forest land, grassland, water areas, and construction land. The economic value coefficient for unused land is 0. Therefore, statistical yearbook data is consulted, and the corresponding economic benefit coefficients for each year are calculated. E i ; Step 1.3: Preliminary construction of function values ​​under the ecological protection scenario in northern Anhui region; The ecological protection scenario aims to maximize ecological benefits, with its core objective focusing on enhancing ecosystem service functions to promote carbon storage growth and improve the ecological environment. Under this scenario, priority is given to improving the regional ecosystem service value through rational optimization of land use structure, thereby maximizing the ecosystem service value of Henan Province. The corresponding objective function can be expressed as follows: In the formula, h ( X This represents the total value of ecosystem services for land use types in northern Anhui. X i For the first i The area of ​​land use type, of which i =1,2,......,6 correspond to six land use types: cultivated land, forest land, grassland, water area, construction land, and unused land; G i Then it is the first iEcological value coefficients for different land use types. The study adopted the ecosystem service value equivalent table proposed by Xie Gaodi et al. Considering the economic development level of northern Anhui, one standard unit of ecosystem service value equivalent was defined as 1 / 7 of the economic value of the average annual grain yield of one hectare of farmland. The study used the average price, average yield, and average planted area of ​​cash crops such as rice, wheat, and corn in northern Anhui from 2010 to 2020 as basic data. The calculation formula is as follows: In the formula, E a This indicates the economic value of the food production services provided by a unit area of ​​farmland ecosystem. i As a type of crop; p i for i Average price of crops (yuan / ton); q i for i Yield per unit area of ​​crops (tons / hmm 2 ); m i for i Planting area of ​​crops ( hmm 2 ); M For all crop planting areas ( hmm 2 Therefore, a preliminary ecological benefit index for northern Anhui Province from 2010 to 2020 was constructed. Step 1.4: Data collection of constraints for simulating future spatial states; The latest administrative boundary map and elevation DEM map of northern Anhui were obtained from Bigmap GIS; data on primary and secondary roads, population density, annual precipitation, and aerosol density were obtained from the resource and environmental science data platform. Step 2: Calculation of surface subsidence data in northern Anhui region; Step 2.1, Impact of surface subsidence values; Land subsidence signifies far more than just a loss of ground elevation; it represents a chain reaction leading from fundamental degradation of land functions to structural failure of spatial planning. Firstly, it directly causes the decline of land's core function as a production and carrying capacity factor: in agriculture, this manifests as reduced productivity due to waterlogging and salinization; in construction, it manifests as decreased safety and asset value due to infrastructure damage. This functional degradation is essentially a continuous depletion of land's natural capital and economic value. Furthermore, by permanently altering the fundamental geographical parameter of surface elevation, subsidence undermines the physical basis of land use control, creating a systemic risk of discrepancies between existing ecological protection red lines, flood control and drainage systems, and urban and rural layouts. Ultimately, this process clearly reveals that human land use intensity has exceeded the critical point of geological carrying capacity and natural recovery, becoming a key negative indicator and an irreversible risk warning for measuring the sustainability of regional development models and the effectiveness of land use governance systems. Step 2.2, Surface settlement calculation; The ultimate goal of the entire InSAR processing is to decompose and calculate the raw phase difference observed by radar into surface deformation. This is based on a fundamental physical relationship: In the formula The total interferometric phase is the phase difference between corresponding pixels in two SAR images, which is the sum of all contributions. , It's radar phase, the satellite is... t 1 and t 2 The phase of the radar echo signal received from the ground unified target; The wavelength of the electromagnetic waves emitted by satellite radar sensors; This refers to the displacement of the ground unmarked along the satellite's line of sight during the two imaging sessions; The phase component that contributes to the actual deformation of the Earth's surface; Phase contribution caused by landmark undulations; The delay caused by the atmospheric effect on the speed of radar signal propagation; This refers to the random phase caused by factors such as sensor noise, temporal decoherence, and spatial decoherence. After data processing to eliminate the phase caused by terrain, atmosphere, and noise, a clean deformation phase is obtained. Therefore, the deformation along the line of sight is calculated. Then convert the viewing deformation into vertical deformation. Calculate the surface subsidence value; Step 3: Calculation of the grey prediction model (GM(1,1)) and the future economic benefit index and ecological benefit index; The grey prediction model GM(1,1) was designed to address the uncertainty of "small sample size and limited information." Its core lies in weakening the randomness of the original sequence through processing such as accumulating limited data. This characteristic allows for effective short-term dynamic simulation and trend prediction of system behavior with only a small amount of data, providing a new quantitative prediction method for land use change research. This method was used to process economic and environmental indices from 2010 to 2020, predicting the economic value coefficients and revised ecological value coefficients for each land use type. Land use data collection and multi-scenario setup; Step 3.1: Land use data collection; The land use data used in this study originated from the China Land Cover Dataset (CLCD) released by a research team at Wuhan University. This dataset has a spatial resolution of 30 meters. Based on the national standard "Classification of Current Land Use" (GB / T 21010-2017), we reclassified the original data into six categories: cultivated land, forest land, grassland, water area, construction land, and unused land. Finally, we extracted and constructed a land use classification dataset covering northern Anhui Province. Step 3.2, Setting the Scene; By setting differentiated scenario frameworks to simulate and predict regional development paths, the core of this study lies in using multiple scenarios to systematically extrapolate future development trends. This helps provide a scientific basis for formulating policies related to economic development and ecological protection, and optimizes the decision-making process. This study sets up eight scenario structures: The scenario includes eight scenarios: a natural development scenario (land use trends are mainly extrapolated from historical data without additional economic or policy intervention), an economic development scenario (maximizing economic benefits and maximizing the economic potential of land resources), an ecological protection scenario (prioritizing the enhancement of ecosystem service functions to maximize ecological benefits), and a comprehensive development scenario (balancing economic and ecological benefits by coordinating the proportions of different land use types to seek a balance between economic development and ecological protection). The scenario comprehensively considers land subsidence data, resulting in a total of eight scenario structures. Step 4: Multi-objective optimization model; Step 4.1: Construct a multi-objective optimization model; Multi-objective optimization models provide a mathematical framework for solving optimization problems with multiple conflicting objectives, expressed as follows: In the formula, F ( x () represents the objective function, which includes One objective function, For decision variables, including nThere are 1 variable. The constraint is defined as follows: In the formula and Let these represent equality constraints and inequality constraints, respectively. k , J These are the number of equality constraints and the number of inequality constraints, respectively. Step 4.2: Selection of a non-dominated sorting genetic algorithm; The algorithm (Non-dominated Sorting Genetic Algorithm II, NSGA-II) is based on fast non-dominated sorting, crowding distance calculation and elite retention strategy. It has the characteristics of high computational efficiency and strong convergence. Its specific steps are as follows: (1) Randomly generate the initial population. ,in It is a vector of decision variables that needs to satisfy initial conditions and constraints, and then the objective function value of each solution is calculated. (2) Define the population into multiple non-dominant levels based on the dominance relationship. And the condition for a solution to dominate is: and The sorting results must satisfy: F 1 The solution contains no other Pareto front solutions. F 2 :quilt F 1 Dominate but not dominate F 1 The set of solutions, and so on; let the solution... X i The number of dominant solutions is Its dominating set is (3) To maintain population diversity, crowding distance is calculated at each level. The formula is In the formula and In the target upper solution X i The objective value of the adjacent solutions, and The goal The maximum and minimum values; (4) Using the binary tournament algorithm, the solution with the lower non-dominated level is selected first. If the levels are the same, the crowding distance is compared, that is, if Then choose X i ;like p = q Then choose the solution with the larger crowding distance, i.e. Time Selection Xi (5) For crossover and mutation operations, simulated binary crossover (SBX) and polynomial mutation are used to generate offspring populations. When performing crossover operations, it is assumed that there are two parent solutions. and To generate child solutions Y 1 , Y 2 : ; in In the formula It is a cross-distribution index Related random numbers, It is a random number belonging to the interval [0,1]; in the mutation operation, for , after mutation Calculated as , It can be expressed by the following formula: In the formula It is the distribution index of variation. It is a random number within the interval [0,1]; (6) Merge parent populations and offspring population Generate a joint population ,right R t Perform non-dominated sorting and crowding distance calculation, and select the top... N Each solution constitutes the next generation population. The joint population selection formula is: If the last level Exceed N Select examples in descending order; if the maximum number of iterations is reached... If the population objective function converges, the algorithm terminates and returns all non-dominated solutions in the population as the final Pareto front solution set.

[0025] The study uses GM(1,1) to predict the economic and ecological value coefficients of land use types in 2030, and constructs the NSGA-II optimization model based on this. Step 5: Spatial Pattern Simulation and Evolution PLUS Model; Step 5.1, Land Expansion Analysis Strategy; The LEAS model, as one of the core modules of the PLUS model, extracts and samples expansion areas of various land use types from the land use data of the study area, and then applies a random forest algorithm to quantify the development probability of different land use types. The core of this module is the random forest algorithm, whose formula is: In the formula Indicates the first i Each unit has different land use types k The probability of growth; d The value is 0 or 1, with a value of 1 indicating that other land use types have been changed to a new land use type. k A value of 0 indicates other transformations; x A vector consisting of multiple driving factors; I ( ) is the indicator function of the decision tree; h ( x ) is the first n The prediction types of each decision tree; M The total number of decision trees Step 5.2, CA model for multiple types of random patch seeds; CARS (CA model based on multiple random patch seeds) PLUS model uses a threshold-decreasing degree-type random patch seed generation mechanism to simulate the spatial evolution of land use types. It generates dynamically changing "seeds" on a land use type development probability surface output by the LEAS module, using the following formula: In the formula, Indicates the first i Land use type of each cell k The probability of growth, Future land use types k The impact of demand, Represents a cell i Neighborhood effect r A random number between 0 and 1. Land use type k The threshold for the generation of new land patches. If a new land use type... c If a player wins a round of competition, a decreasing threshold is applied. The evaluation is based on the following formula: , In the formula and yes t-1 generation and first t During iteration, land use type c The difference between the current quantity and future demand. Decreasing threshold The attenuation factor ranges from 0 to 1. rl The values ​​are normally distributed random values ​​with a mean of 1. l It is a decay series. It defines whether land use types are permitted. k Convert to c The transformation matrix.

[0026] This study simultaneously defines constraints and objective functions, and combines calculated surface subsidence data to establish multiple development scenarios, optimizing both the quantity and spatial aspects of land use types. First, a land use quantity optimization scheme is established using the NSGA-II model. Then, the PLUS model is used to predict the development probability and expansion impact of various land use types under different development scenarios, establishing a land use spatial optimization scheme. Finally, the land use status of northern Anhui Province under multiple scenarios is simulated.

[0027] The above embodiments are only used to illustrate the present invention and are not intended to limit the technical solutions described herein. Although the present invention has been described in detail with reference to the above embodiments, the present invention is not limited to the specific embodiments described above. Therefore, any modifications or equivalent substitutions to the present invention, as well as all technical solutions and improvements that do not depart from the spirit and scope of the invention, are covered within the scope of the claims of the present invention.

Claims

1. A multi-objective land use optimization method based on spatial data, characterized in that, Includes the following steps: Step 1: Construct a dataset of economic and ecological benefit functions and constraints for the northern Anhui region; Step 2: Calculate surface subsidence data for the northern Anhui region; Step 3: Use the grey prediction model (GM(1,1)) to calculate the future economic benefit index and ecological benefit index, and collect land use data and set up multiple scenarios; Step 4: Construct a multi-objective optimization model and solve the model using the Non-Dominated Sorting Genetic Algorithm (NSGA-II); Step 5: Simulate and evolve the spatial pattern based on the PLUS model.

2. The multi-objective land use optimization method based on spatial data according to claim 1, characterized in that, Step 1 specifically includes: Step 1.1: Collect statistical yearbook data for cities in northern Anhui; Obtain statistical yearbook data from the statistical bureaus of various cities in northern Anhui, and add up the data for the corresponding years to obtain the overall statistical data for northern Anhui. Step 1.2: Preliminary construction of function values ​​under the economic development scenario of northern Anhui region; The economic development scenario sets maximizing economic benefits as its core orientation, emphasizing increasing the proportion of high-economic-value land use types. The core of this scenario lies in fully releasing the economic potential of land resources through optimal allocation, thereby enhancing the region's economic output capacity. The corresponding objective function can be expressed as: In the formula: f ( X) This indicates the economic value corresponding to land use in northern Anhui. X i For the first i Area of ​​land use type; i =1,2,......,6 correspond to six land use types: cultivated land, forest land, grassland, water area, construction land, and unused land; E i For the first i The economic output coefficient of land use type; the economic value created per unit area by agriculture, forestry, animal husbandry, fishery, and the secondary and tertiary industries is used as the economic value coefficient for cultivated land, forest land, grassland, water area, and construction land, while the economic value coefficient for unused land is 0; therefore, statistical yearbook data is consulted and the corresponding economic benefit coefficient for each year is calculated. E i .

3. The multi-objective land use optimization method based on spatial data according to claim 2, characterized in that, Step 1.3: Preliminary construction of function values ​​under the ecological protection scenario in northern Anhui region; The ecological protection scenario aims to maximize ecological benefits, with its core objective focusing on enhancing ecosystem service functions to promote carbon storage growth and improve the ecological environment. Under this scenario, priority is given to improving the regional ecosystem service value through rational optimization of land use structure, thereby maximizing the ecosystem service value in northern Anhui Province. The corresponding objective function can be expressed as follows: In the formula, h ( X This represents the total value of ecosystem services for land use types in northern Anhui. X i For the first i The area of ​​land use type, of which i =1,2,......,6 correspond to six land use types: cultivated land, forest land, grassland, water area, construction land, and unused land; G i Then it is the first i Ecological value coefficients for different land use types were used in this study. The ecosystem service value equivalent table proposed by Xie Gaodi et al. was adopted. Considering the economic development level of northern Anhui, the ecosystem service value equivalent of one standard unit was defined as 1 / 7 of the economic value of the average annual grain yield of one hectare of farmland. The average price, average yield, and average planting area of ​​cash crops such as rice, wheat, and corn in northern Anhui were used as basic data. The calculation formula is as follows: In the formula, E a This indicates the economic value of the food production services provided by a unit area of ​​farmland ecosystem. i As a type of crop; p i for i Average price of crops (yuan / ton); q i for i Yield per unit area of ​​crops (tons / hm 2 ); m i for i Planting area of ​​crops ( hm 2 ); M For all crop planting areas ( hm 2 Therefore, an ecological benefit index for the northern Anhui region was initially constructed. Step 1.4: Data collection of constraints for simulating future spatial states; The latest administrative boundary map and elevation DEM map of the northern Anhui region were obtained from Bigmap GIS; data on primary and secondary roads, population density, annual precipitation, and aerosol density were obtained from the resource and environmental science data platform.

4. The multi-objective land use optimization method based on spatial data according to claim 3, characterized in that, Step 2 specifically includes: Step 2.1, Impact of surface subsidence values; Land subsidence signifies far more than just a loss of ground elevation; it represents a chain reaction leading from fundamental degradation of land functions to structural failure of spatial planning. Firstly, it directly causes the decline of land's core function as a production and carrying capacity factor: in agriculture, this manifests as reduced productivity due to waterlogging and salinization of arable land; in construction, it manifests as depreciation in safety levels and asset value due to infrastructure damage. This functional degradation is essentially a continuous depletion of land's natural capital and economic value. Furthermore, by permanently altering the fundamental geographical parameter of surface elevation, subsidence undermines the physical basis of land use control, causing systemic risks of discrepancies between maps and reality to emerge from ecological protection red lines, flood control and drainage systems, and urban and rural layouts delineated based on the original topography. Ultimately, it clearly reflects that human land use intensity has exceeded the critical point of geological carrying capacity and natural recovery, becoming a key negative indicator and an irreversible risk warning for measuring the sustainability of regional development models and the effectiveness of land use governance systems.

5. The multi-objective land use optimization method based on spatial data according to claim 4, characterized in that, Step 2.2, Surface settlement calculation; The ultimate goal of the entire InSAR processing is to decompose and calculate the raw phase difference observed by radar into the amount of surface deformation. It is based on a fundamental physical relationship: In the formula The total interference phase is for two scenes. The phase difference between corresponding pixels in an image is the sum of all contributions; , It's radar phase, the satellite is... t 1 and t 2 The phase of the radar echo signal received from the ground unified target; The wavelength of the electromagnetic waves emitted by satellite radar sensors; This refers to the displacement of the ground target along the satellite's line of sight during the two imaging sessions; The phase component contributing to the actual surface deformation; Phase contribution caused by surface undulations; The delay caused by the atmospheric effect on the speed of radar signal propagation; The random phase is caused by factors such as sensor noise, temporal decoherence, and spatial decoherence; after eliminating the terrain, atmospheric, and noise phase through data processing, a clean deformation phase is obtained. Therefore, the deformation along the line of sight is calculated. Then convert the viewing deformation into vertical deformation. The surface subsidence value was calculated.

6. The multi-objective land use optimization method based on spatial data according to claim 5, characterized in that, Step 3 specifically includes: Step 3.1: Land use data collection; The land use data used in this study comes from the China Land Cover Dataset (CLCD) released by a research team from Wuhan University. This dataset has a spatial resolution of 30 meters. Based on the national standard "Classification of Current Land Use" (GB / T 21010-2017), we reclassified the original data into six types: cultivated land, forest land, grassland, water area, construction land, and unused land. Finally, we extracted and constructed a land use classification dataset covering northern Anhui Province. Step 3.2, Setting the Scene; By setting up differentiated scenario frameworks to simulate and predict regional development paths, the core of this study lies in using multiple scenarios to systematically extrapolate future development trends. This helps provide a scientific basis for formulating policies related to economic development and ecological protection, and optimizes the decision-making process. This study sets up eight scenario structures: The natural development scenario, where land use trends are mainly extrapolated from historical data without additional economic or policy intervention; the economic development scenario, which aims to maximize economic benefits and maximize the economic potential of land resources; the ecological protection scenario, which prioritizes enhancing ecosystem service functions to maximize ecological benefits; and the comprehensive development scenario, which balances economic and ecological benefits by coordinating the proportions of different land use types to seek a balance between economic development and ecological protection. Additionally, eight scenario structures are set up, including the natural development scenario, economic development scenario, ecological protection scenario, and comprehensive development scenario, all taking into account land subsidence data.

7. The multi-objective land use optimization method based on spatial data according to claim 6, characterized in that, Step 4 specifically includes: Step 4.1: Construct a multi-objective optimization model; Multi-objective optimization models provide a mathematical framework for solving optimization problems with multiple conflicting objectives, expressed as follows: In the formula, F ( x () represents the objective function, which includes m One objective function, For decision variables, including n There are 1 variable; the constraint condition is defined as: In the formula and Let these represent equality constraints and inequality constraints, respectively. k , J These are the number of equality constraints and the number of inequality constraints, respectively.

8. A multi-objective land use optimization method based on spatial data according to claim 7, characterized in that, Step 4.2: Selection of a non-dominated sorting genetic algorithm; The algorithm (Non-dominated Sorting Genetic Algorithm II, NSGA-II) is based on fast non-dominated sorting, crowding distance calculation and elite retention strategy. It has the characteristics of high computational efficiency and strong convergence. Its specific steps are as follows: (1) Randomly generate the initial population. ,in It is a vector of decision variables that needs to satisfy initial conditions and constraints, and then the objective function value of each solution is calculated. (2) Define the population into multiple non-dominant levels based on the dominance relationship. And the condition for a solution to dominate a solution is: and The sorting results must satisfy: F 1 The solution contains no other Pareto front solutions. F 2 :quilt F 1 Dominate but not dominate F 1 The set of solutions, and so on; let the solution... X i The number of dominant solutions is Its dominating set is (3) To maintain population diversity, crowding distance is calculated at each level. The formula is In the formula and In the target upper solution X i The objective value of the adjacent solutions, and The goal The maximum and minimum values; (4) Using binary tournament selection, priority is given to solutions with lower non-dominated levels. If the levels are the same, the crowding distance is compared, i.e., if Then choose X i ;like p = q Then choose the solution with the larger crowding distance, i.e. Time Selection X i (5) For crossover and mutation operations, simulated binary crossover is used. And polynomial mutation generates offspring population When performing crossover operations, it is assumed that there are two parent solutions. and To generate child solutions Y 1 , Y 2 : in In the formula It is a cross-distribution index Related random numbers, It is a random number belonging to the interval [0,1]; in the mutation operation, for , after mutation Calculated as , It can be expressed by the following formula: In the formula It is the distribution index of variation. It is a random number within the interval [0,1]; (6) Merge parent populations and offspring population Generate a joint population ,right R t Perform non-dominated sorting and crowding distance calculation, and select the top... N Each solution constitutes the next generation population. The joint population selection formula is: If the last level Exceed N Select examples in descending order; If the maximum number of iterations is reached If the population objective function converges, the algorithm terminates and returns all non-dominated solutions in the population as the final Pareto front solution set. The economic and ecological value coefficients of land use types in 2030 are predicted by GM(1,1), and an NSGA-II optimization model is constructed based on this.

9. A multi-objective land use optimization method based on spatial data according to claim 8, characterized in that, Step 5 specifically includes: Step 5.1, Land Expansion Analysis Strategy; The LEAS model, as one of the core modules of the PLUS model, extracts and samples the expansion areas of various land use types from the land use data of the study area, and then applies the random forest algorithm to quantify the development probability of different land use types. The core of this module is the random forest algorithm, whose formula is: ， In the formula Indicates the first i Each unit has different land use types k The probability of growth; d The value is 0 or 1. A value of 1 indicates that other land use types have been converted to land use type k, and a value of 0 indicates other conversions. x A vector consisting of multiple driving factors; I ( ) is the indicator function of the decision tree; h ( x ) is the first n The prediction types of each decision tree; M This represents the total number of decision trees.

10. A multi-objective land use optimization method based on spatial data according to claim 9, characterized in that, Step 5.2, CA model for multiple types of random patch seeds; CARS (CA Model Based on Multiple Random Patch Seeds): The PLUS model uses a threshold-decreasing degree-type random patch seed generation mechanism to simulate the spatial evolution of land use types. It generates dynamically changing "seeds" on a land use type development probability surface output by the LEAS module, using the following formula: In the formula, Indicates the first i Land use type of each cell k The probability of growth, Future land use types k The impact of demand, Represents a cell i Neighborhood effect r A random number between 0 and 1. The threshold for generating new land patches for land use type k; if a new land use type c If a player wins a round of competition, a decreasing threshold is applied. The evaluation is based on the following formula: , In the formula and It is the land use type during the t-1 generation and the t-th generation iteration. c The difference between the current quantity and future demand. Decreasing threshold The attenuation factor ranges from 0 to 1. rl The values ​​are normally distributed random values ​​with a mean of 1. l It is a decay series. It defines whether land use types are permitted. k Convert to c The transformation matrix is ​​determined. By simultaneously defining constraints and objective functions, and combining the calculated surface subsidence data, various development scenarios are established to optimize both the quantity and spatial aspects of land use types. First, a land use quantity optimization scheme is established using the NSGA-II model. Then, the PLUS model is used to predict the development probability and expansion impact of various land use types under different future development scenarios, establishing a land use spatial optimization scheme. Finally, the land use status of northern Anhui region under various scenarios is simulated.

Images