CN-121998325-A - Multi-target land use optimization method based on spatial data

CN121998325ACN 121998325 ACN121998325 ACN 121998325ACN-121998325-A

Abstract

The invention provides a multi-objective land utilization optimization method based on space data, which integrates land subsidence data into an optimization model as key constraint variables, and the method utilizes sentinel SAR data to quantify land subsidence by integrating land utilization, social economy and ecological data, so as to construct a multi-objective function which simultaneously pursues maximization of economic benefit and ecological benefit; and then adopting an improved non-dominant genetic algorithm (NSGA-II) to carry out multi-objective optimization, and using a PLUS model to realize the space refined distribution of the optimized land utilization structure. The method overcomes the defect that the prior method ignores the dynamic change of the geological environment, and enables the optimization result to truly reflect the earth surface subsidence risk, thereby providing scientific decision basis for land utilization planning taking safety, development and protection into consideration in rapid development and geological sensitive areas.

Inventors

TAO TINGYE
CHENG WEI
HUANG JIANWEI
Wang Zidai
CHEN WENJIE
CHEN TAO
WU YONGBO

Assignees

合肥工业大学

Dates

Publication Date: 20260508
Application Date: 20260119

Claims (10)

1. The multi-target land use optimization method based on the spatial data is characterized by comprising the following steps of: step 1, constructing economic benefit and ecological benefit function data and constraint condition data sets of the north Anhui region; Step 2, calculating earth surface subsidence data of the north Anhui region; step3, calculating a future economic benefit index and an ecological benefit index by adopting a gray prediction model (GM (1, 1)), collecting land utilization data and setting multiple scenes; step 4, constructing a multi-objective optimization model, and solving the model by adopting a non-dominant order genetic algorithm (NSGA-II); And 5, performing space pattern simulation evolution based on the PLUS model.
2. The multi-objective land use optimizing method based on spatial data as set forth in claim 1, wherein step 1 specifically includes: Step 1.1, collecting urban statistics annual-image data of the north Anhui region; acquiring statistical annual-image data in each city statistical bureau in the north of Anhui area, and adding the corresponding annual-image data to obtain the global north of Anhui area statistical data; step 1.2, preliminarily constructing a function value under the economic development situation of the north Anhui region; the economic development scenario sets the maximization of economic benefit as core guide to emphasize the improvement of the occupation ratio of land utilization types with high economic value, the core of the scenario is to fully release the economic potential of land resources through the optimal configuration of the land resources so as to enhance the economic output capacity of the area, and the corresponding objective function can be expressed as follows: Wherein f (X) represents the economic value corresponding to land utilization in the north Anhui region, X i represents the area of the type i land utilization, i=1, 2, and the number of the first to sixth land utilization types, 6 represents the economic value coefficient of the type i land utilization, E i represents the economic value coefficient of the type i land utilization by using the economic value coefficient of the unit area agriculture, forestry, animal husbandry, fishery and the second and third industries as the economic value coefficient of the cultivated land, the woodland, the grassland, the water area and the construction land, and the economic value coefficient of the unused land is 0, thus, inquiring statistical annual-image data and calculating the corresponding economic benefit coefficient E i each year.
3. The multi-objective land use optimizing method based on spatial data as claimed in claim 2, wherein step 1.3, preliminarily constructing function values under the environment of ecology protection of the north Anhui area; The ecological protection scenario aims at realizing the maximization of ecological benefit, the core aim is to promote the service function of the ecological system so as to promote the increase of carbon reserves and improve the ecological environment, in the scenario, the priority is to improve the service value of the regional ecological system by reasonably optimizing the land utilization structure so as to enable the service value of the ecological system in the north of Anhui province to reach the highest level, and the corresponding objective function can be expressed as follows: Wherein h (X) represents the total value of the ecosystem service of the land use type in the north Anhui area, X i is the area of the land use type of the i type, wherein i=1, 2, the number of the land use types of the i type is 6, the cultivated land, the woodland, the grassland, the water area, the construction land and the unused land are corresponding to six land use types, G i is the ecological value coefficient of the land use type of the i type, an ecosystem service value equivalent table proposed by Xie Gaode and the like is adopted in the research, 1 standard unit ecosystem service value equivalent is defined as 1/7 of the economic value of the annual average grain yield of 1 hectare in the north Anhui area in combination with the economic development level of the north Anhui area, and the average price, average yield and average planting area of economic crops such as rice, wheat and corn in the north Anhui area are counted as follows: Wherein E a represents the economic value of providing food production service function by the farmland ecosystem in unit area, i is the crop type, p i is the average price (yuan/ton) of i crops, q i is the yield in unit area of i crops (ton/hm 2 ）;m i is the planting area (hm 2 ) of i crops, M is the planting area (hm 2 ) of all crops, and therefore, the ecological benefit index of the northern Anhui area is initially constructed; step 1.4, collecting future space state simulation constraint condition data; Acquiring the latest administrative boundary map and elevation DEM map of the north Anhui region from Bigmap GIS, and acquiring the data of primary roads, secondary roads, population density, annual precipitation, aerosol density and the like from a resource environment science data platform.
4. A multi-objective land use optimizing method based on spatial data as claimed in claim 3, wherein step 2 specifically comprises: step 2.1, the influence of the earth surface sedimentation value; The method is characterized in that the ground surface subsidence is marked by far-non-simple ground elevation loss, but a field is a chain reaction from ground function basic degradation to space planning structural failure, the core function degradation of the ground serving as production and bearing elements is firstly directly caused, the productivity degradation caused by farmland waterlogging and salinization is represented in the agricultural field, the safety level and asset value detraction caused by infrastructure damage is represented in the building field, the degradation of the function is essentially continuous consumption of natural capital and economic value of the ground, furthermore, the physical foundation of national soil space use control is mobilized by the permanent change of the ground surface elevation, so that the systematic risk that the original land is not consistent with the ecological protection red line, flood control and drainage system and urban and rural layout map defined according to the original topography is realized, and finally, the key index and irreversible early warning about whether the development mode of the area is sustainable and the national soil space use system is effective or not are measured and the critical point of reverse recovery is clearly reflected.
5. The multi-objective land use optimizing method based on space data according to claim 4, wherein step 2.2, calculation of surface subsidence; The final target of the whole InSAR processing is to decompose and settle the original phase difference observed by the radar into the surface deformation, and the final target is based on a basic physical relationship: In the middle of Is the total interference phase and is two views The phase difference value of the corresponding pixel point of the image is the sum of all contributions; , Is the radar phase, the phase of the radar echo signal received by the satellite from the ground unity target at t 1 and t 2 ; The wavelength of electromagnetic waves emitted by the satellite radar sensor; displacement of the ground target along the satellite line of sight during the two imaging; The phase part contributing to the true deformation of the earth surface Phase contribution due to surface relief; Delay caused by the atmosphere to the radar signal propagation speed; The method is used for eliminating the terrain, atmosphere and noise phases through data processing to obtain pure deformation phases Thereby calculating the distortion of the visual line Then the visual direction deformation is converted into vertical deformation And calculating the surface subsidence value.
6. The multi-objective land use optimizing method based on spatial data as set forth in claim 5, wherein step 3 specifically includes: step 3.1, land utilization data collection; Land utilization data adopted by the research are derived from a Chinese land coverage data set (CLCD) issued by a research team of the university of Wuhan, wherein the spatial resolution of the data set is 30 meters, and according to the national standard of land utilization status Classification (GB/T21010-2017), the original data are reclassified into six types, namely cultivated land, woodland, grassland, water area, construction land and unused land; Step 3.2, setting a situation; The modeling prediction of the regional development path is carried out by setting a differentiated scene framework, and the modeling prediction is characterized in that the systematic deduction of the future development situation is realized by setting multiple scenes, so that the modeling prediction is beneficial to providing scientific basis and optimizing decision-making process when preparing policies related to economic development and ecological protection, wherein eight scene structures are set in the study: The land utilization trend is deduced mainly according to historical data, economic or policy intervention is not added additionally, economic development is achieved, economic potential of land resources is exerted to the greatest extent by taking economic benefit maximization as a main target, ecological protection is achieved by improving service functions of an ecological system preferentially, ecological benefit maximization is achieved, comprehensive development is achieved, economic benefit and ecological benefit are achieved, balance between economic development and ecological protection is sought by coordinating proportion among different land utilization types, and meanwhile, eight scene structures are set and comprehensively considered.
7. The multi-objective land use optimizing method based on spatial data as set forth in claim 6, wherein step 4 specifically includes: step 4.1, constructing a multi-objective optimization model; The multi-objective optimization model can be used for solving the mathematical framework of the optimization problem of a plurality of mutually conflicting objectives, and expressed in the form of: wherein F (x) represents an objective function, including m objective functions, For decision variables, n variables are included, and the constraint is defined as: In the following And The equality constraint and the inequality constraint are respectively expressed, and k and J are the equality constraint number and the inequality constraint number respectively.
8. The method for optimizing multi-objective land use based on spatial data as claimed in claim 7, wherein step 4.2, selection of non-dominant ranking genetic algorithm; The algorithm (Non-dominated Sorting Genetic Algorithm II, NSGA-II) uses the rapid Non-dominant sorting, crowding distance calculation and elite retention strategy as the core, and has the characteristics of high calculation efficiency, strong convergence and the like, and the specific steps are that (1) an initial population is randomly generated Wherein Is a vector of decision variables, and needs to satisfy initial conditions and constraint values so as to calculate an objective function value of each solution (2) Defining the population into a plurality of non-dominant classes according to the dominant relationship And the conditions for solving the dominant solution are And is also provided with The sorting result needs to satisfy that the solution of F 1 has no other set of pareto front solutions, F 2 is the set of solutions which are dominated by F 1 but not dominated by F 1 , so on, and the dominated solution number of the solution X i is set as Which is governed by (3) Calculating the crowding distance in each class for maintaining population diversity The formula is In the following And Is at the target The target value of the adjacent solution of the upper solution X i , And Is the object of With binary tournament selection, preferably selecting solutions with low non-dominant ranking, if the ranking is the same, then comparing crowding distances, i.e. if X i is selected, and if p=q, a solution with a larger crowding distance is selected, i.e Time-select X i (5) for crossover and mutation operations, use is made of analog binary crossover And polynomial variation to generate offspring populations Two parent solutions are assumed in the crossover operation And To generate a child solution Y 1 ,Y 2 : Wherein the method comprises the steps of In the middle of Is a cross distribution index The random number of the code in question, Is a random number within interval 0,1, and in the variation operation, for After mutation Calculated as , Is represented by the following formula: In the middle of In order to obtain the mutation distribution index, Is a random number in interval 0,1, and (6) merging parent population And offspring populations Generating a combined population Performing non-dominant sorting and crowding distance calculation on R t , and selecting the first N solutions to form a next generation population The joint population selection formula is If the last stage Exceeding N, selected in descending order by way of example; If the maximum iteration number is reached Or the objective function of the population converges, the algorithm is terminated, and all non-dominant solutions in the population are returned to be used as a final pareto front solution set; the economic value coefficient and the ecological value coefficient of the land use type in 2030 year are obtained through GM (1, 1) prediction, and an NSGA-II optimization model is built based on the economic value coefficient and the ecological value coefficient.
9. The multi-objective land use optimizing method based on spatial data as set forth in claim 8, wherein step 5 specifically includes: Step 5.1, land dilation analysis strategy; The LEAS model is used as one of core modules of the PLUS model, and is used for extracting and sampling expansion areas of various land types in land utilization data of a research area and simultaneously giving random forest algorithms to quantify development probabilities of different land utilization types, wherein the core of the module is the random forest algorithm, and the formula is as follows: , In the middle of Indicating the growth probability of each land utilization type k in the ith unit, d is 0 or 1, 1 indicates that other land utilization types are converted into the land utilization type k, 0 indicates other conversion, x indicates a vector composed of a plurality of driving factors, I # ) Is an indication function of decision trees, h (x) is the prediction type of each decision tree, and M is the total number of decision trees.
10. The multi-objective land use optimizing method based on spatial data as claimed in claim 9, wherein step 5.2, CA model of a plurality of kinds of random plaque seeds; CARS (CA model based on multiple types of random plaque seeds) is characterized in that the PLUS model adopts a random plaque seed generation mechanism based on a degree type with decreasing threshold value to simulate the space evolution process of the land utilization type, and dynamically-changed seeds are generated on one land utilization type development probability surface output by the LEAS module, wherein the formula is as follows: in the formula, Representing the increasing probability of the i-th cell land utilization type k, The impact of future land use type k demand, Representing the neighborhood effect of cell i, r is a random number from 0 to 1, Generating a threshold value of new land patches for land use type k, and if a new land use type c wins in a round of competition, taking a decrementing threshold value To evaluate, the calculation formula is: , In the following And Is the difference between the current number of land utilization types c and the future demand at the t-1 generation and the t-th generation of iteration, To decrease the threshold value In the range of 0-1, rl is a normal distributed random value and the mean value is 1, l is the number of decay stages, Defining whether to allow the land utilization type k to be converted into a conversion matrix of c, and simultaneously setting up a plurality of different development scenes by combining the calculated surface subsidence data and simultaneously fixing constraint conditions and an objective function, so as to optimize the land utilization type in two aspects of quantity and space. Firstly, establishing a land use quantity optimization scheme through an NSGA-II model, then respectively predicting various land use type development probabilities and expansion influences under various future development situations by using a PLUS model, establishing a land use space optimization scheme, and finally simulating land use conditions of a north Anhui region under various situations.

Description

Multi-target land use optimization method based on spatial data Technical Field The invention relates to the field of land use optimization and ecological system service value evaluation, in particular to a multi-target land use optimization method based on spatial data. Background The reasonable utilization of land resources is directly related to the sustainability of the ecological environment and the healthy development of economy. With the acceleration of the urban process, land utilization problems are increasingly developed, so that ecological system services are threatened, and a series of problems such as resource exhaustion, environmental deterioration and the like are caused. The services offered by the ecosystem, including climate control, water resource supply, soil protection, and biodiversity maintenance, are critical and of irreplaceable economic and environmental value from a global perspective. Numerous studies have shown that the economic value of an ecosystem service can be quantified not only by its direct products, but also by a unit area value equivalent factor for systematic evaluation. The assessment method can provide clear economic indexes for services of different ecological system types, and further provides quantitative support for land utilization decisions. However, traditional land use decisions focus on maximization of economic benefit, often neglecting sustainability and long-term benefits of the ecological environment, which can lead to impaired health of the ecological system, reducing its potential sustainable development. If in some resource-depleted cities, excessive development and neglecting of ecological protection measures eventually lead to a sharp drop of municipal conditions, affecting the overall social benefit. In this context, multi-objective optimization methods are increasingly gaining importance. The method can comprehensively consider the factors of multiple layers such as economy, society, ecology and the like, thereby realizing the economic growth and simultaneously protecting the ecology environment to the greatest extent. However, the consideration of the important influence factor of the surface subsidence in the prior study is particularly lacking. The invention introduces the earth surface subsidence data into the multi-objective land utilization optimizing process for the first time as important variable conditions to enhance the scientificity and the practical applicability of the model. By integrating the ground surface subsidence data, the influence of land utilization decisions can be comprehensively evaluated, effective support based on data analysis is provided for policy makers, and the boosting achieves the win-win situation of economic development and ecological protection. Therefore, a multi-objective land use optimization method based on spatial data is proposed. Disclosure of Invention The invention provides a multi-target land utilization optimization method based on spatial data, which is characterized in that ground surface subsidence data is integrated into an optimization model as key constraint variables for the first time. The method constructs a multi-objective function which simultaneously pursues maximization of economic benefit and ecological benefit by integrating land utilization, social economy and ecological data and quantifying earth surface subsidence by utilizing the sentinel SAR data, then adopts an improved non-dominant genetic algorithm (NSGA-II) to carry out multi-objective optimization, and utilizes a PLUS model to realize space refined distribution of the optimized land utilization structure. The method overcomes the defect that the prior method ignores the dynamic change of the geological environment, and enables the optimization result to truly reflect the earth surface subsidence risk, thereby providing scientific decision basis for land utilization planning taking safety, development and protection into consideration in rapid development and geological sensitive areas. The invention provides a multi-target land use optimizing method based on space data, which comprises the following steps: step 1, constructing economic benefit and ecological benefit function data and constraint condition data sets of the north Anhui region; Step 2, calculating earth surface subsidence data of the north Anhui region; step3, calculating a future economic benefit index and an ecological benefit index by adopting a gray prediction model (GM (1, 1)), collecting land utilization data and setting multiple scenes; step 4, constructing a multi-objective optimization model, and solving the model by adopting a non-dominant order genetic algorithm (NSGA-II); And 5, performing space pattern simulation evolution based on the PLUS model. As a preferable technical scheme of the invention, the step 1 specifically comprises the following steps: Step 1.1, collecting urban statistics annual-image data of the north Anhui region; acquiring stati