CN113642764B - Village and town aggregation space evolution simulation prediction method and computer equipment - Google Patents

Abstract

The embodiment of the application discloses a village and town aggregation space evolution simulation prediction method and equipment, comprising the following steps: a dependent variable layer and an independent variable layer required by the GeoDetector model are manufactured; making an exclusion layer and other layers required by SLEETH model simulation; inputting a plurality of layers into a SLEETH model for parameter correction to obtain optimal parameters; setting the initial year and the final year of SLEETH model simulation, and adopting optimal parameters to simulate and predict the village and town aggregation space evolution of a research area. According to the embodiment of the application, the GeoDetector model and the SLEETH model are fused to simulate and predict the evolution of the village and town aggregation space, and the driving relation of different transformation driving forces to the village and town aggregation space is inlaid into the land utilization simulation model, so that the advantage complementation of the two models is realized, the precision of the simulation model can be effectively improved, and the evolution trend of the village and town aggregation space can be better depicted.

Description

A method and computer equipment for simulating and predicting the spatial evolution of settlements in villages and towns

technical field

The invention relates to the technical fields of human geography and urban-rural planning, in particular to a method for simulating and predicting the spatial evolution of village and town settlements.

Background technique

Land use is an important indicator reflecting the spatial changes of villages and towns. It is affected by various transformational dynamics such as economy, society, population, and policies. It has always been the focus of research in the fields of human geography and urban-rural planning. The dynamic change simulation of land use can analyze and predict the change process of village and town space, predict the expansion direction and scale of village and town construction land, and help land managers formulate effective village and town development policies, and the land based on cellular automata (CA) model Leveraging change simulation and forecasting is a hot topic among them. Among them, one of the key issues and difficulties is how to identify the driving force of the spatial transformation of villages and towns, and how to effectively integrate the driving force with the land use simulation model, so as to improve the accuracy of the simulation model and make it better predict the evolution trend of villages and towns space .

However, with the existing technology, the existing cellular automaton (CA) model still cannot well embedding the spatial transformation dynamics of village and town settlements and the simulation of land use dynamic changes in space, resulting in the inability to effectively predict different The dynamic change process of spatial expansion of construction land in villages and towns during the period.

Contents of the invention

In view of the above-mentioned technical defects, the purpose of the embodiments of the present invention is to provide a method for simulating and predicting the spatial evolution of village and town settlements and computer equipment, so as to effectively integrate the recognition model of the spatial transformation dynamics of village and town settlements with the simulation model of land use dynamic change, and realize the two models Complementary advantages of each other, and finally scientifically describe the development trend and spatial layout of the future land use of villages and towns.

In order to achieve the above purpose, in the first aspect, the embodiment of the present invention provides a method for simulating and predicting the spatial evolution of village and town settlements, including:

S1: Obtain the historical and current remote sensing images of the preset research area, and make the dependent variable layer required by the GeoDetector model according to the land use interpretation results of the remote sensing images;

S2: Collect the driving factor data within the scope of the preset research area, and make the independent variable layer required by the GeoDetector model according to the driving factor data;

S3: use the GIS platform to make sampling points, extract the data in the dependent variable layer and the independent variable layer according to the sampling points, and export the sampling results in excel format;

S4: Import the sampling results into the GeoDetector model for calculation, and obtain the driving strength of each driving factor on the spatial evolution of village and town settlements;

S5: making an exclusion layer required for SLEUTH model simulation according to the driving strength;

S6: Make the slope layer, land use layer, city scale layer, traffic layer and hillshade layer required for SLEUTH model simulation;

S7: Input multiple layers of the SLEUTH model into the SLEUTH model for parameter correction, and obtain the optimal parameters for the SLEUTH model simulation;

S8: Set the start year and end year of the SLEUTH model simulation, and use the optimal parameters to simulate and predict the spatial evolution of village and town settlements in the study area.

As a specific implementation manner of the present application, step S1 specifically includes:

S11: Obtain historical and current remote sensing images of the preset research area in multiple periods;

S12: Interpreting the remote sensing images in each period to obtain the land use status vector data of villages and towns in multiple time periods;

S13: According to the vector data of the current land use status, use the GIS platform to make the dependent variable layer required by the GeoDetector model.

As a specific implementation manner of the present application, step S5 specifically includes:

S51: According to the driving strength of each driving factor, use the GIS platform to weight and superimpose the layers of each driving factor within the research range to generate a dynamic distribution map of the spatial evolution of village and town settlements;

S52: Obtain various layers of vegetation, waters, terrain, and ecological control elements within the research area, and use the GIS platform to weight and superimpose the above layers to generate a spatial ecological sensitivity distribution map of villages and towns;

S53: Using the GIS platform to generate a probability map of the spatial evolution of village and town settlements from the village and town settlement spatial evolution dynamic distribution map and the village and town settlement spatial ecological sensitivity distribution map;

S54: Reclassify the data, and make the spatial evolution probability map of village and town settlements as an exclusion layer required for SLEUTH model simulation.

As a specific implementation manner of the present application, step S6 specifically includes:

S61: Obtain the DEM data within the scope of the study area and make it into a slope layer and a hillshade layer;

S62: Making the land use status vector data in multiple periods into a land use layer;

S63: Make the vector data of village and town settlement construction land in multiple periods into a city-wide layer;

S64: Make the vector data depicting the current road traffic as a traffic layer.

As a specific implementation manner of the present application, step S7 specifically includes:

The forced Monte Carlo iterative calculation method is used for parameter correction. The parameter correction is divided into four stages: rough correction, fine correction, final correction and simulation parameter acquisition. A set of growing parameter sets obtained in each step is used for the next step. Step parameter calibration, finally get the optimal parameters of SLEUTH model simulation.

Further, as a preferred embodiment of the present application, the method further includes:

Use the Kappa coefficient to evaluate the consistency of the simulation results. When the simulation results are completely consistent with the reference, Kappa reaches the maximum value of 1. The larger the Kappa value, the better the consistency; Small: When 0.4≤Kappa<0.75, the consistency between the two is average, and the change is more obvious. When Kappa<0.4, the consistency between the two is low, and the change is large.

In a second aspect, an embodiment of the present invention provides a computer device, including a processor, an input device, an output device, and a memory, and the processor, input device, output device, and memory are connected to each other, wherein the memory is used to store A computer program, the computer program includes program instructions, and the processor is configured to invoke the program instructions to execute the method in the first aspect above.

Implement the embodiment of the present invention, use the method of fusion of GeoDetector model and SLEUTH model to simulate and predict the evolution of village and town settlement space, and embed the driving relationship of different transformation driving forces on village and town settlement space into the land use simulation model, realizing two The complementary advantages of the models can effectively improve the accuracy of the simulation model and enable it to better describe the evolution trend of the settlement space of villages and towns.

Description of drawings

In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings that need to be used in the description of the specific embodiments or the prior art.

Fig. 1 is a flow chart of the method for simulating and predicting the spatial evolution of village and town settlements provided by the embodiment of the present invention;

Figure 2 is a land use classification map of Hancheng City;

Fig. 3 is a schematic diagram of GeoDetector model dependent variable (Y) layer;

Fig. 4 is a schematic diagram of GeoDetector model independent variable (X) layer;

Fig. 5 is a schematic diagram of a sampling point layer;

Figure 6 is a schematic diagram of the exclusion layer of the SLEUTH model;

Figure 7 is all input layers of the SLEUTH model;

Fig. 8 is a display diagram of the simulation results of the SLEUTH model;

Fig. 9 is a structural diagram of a computer device provided by an embodiment of the present invention.

Detailed ways

The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some of the embodiments of the present invention, but not all of them. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

Please refer to Fig. 1, the embodiment of the present invention provides a method for simulating and predicting the spatial evolution of village and town settlements that integrates GeoDetector and SLEUTH models, which mainly includes the following steps:

S1: Obtain remote sensing images of the history and current status of the preset research area in multiple periods.

S2: Interpret the remote sensing images in each period to obtain the vector data of land use status in the settlement space of villages and towns in multiple time periods.

S3: Collect the driving factor data in the study area, perform preliminary processing, and form a data set.

S4: Use the GIS platform to make the multi-period land use status vector data obtained in step S2 into spatial growth change data in different time periods, and generate the dependent variable (Y) layer required by the GeoDetector model.

S5: Making the driving factor data obtained in step S3 as an independent variable (X) layer required by the GeoDetector model.

S6: Use the GIS platform to make sampling points, extract the data in the dependent variable layer and independent variable layer according to the sampling points, and export the sampling results in excel format.

S7: Import the table data obtained in step S6 into the GeoDetector model for calculation, and obtain the driving strength of each driving factor on the spatial evolution of village and town settlements.

S8: According to the driving strength of each driving factor obtained in step S7, use the GIS platform to weight and superimpose the layers of each driving factor within the research range to generate a dynamic distribution map of the spatial evolution of village and town settlements.

S9: Obtain various layers of vegetation, waters, terrain, and ecological control elements within the research area, and use the GIS platform to weight and superimpose the above layers to generate a spatial ecological sensitivity distribution map of villages and towns.

S10: Use the GIS platform to generate a probability map of the spatial evolution of village settlements from the maps obtained in steps S8 and S9.

S11: Reclassify the data, and make the spatial evolution probability map of village and town settlements obtained in step S10 as the exclusion layer required for SLEUTH model simulation.

S12: Other data layers required for making SLEUTH model simulations: Slope, Landuse, Urban, Transportation, Hillshade.

S13: Input multiple layers of the SLEUTH model into the SLEUTH model, and use the Brute-force Monte Carlo method to correct the parameters. The parameter correction is divided into rough correction, fine correction, final correction and simulation parameters Acquisition is carried out in four stages, and a set of growing parameter sets obtained in each step is used for parameter calibration in the next step, and finally the optimal parameters for SLEUTH model simulation are obtained.

S14: Use the optimal parameters obtained by the final calibration to carry out simulation verification of land use evolution in the study area, use the layer data of the first period as the initial year of simulation, simulate and generate the land use map of the current year, and compare it with The actual status quo land use in the current year is compared and analyzed to quantitatively evaluate the accuracy of the model simulation.

S15: Use the Kappa coefficient to evaluate the consistency of the simulation results. When the result is completely consistent with the reference, Kappa reaches the maximum value of 1. The larger the Kappa value, the better the consistency; when Kappa≥0.75, the consistency between the two is high. The change is small: when 0.4≤Kappa<0.75, the consistency between the two is average, and the change is more obvious. When Kappa<0.4, the consistency between the two is low and the change is large.

S16: Use the model after the consistency evaluation to simulate the optimal parameters, and use the layer data of the current year as the initial year to simulate and predict the spatial evolution of village and town settlements in the study area.

S17: Obtain the simulation and prediction results of the spatial evolution of village and town settlements in the target year.

Further, in order to better understand the embodiments of the present invention, the following examples illustrate:

(1) Through the interpretation of 30-meter-accurate satellite images, the vector data of land use classification in four periods (2000, 2005, 2013, and 2018) of Hancheng City in Weinan, Shaanxi Province were obtained, as shown in Figure 2.

(2) Extract the map spots used for village and town settlement construction in the most recent period (2013 to 2018), calculate the spatial variation with the town as a unit, and use the GIS platform to make the dependent variable (Y) layer required by GeoDetector, as shown in Figure 3 Show.

(3) Select the independent variable factor (X) required by GeoDetector, that is, the dynamic factor, such as the total population (X1), population density (X2), urban distance (X3), road distance (X4), etc., and make each factor as For the corresponding layers, use the natural breakpoint method to process each layer into 5 classifications, and reclassify as 1, 2, 3, 4, 5, as shown in Figure 4.

(4) Use the fishing net module of the sampling function in the GIS data management tool to make a layer of sampling points with a spacing of 300 meters (10 times the accuracy, which can be adjusted according to the size of the area), and use the sampling module extracted and analyzed in the spatial analysis tool to export to excel Format sampling results, as shown in Figure 5.

(5) Import the sampling results into GeoDetector for calculation, and obtain the q values of each dynamic factor of the factor detector. The q value is also called factor explanatory power. Its value is between 0 and 1, indicating how much a dynamic factor (X) explains the spatial differentiation of the dependent variable (Y). The larger the q value, the more significant the independent variable X is. The stronger the explanatory power of the dependent variable Y, and vice versa, the weaker, the expression is:

In the formula: h=1, ..., L is the stratification of variable Y or factor X, i.e. classification or partition; N _h and N are respectively the number of units in layer h and the whole district; and ^σ2 are the variance of the Y values of layer h and the whole area, respectively. SSW and SST are the sum of variance within a layer and the total variance of the whole region, respectively.

The q values of GeoDetector calculation results X1, X2, X3, and X4 are: , and the ratio after normalization processing is , which can be used as the weight of each dynamic factor.

Among them, the GeoDetector operation results are shown in Table 1:

(6) Use the weighted sum module of the overlay analysis in the GIS spatial analysis tool to assign weights to each dynamic factor layer, and then export it to a 30-meter-precision 8-bit grayscale GIF file format, which can be used as a SLEUTH model to exclude layers (Figure 6 ). The value of the pixel attribute of the exclusion layer is 0-100, indicating the impossibility that the area can be converted into settlement construction land. The higher the value of the exclusion layer, the less likely the area will be converted into settlement construction land. For example, if the value is 100, it means that the area is completely impossible to be reconstructed into settlement construction land. On the contrary, the value is 0, which means that the area can be converted The most likely land for settlement construction. In the 8bit grayscale GIF format, the color stretching range is 0-255, which corresponds to the 0-100 value assigned to the layer pixel attribute.

(7) SLEUTH model simulation requires 6 layers, namely Slope, Landuse, Exclusion, Urban, Transportation, and Hillshade. Obtain the 30-meter precision DEM data of Hancheng City, make it into a slope layer and a hillshade layer, make the vector data of land use status at 4 time nodes into a land use layer, and make the vector data of village and town settlement construction land at 4 time points into a city Range layer, which depicts the current road traffic vector data at four time nodes, is made as a traffic layer, and all layers are exported as an 8-bit grayscale GIF file format with uniform coordinates and consistent resolution, as shown in Figure 7.

(8) Input each layer into the SLEUTH model for model calibration. In the calibration stage, OSM_NS is used as the best fit index to determine the model.

OSM_NS=compare×pop×edges×clusters×xmean×ymean.

In the formula, compare is the ratio of the total number of urbanization pixels in the simulated last year to the actual total number of urbanization pixels in the last year, and pop is the least squares method for the ratio of the number of simulated urbanization pixels to the actual number of urbanization pixels in the calibration year Regression correlation coefficient value, edges is the least squares regression correlation coefficient value of the ratio of the simulated urban boundary number to the real urban boundary number in the calibration year, and clusters is the least squares regression correlation of the simulated urban cluster and the real urban cluster ratio in the calibration year Coefficient value, xmean is the least square regression correlation coefficient value of the ratio of the average x coordinate value of the simulated urbanization pixel to the average x coordinate value of the real urbanization pixel in the calibration year, and ymean is the average value of the simulated urbanization pixel The least square regression correlation coefficient value of the ratio between the y coordinate value and the average y coordinate value of the real urbanization pixel in the calibration year.

(9) After three calibration stages of rough calibration, fine calibration and rough calibration, a set of optimal parameters are obtained: the diffusion coefficient is 31, the reproduction coefficient is 65, the dispersion coefficient is 1, the slope coefficient is 62, and the road coefficient is 44.

(10) Input the optimal parameters, set the start year of the SLEUTH model simulation as 2018, and the end year as 2030, and then the simulation prediction results of the spatial evolution of village and town settlements in Hancheng City from 2018 to 2030 can be obtained, as shown in Figure 8.

Implement the above-mentioned method for simulating and predicting the evolution of village and town settlement space of the present invention, adopt the method of fusion of GeoDetector model and SLEUTH model to simulate and predict the evolution of village and town settlement space, and embed the driving relationship of different transformation driving forces on village and town settlement space into the land use simulation In the model, the accuracy of the simulation model can be effectively improved, so that it can better describe the evolution trend of the settlement space of villages and towns.

Based on the same inventive concept, an embodiment of the present invention provides a computer device. As shown in FIG. 9, the user mobile terminal may include: one or more processors 101, one or more input devices 102, one or more The output device 103 and the memory 104 , the processor 101 , the input device 102 , the output device 103 and the memory 104 are connected to each other through the bus 105 . The memory 104 is used to store a computer program, the computer program includes program instructions, and the processor 101 is configured to call the program instructions to execute the methods in the foregoing method embodiments.

It should be understood that in the embodiment of the present invention, the so-called processor 101 may be a central processing unit (Central Processing Unit, CPU), and the deep learning graphics card (such as: Huawei NPU, Nvidia GPU, Google TPU) The processor may also be other general-purpose Processor, digital signal processor (Digital Signal Processor, DSP), application specific integrated circuit (Application Specific Integrated Circuit, ASIC), off-the-shelf programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate Or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor, or the processor may be any conventional processor, or the like.

The input device 102 may include a keyboard, etc., and the output device 103 may include a display (LCD, etc.), a speaker, and the like.

The memory 104 may include read-only memory and random-access memory, and provides instructions and data to the processor 101 . A portion of memory 104 may also include non-volatile random access memory. For example, memory 104 may also store device type information.

In specific implementation, the processor 101, input device 102, and output device 103 described in the embodiment of the present invention can execute the implementation described in the embodiment of the method for simulating and predicting the spatial evolution of village and town settlements provided by the embodiment of the present invention, which is not described here. Let me repeat.

It should be noted that, for a more specific work flow and related details of the computer device in the embodiment of the present invention, please refer to the foregoing method embodiments, and details are not repeated here.

The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person familiar with the technical field can easily think of various equivalents within the technical scope disclosed in the present invention. Modifications or replacements shall all fall within the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims (2)

1. A village and town aggregation space evolution simulation prediction method is characterized by comprising the following steps:

s1: acquiring historical and current remote sensing images of a preset research area range, and manufacturing a dependent variable layer required by a GeoDetector model according to land utilization interpretation results of the remote sensing images;

s2: collecting driving factor data in a preset research area range, and manufacturing an independent variable layer required by a GeoDetector model according to the driving factor data;

s3: manufacturing sampling points by using a GIS platform, sampling and extracting the dependent variable layer and the data in the independent variable layer according to the sampling points, and exporting a sampling result in an excel format;

s4: the sampling result is imported into a GeoDetector model for calculation, and driving strength of each driving factor on village and town aggregation space evolution is obtained;

s5: manufacturing an exclusion layer required by SLEETH model simulation according to the driving strength;

s6: manufacturing a gradient layer, a land utilization layer, a city range layer, a traffic layer and a mountain shadow layer required by SLEETH model simulation;

s7: inputting a plurality of layers of the SLUTH model into the SLUTH model for parameter correction to obtain optimal parameters simulated by the SLUTH model;

s8: setting the initial year and the final year of SLEETH model simulation, and adopting the optimal parameters to simulate and predict the village and town aggregation space evolution of a research area;

the step S1 specifically comprises the following steps:

s11: acquiring historic and current remote sensing images of a preset research area range in a plurality of periods;

s12: interpreting the remote sensing image of each period to obtain land utilization status vector data of villages and towns in a plurality of time periods;

s13: according to the land utilization current situation vector data, utilizing a GIS platform to manufacture a dependent variable layer required by a GeoDetector model;

the step S5 specifically comprises the following steps:

s51: according to the driving strength of each driving factor, weighting and superposing the layers of each driving factor in the research range by adopting a GIS platform to generate a village and town aggregation space evolution power distribution diagram;

s52: acquiring various vegetation, water areas, terrains and ecological management and control element layers in a research range, and carrying out weighted superposition on the layers by adopting a GIS platform to generate a village and town aggregate space ecological sensitivity distribution map;

s53: generating a village and town aggregation space evolution probability map by adopting a GIS platform to obtain the village and town aggregation space evolution power distribution map and a village and town aggregation space ecological sensitivity distribution map;

s54: reclassifying the data, and manufacturing the village and town aggregation space evolution probability map as an exclusion map layer required by SLEEUTH model simulation;

the step S6 specifically comprises the following steps:

s61: obtaining DEM data in the range of a research area, and manufacturing the DEM data into a gradient layer and a mountain shadow layer;

s62: preparing land use current situation vector data in a plurality of periods as a land use layer;

s63: the method comprises the steps of manufacturing urban area range map layers by using village and town aggregation construction land vector data in a plurality of periods;

s64: the current road traffic vector data is drawn to be a traffic layer;

the step S7 specifically comprises the following steps:

carrying out parameter correction by adopting a forced Monte Carlo iterative calculation method, wherein the parameter correction is carried out in 4 stages of coarse correction, fine correction, final correction and parameter acquisition, and a set of increased parameter sets obtained in each step are used for parameter correction of the next step to finally obtain optimal parameters simulated by a SLUTH model;

consistency evaluation is carried out on the simulation results by using Kappa coefficients, when the simulation results are completely consistent with the reference, kappa reaches a maximum value of 1, and the larger the Kappa value is, the better the consistency is; when Kappa is more than or equal to 0.75, the consistency of the two is higher, and the change is smaller: when Kappa is more than or equal to 0.4 and less than or equal to 0.75, the consistency of the two is common, and when Kappa is more obvious and less than 0.4, the consistency of the two is lower, and the change is larger.

2. A computer device comprising a processor, an input device, an output device, and a memory, the processor, the input device, the output device, and the memory being interconnected, wherein the memory is configured to store a computer program comprising program instructions, the processor being configured to invoke the program instructions to perform the method of claim 1.