CN-121998260-A - Multi-target composite ecological network construction and local synergy method, device and equipment

CN121998260ACN 121998260 ACN121998260 ACN 121998260ACN-121998260-A

Abstract

The invention relates to the field of urban planning and management, in particular to a multi-target composite ecological network construction and local synergy method, device and equipment, creatively provides a composite ecological network integrating the double targets of urban cold island effect and habitat quality, and is based on a hierarchical optimization method of macroscopic network pattern constraint, machine learning interpretation and space threshold regulation, in a heterogeneous composite ecological network, the complex nonlinear relation between the landscape index and the double ecological targets is captured by utilizing the high-dimensional nonlinear fitting capability of the machine learning model, and the SHAP model is introduced to identify the dominant driving factor of the synergy and the quantization threshold value thereof in the composite ecological network, so that the abstract model prediction result is inverted into a planning control index with definite physical meaning, a targeted local synergy optimization strategy is generated, and the ecological restoration efficiency and the refined management and control effect are greatly improved.

Inventors

XIANG YANG
HE QINGJUN
HUANG CHUNBO
WANG XIAOSHUANG
ZHANG YANGYANG

Assignees

中国地质大学（武汉）

Dates

Publication Date: 20260508
Application Date: 20260408

Claims (10)

1. The multi-target composite ecological network construction method is characterized by comprising the following steps of: step 1, multi-source remote sensing data of a target city are obtained and preprocessed, wherein the multi-source remote sensing data comprise earth surface temperature data and geospatial data; Step 2, constructing a threat source parameter table and a habitat type sensitivity table based on the geospatial data, calculating habitat quality by adopting a InVEST model, and generating a habitat quality index distribution diagram; Step 3, respectively carrying out reverse and forward normalization and space superposition on the surface temperature data and the habitat quality index distribution map, and carrying out morphological space pattern analysis to extract a composite ecological source land which simultaneously meets the dual attribute of the biological habitat and the urban cold source; and 4, calculating a comprehensive resistance surface based on the composite ecological source land to identify a composite ecological corridor and a composite obstacle point, and constructing a multi-target composite ecological network.
2. The multi-objective complex ecological network construction method according to claim 1, wherein the geospatial data includes at least one or more of land utilization/land coverage data, digital elevation model, local climate zone, night light data, population data and road data.
3. The multi-objective complex ecological network construction method according to claim 1, wherein the surface temperature data is acquired and preprocessed, comprising the steps of: step 101, screening an initial remote sensing image meeting preset conditions in Google EARTH ENGINE; 102, performing mask processing on a low-quality area in the initial remote sensing image by using a quality evaluation wave band, and synthesizing single-frame surface temperature data through a median; And 103, performing primary filling on the single-frame ground surface temperature data by adopting a preset interpolation strategy, and selecting an optimal semi-variance model to perform secondary filling by means of average error and/or standardized root mean square error to generate seamless target ground surface temperature data.
4. The multi-objective complex ecological network construction method according to claim 1, wherein the threat source parameter table contains a plurality of types of key threat factors, and influence parameter values for representing stress intensities of different ecological types are set for each type of key threat factors, wherein the key threat factors comprise watertight surfaces, roads, cultivated lands, bare lands, night lights, population distributions and railways, and the influence parameter values comprise maximum influence distances, relative weights and space attenuation function forms; the habitat type sensitivity table includes suitability scores for different habitats as habitats and their sensitivity to the key threat factors.
5. The multi-objective complex ecological network constructing method according to any one of claims 1 to 4, wherein the identifying the complex ecological source place comprises the steps of: step 301, reversely normalizing the surface temperature data by adopting an inverse sequence form of a range normalization method, positively normalizing the habitat quality data, carrying out weighted superposition on the normalized cold island intensity grid graph and the habitat quality grid graph, and calculating to generate a composite ecological index; Step 302, reclassifying the composite ecological index by using a quantile classification method, and screening out a high-value area as a potential composite ecological source; Step 303, extracting core plaques of the potential composite ecological source land through morphological space pattern analysis, and filtering finely-broken interference plaques in the core plaques through an area screening mechanism; and step 304, extracting preferred core plaques with high connectivity according to the possible connectivity index and/or the overall connectivity index, and taking the preferred core plaques as a target composite ecological source.
6. The multi-objective complex ecological network constructing method according to claim 5, wherein the identifying complex ecological corridor and complex obstacle point comprises the steps of: Step 401, constructing a first resistance factor affecting the cold island effect and a second resistance factor affecting the habitat quality, acquiring an interpretation force q value of each resistance factor on the surface temperature and the habitat quality space difference characteristic by adopting a factor detector in a geographic detector, normalizing the q value, and then taking the normalized q value as the weight of each resistance factor, and weighting and superposing to generate a comprehensive resistance surface; Step 402, using the target composite ecological source land as a source point and the comprehensive resistance surface as a consumption substrate, and calculating a lowest resistance path based on a minimum accumulated resistance model to identify the composite ecological corridor; And step 403, performing global obstacle point scanning on the composite ecological corridor and the buffer area thereof based on a preset obstacle point detection model and a Barrier Mapper tool, and taking the obstacle point with the highest obstacle value level as the composite obstacle point.
7. A local synergy method based on the multi-objective composite ecological network constructed by the method of any one of claims 1-6, characterized by comprising the following steps: Step 5, identifying dominant driving factors and corresponding threshold intervals of the multi-target composite ecological network based on an interpretable machine learning model; And 6, making a space optimization strategy for a key area of the multi-target composite ecological network according to the dominant driving factor and the threshold interval thereof so as to realize the collaborative improvement of the cold island effect and the habitat quality.
8. The local synergy method of claim 7, wherein identifying the dominant driving factor and the corresponding threshold interval in step 5 comprises the steps of: Step 501, constructing a plurality of interpretable machine learning models and training, wherein the input of the interpretable machine learning models is the landscape index of the multi-target composite ecological network, and the output of the interpretable machine learning models is the surface temperature and the habitat quality; step 502, performing grid search and super-parameter tuning on the plurality of interpretable machine learning models, and selecting a machine learning model with highest precision as an optimal prediction model based on a preset evaluation index; step 503, building a landscape index response curve under different granularities, and obtaining an optimal grid resolution when a preset numerical value stability condition is met; Step 504, calculating SHAP values of all landscape indexes in the optimal prediction model under the optimal grid resolution by adopting a SHAP model, namely, feature contribution degree; step 505, summarizing the SHAP absolute value average values of all samples to show the global importance ordering and local interpretation of each landscape index on the surface temperature and the habitat quality, and obtaining the dominant driving factor; step 506, establishing a SHAP local dependency graph of the dominant driving factor, and generating a key threshold of the dominant driving factor based on the SHAP local dependency graph analysis, wherein the key threshold is an inflection point of the SHAP value from negative to positive or from positive to negative.
9. A local synergy device is characterized by comprising a data processing module, a habitat calculation module, a network construction module, a threshold identification module and a strategy generation module, The data processing module is used for acquiring and preprocessing multi-source remote sensing data of a target city, wherein the multi-source remote sensing data comprises earth surface temperature data and geographic space data; the habitat calculation module is used for constructing a threat source parameter table and a habitat type sensitivity table based on the geographic space data, calculating habitat quality by adopting a InVEST model and generating a habitat quality index distribution diagram; The network construction module is used for respectively carrying out reverse normalization and forward normalization and space superposition on the surface temperature data and the habitat quality index distribution map, carrying out morphological space pattern analysis, extracting a composite ecological source land which simultaneously meets the double attributes of the biological habitat and the urban cold source, and calculating a comprehensive resistance surface based on the composite ecological source land to identify a composite ecological corridor and a composite obstacle point, so as to construct a multi-target composite ecological network; The threshold value identification module is used for identifying dominant driving factors and corresponding threshold value intervals of the multi-target composite ecological network based on an interpretable machine learning model; The strategy generation module is used for making a space optimization strategy for a key area of the multi-target composite ecological network according to the dominant driving factor and the threshold interval thereof so as to realize the collaborative promotion of the cold island effect and the habitat quality.
10. A local synergy apparatus comprising a computer readable storage medium and a processor, wherein said processor, when executing a computer program on said computer readable storage medium, implements the steps of the local synergy method of claim 7 or 8.

Description

Multi-target composite ecological network construction and local synergy method, device and equipment Technical Field The invention relates to the field of urban planning and management, in particular to a method, a device and equipment for constructing a multi-target composite ecological network and locally enhancing effect. Background At present, related researches for relieving urban heat islands are focused on cooling effect analysis of single landscape elements (such as water bodies and greenbelts) or constructing a cold island network based on a plaque-gallery-matrix theory. However, the existing method still has obvious limitations that firstly, the construction of the traditional cold island network depends on the physical index of surface temperature, comprehensive consideration of population dimension is lacking, thermal hazard and thermal exposure and thermal vulnerability are not organically combined, so that a release strategy possibly generates ' resource mismatch ' - ' namely, a cooling measure is excessively configured in a low population thermal exposure area, but an actual high risk area with higher temperature and high population density is ignored, secondly, the existing driving mechanism analysis assumes that urban thermal environment driving factors are in a homogeneous and linear relationship in space, the heterogeneous influence of different spatial functions is ignored, and the essential difference of dominant driving factors and thresholds in surfaces serving as thermal risks, galleries serving as thermal transmission channels and barrier points serving as thermal barriers cannot be effectively distinguished. Therefore, a comprehensive evaluation framework capable of integrating thermal hazard, thermal exposure and thermal vulnerability is urgently needed, a high/low hot air risk network is constructed based on the framework, and meanwhile, a technical method for accurately analyzing nonlinear morphological thresholds under different network functions is provided, so that the accuracy and efficiency of urban hot air risk regulation are supported. Disclosure of Invention The invention provides a method, a device and equipment for constructing and locally enhancing a multi-target composite ecological network, which solve the technical problems. The first aspect of the embodiment of the invention provides a multi-target composite ecological network construction method, which comprises the following steps: step 1, multi-source remote sensing data of a target city are obtained and preprocessed, wherein the multi-source remote sensing data comprise earth surface temperature data and geospatial data; Step 2, constructing a threat source parameter table and a habitat type sensitivity table based on the geospatial data, calculating habitat quality by adopting a InVEST model, and generating a habitat quality index distribution diagram; Step 3, respectively carrying out reverse and forward normalization and space superposition on the surface temperature data and the habitat quality index distribution map, and extracting a potential composite ecological source land which simultaneously meets the dual attribute of the biological habitat and the urban cold source; And 4, performing morphological space pattern analysis on the potential composite ecological source land, identifying a target composite ecological source land, calculating a comprehensive resistance surface to identify a composite ecological corridor and a composite obstacle point, and constructing a multi-target composite ecological network. The second aspect of the embodiment of the invention provides a local synergy method based on the multi-target composite ecological network, which comprises the following steps: Step 5, identifying dominant driving factors and corresponding threshold intervals of the multi-target composite ecological network based on an interpretable machine learning model; And 6, making a space optimization strategy for a key area of the multi-target composite ecological network according to the dominant driving factor and the threshold interval thereof so as to realize the collaborative improvement of the cold island effect and the habitat quality. A third aspect of the embodiments of the present invention provides a local synergy apparatus comprising a computer readable storage medium and a processor implementing the steps of the local synergy method described above when executing a computer program on the computer readable storage medium. A fourth aspect of the embodiments of the present invention provides a local synergy device, comprising a data processing module, a habitat computing module, a network construction module, a threshold identification module and a policy generation module, The data processing module is used for acquiring and preprocessing multi-source remote sensing data of a target city, wherein the multi-source remote sensing data comprises earth surface temperature data and geographic space