Comprehensive city toughness regulation and control method based on multi-source data mining and threshold effect

Abstract

The invention discloses a comprehensive city toughness regulation and control method based on multi-source data mining and a threshold effect, and the method comprises the steps: completing city toughness evaluation through collected city toughness influence factor data, and carrying out the analysis to obtain a weak region where the city toughness is distributed and developed in an unbalanced manner; the method for measuring and calculating the key influence factors of the city toughness is constructed by analyzing internal characteristics of basic regulation and control factors of the city, and threshold points of the key influence factors are analyzed by utilizing an established threshold effect model of the key influence factors of the city toughness. The urban toughness level distribution and the threshold effect of the key influence factors of the urban toughness are combined to realize comprehensive and practical urban toughness regulation and control.

Description

Technical Field

[0001] The present invention belongs to the technical field of resilient cities, and specifically relates to a comprehensive urban resilience regulation method based on multi-source data mining and threshold effects. Background Art

[0002] With the rapid development of urban economies, cities are expanding and urbanization levels are significantly increasing. While this process has improved residents' living standards, it has also triggered a series of urban safety issues. For example, the continued growth of urban populations worldwide has directly exacerbated waste generation and greenhouse gas emissions, in turn impacting global climate change and triggering a variety of natural disasters, bringing economic and social pressures around the world. Traffic congestion, resource shortages, and environmental pollution have further exacerbated the complexity of urban development, posing even more severe challenges to urban governance. Against this backdrop, traditional passive urban safety models are no longer sufficient to meet the needs of increasingly complex urban systems. To optimize urban development structures and achieve sustainable urbanization, we need to proactively build resilient cities to enhance their ability to cope with diverse risks, reduce losses from accidents and disasters, and maintain normal urban operations.

[0003] Traditional urban resilience research has focused primarily on guiding improvements to urban resilience. However, these efforts suffer from overly generalized resilience measures, a high degree of subjectivity in identifying key factors influencing urban resilience, and a lack of direct application to urban development practices. In recent years, resilient city development has become a new trend in urban planning and development, with numerous locations around the world formulating resilient city development plans. However, existing methods can only analyze urban resilience levels and propose reference measures to enhance urban resilience. These methods fail to objectively identify key factors influencing urban resilience, leading to a high degree of subjectivity in improving urban resilience. Furthermore, a lack of research on urban resilience regulation makes it difficult to directly apply common resilience measures to urban development practices.

[0004] By evaluating the urban resilience levels of various regions within a given city, we found significant imbalances in urban resilience. By collecting and observing the resilience distribution of other cities in existing research, we also found similar imbalances. These studies (such as CN118115036A) often use a single method to select the most important factors influencing urban resilience and propose guiding measures to enhance urban resilience. Therefore, it is urgent to develop a method that addresses the imbalances in resilience development across urban areas. By analyzing the inherent characteristics of urban resilience elements, we can objectively identify key influencing factors, combine the distribution of urban resilience levels with the threshold effects of key influencing factors, and establish a comprehensive approach to urban resilience regulation. Summary of the Invention

[0005] The main purpose of this invention is to overcome the shortcomings and deficiencies of the existing technology, and to provide a comprehensive urban resilience regulation method based on multi-source data mining and threshold effects. By constructing an urban resilience assessment system, the imbalance of resilience development in urban areas is weighted and assessed. By analyzing the inherent characteristics of urban resilience elements, the key influencing factors are objectively determined. Combined with the distribution of urban resilience levels and the threshold effects of key influencing factors of urban resilience, the practical significance of urban resilience research is enhanced.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] In a first aspect, the present invention provides a comprehensive urban resilience control method based on multi-source data mining and threshold effects, comprising the following steps:

[0008] Constructing an urban resilience assessment system, including a target layer, a criterion layer, and an indicator layer. The criterion layer includes multiple resilience criteria, and the indicator layer includes multiple resilience assessment values. Each resilience criterion corresponds to several resilience assessment values.

[0009] Collect and process raw urban data, conduct resilience assessments of various urban areas based on resilience-influencing factors, obtain weights for each resilience assessment value, calculate the current regional resilience level based on the resilience assessment values ​​and their weights, rank and rank the cities based on the calculated results, and identify areas with low resilience levels. The raw urban data includes time, urban area planning, and specific values ​​of resilience-influencing factors.

[0010] Analyze the inherent characteristics of basic urban regulatory elements in multiple dimensions to obtain the relevant characteristics between each level. Determine the criticality of basic urban regulatory elements based on the relevant characteristics between each level. Calculate and analyze the criticality to obtain the key influencing factors of resilience. The basic urban regulatory elements include multiple levels.

[0011] The threshold effect analysis method is used to analyze the key influencing factors of resilience, obtain the critical values ​​of the key influencing factors of resilience, and implement regulation according to the regional resilience level and the critical values ​​of the key influencing factors of resilience.

[0012] As a preferred technical solution, the criterion layer includes urban facility resilience, economic resilience, community resilience and ecological resilience. Each criterion layer has multiple characterization levels, including absorptive capacity, adaptability and recovery capacity; the resilience assessment value selects the resilience assessment value according to each characterization level.

[0013] As a preferred technical solution, the collection and processing of raw urban data and the assessment of resilience of various urban areas based on resilience influencing factors include:

[0014] Standardize and normalize the original urban data to obtain standard data;

[0015] Based on the standard data corresponding to the resilience influencing factors, the entropy method is used to evaluate the resilience of various urban areas. Specifically:

[0016] S21. Use normalization to make each toughness evaluation value dimensionless, as shown in the following formula:

[0017] Positive indicators:

[0018] Negative indicators:

[0019] Where i represents the i-th resilience subsystem, j represents the j-th resilience index, and R' ij is the corresponding X ij Indicator data; T ij represents the normalized value of the jth index of the i-th resilience subsystem, min(R' ij ) and max(R' ij ) represent R' ij The minimum and maximum values ​​of

[0020] S22. Calculate the information entropy value e of each indicator ij , as follows:

[0021]

[0022] Among them, n represents the number of indicators, q ij represents the proportion of the jth indicator in the i-th resilience subsystem;

[0023] S23. Calculate the weight W of each indicator ij , as follows:

[0024] As a preferred technical solution, the current regional resilience level is calculated based on the resilience assessment value and its weight, specifically: the weighted sum of the resilience assessment values ​​is calculated to obtain the current regional resilience level representation, as shown in the following formula:

[0025]

[0026] Among them, W ij is the weight of each indicator, T ij represents the normalized value of the jth index of the i-th toughness criterion, W ij T ij Used to calculate the resilience evaluation value of the positive indicator, W ij (1-T ij ) is used to calculate the resilience assessment value of negative indicators.

[0027] As an optimal technical solution, the multi-dimensional analysis of the intrinsic characteristics of the basic regulatory elements of the city includes correlation analysis, difference analysis and sensitivity analysis; the basic regulatory elements of the city include resilience indicators, urban subsystems and regional resilience, the regional resilience includes the resilience of multiple urban subsystems, and the resilience evaluation of each urban subsystem has multiple resilience indicators.

[0028] As a preferred technical solution, the correlation analysis includes:

[0029] S31. Correlation analysis between resilience indicators and urban subsystems: The Spearman correlation coefficient was used to measure the correlation between resilience influencing factors within the urban area to be studied, as shown in the following formula:

[0030]

[0031] Among them, d i is the difference in the rank value of the i-th data pair, n is the total number of observation samples, and the correlation coefficient ρ is [-1,1];

[0032] S32. Correlation between urban subsystems: The coupling correlation between the urban subsystems to be studied is calculated using the coupling coordination model, as shown in the following formula:

[0033]

[0034] Where k is the number of urban subsystems, R i is the toughness evaluation value of each criterion layer;

[0035] S33. Correlation between urban subsystems and regional resilience: Using GeoDa software, the local Moran's I index in spatial autocorrelation was used to measure the spatial correlation between the urban subsystems to be studied and the regional resilience.

[0036] As a preferred technical solution, the difference analysis includes: calculating the contribution rate of the resilience index to the corresponding urban subsystem, as shown in the following formula:

[0037]

[0038] Among them, R ij is the jth resilience evaluation value of the i-th criterion layer, R i is the comprehensive evaluation value of each criterion layer.

[0039] As a preferred technical solution, the sensitivity analysis is specifically: performing global sensitivity calculation using a Bootstrap-based RandomForest algorithm, including:

[0040] Define input variables and output variables and obtain the original data set;

[0041] Extract several Bootstrap samples from the original data set with replacement and set the sample size;

[0042] For each Bootstrap sample, a set number of features are randomly selected for node splitting to generate an unpruned decision tree to obtain a random forest model;

[0043] Calculate the mean MSE reduction resulting from splitting nodes in all trees for each input variable;

[0044] Importance scores were calculated by Bootstrap or permutation test, and the importance scores of each item were counted.

[0045] As a preferred technical solution, the method of determining the criticality of basic urban regulatory elements based on the relevant characteristics between various levels includes:

[0046] Through numerical analysis of the correlation characteristics between each level, the criticality (KD) of resilience influencing factors is constructed as follows: KD = CM + DM + SM, where CM is the correlation of basic urban regulatory factors, DM is the difference of basic urban regulatory factors, and SM is the sensitivity of basic urban regulatory factors. Basic urban regulatory factors include resilience indicators, urban subsystems, and regional resilience. Regional resilience includes the resilience of multiple urban subsystems, and the resilience evaluation of each urban subsystem has multiple resilience indicators.

[0047] The numerical analysis of the relevant features between the various levels includes:

[0048] The CM analysis of the correlation between the basic urban control factors is as follows: the correlation between all urban subsystems and resilience indicators is calculated and the average is taken to obtain the mean correlation coefficient ρ xij ; Calculate the coupling degree between each urban subsystem and take the average value to obtain the mean resilience coupling degree Calculating spatial correlations between urban subsystems and regional resilience Establish the correlation degree CM of basic urban control elements as follows:

[0049] When ρ xij >0,

[0050] When ρ xij <0,

[0051] Where i is the resilience subsystem of the i-th city, k is the city subsystem other than the i-th city, j is the resilience index of the j-th city, n is the year, R ij is the evaluation value of each toughness index, R iis the resilience assessment value of each urban subsystem;

[0052] The DM analysis of the differences between the various levels includes: the contribution of each resilience indicator to the resilience of the urban subsystem to which it belongs. xij , analyze the differences in basic urban regulatory factors, combine the evaluation and assignment of various resilience indicators, and establish the difference degree DM of basic urban regulatory factors, as follows:

[0053] For positive indicator factors, DM=R ij I xij ;

[0054] For negative indicator factors, DM=(1-R ij )I xij ;

[0055] The sensitivity SM between each level includes: performing sensitivity analysis on the basic urban regulatory factors through the Random Forest algorithm, calculating the sensitivity of each resilience indicator to the resilience of the urban subsystem to which it belongs, denoted as S xij , and combined with the evaluation and assignment of each factor, the sensitivity SM of the factors affecting urban resilience is established as follows:

[0056] For positive indicator factors, SM=R ij S xij ;

[0057] For negative indicator factors, SM=(1-R ij )S xij .

[0058] As a preferred technical solution, the threshold effect analysis method is used to analyze the key factors affecting urban resilience, including:

[0059] Using the cubic spline interpolation algorithm, based on the resilience assessment values ​​of key influencing factors of urban resilience in different regions arranged in ascending order, we obtained an impact trend map, including a scatter plot and smooth curve of the distribution of key influencing factors;

[0060] Using the Gaussian kernel density algorithm, we established a kernel density estimation function based on the distribution of key influencing factors of urban resilience at different levels of urban resilience, and selected an appropriate bandwidth to obtain a smooth density curve.

[0061] The critical transition point of the influencing mechanism is determined by analyzing the scatter distribution and inflection points of the impact trend diagram, and the significance of the threshold interval is verified by the kurtosis characteristics of the density curve. Combined with the analysis of the critical transition point and threshold interval, the threshold point of the key influencing factors of urban resilience is obtained.

[0062] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0063] (1) This paper divides the complex urban system into four subsystems: urban facilities, economy, community and ecology. Following the principles of scientificity, feasibility and systematicity, 32 indicators are selected from the three characterization levels of resilience, namely, absorptive capacity, adaptability and recovery capacity, as the basic elements of urban resilience regulation, thereby improving the ability to identify weak areas with uneven distribution of urban resilience.

[0064] (2) This invention considers the intrinsic characteristics of the basic regulatory factors of urban resilience, analyzes the connections and interactions between different levels of urban resilience from the three dimensions of correlation, difference, and sensitivity, and simultaneously uses the three dimensions to construct a criticality calculation method for factors affecting urban resilience, which helps to avoid the subjectivity of manual selection and considers a more systematic and comprehensive scope.

[0065] (3) By determining the threshold points of key influencing factors of urban resilience, the present invention ensures that the region can significantly improve the level of urban resilience without investing infinitely high construction costs, while reducing the cost of urban resilience construction and effectively achieving a phased leap in the level of regional resilience.

[0066] (4) The urban resilience regulation method constructed by the present invention adds a method for objectively solving the key influencing factors of urban resilience on the basis of traditional resilience enhancement measures for low-resilience areas, provides threshold settings with high practical significance for key influencing factors, and comprehensively regulates the balanced development of resilience in urban areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0068] Figure 1 This is a flow chart of a comprehensive urban resilience control method based on multi-source data mining and threshold effect according to an embodiment of the present invention;

[0069] Figure 2 This is an analysis of the thresholds of key regulatory factors for urban facility resilience according to an embodiment of the present invention;

[0070] Figure 3 This is an analysis of the thresholds of key regulatory factors for economic resilience in accordance with an embodiment of the present invention;

[0071] Figure 4 This is the threshold analysis of key regulatory factors for community resilience in an embodiment of the present invention;

[0072] Figure 5This is a threshold analysis of key regulatory factors for ecological resilience in an embodiment of the present invention. DETAILED DESCRIPTION

[0073] In order to enable those skilled in the art to better understand the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0074] References to "embodiments" in this application mean that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute an independent or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described in this application may be combined with other embodiments.

[0075] See also Figure 1 This embodiment provides a comprehensive urban resilience control method based on multi-source data mining and threshold effect, including the following steps:

[0076] S1. Build an urban resilience assessment system based on actual needs.

[0077] The urban resilience assessment system of this embodiment is set up as a target layer, a criterion layer, and an indicator layer. The target layer is the resilience of a safe development city, and the criterion layer is the resilience of urban facilities, economic resilience, community resilience, and ecological resilience. Each resilience criterion has multiple representation levels, including absorptive capacity, adaptability, and recovery capacity. Through these representation level settings, 32 indicators are selected as resilience assessment values, as shown in Table 1.

[0078] Table 1 Urban resilience assessment system

[0079]

[0080] (1) Resilience of urban facilities.

[0081] A reasonable urban structure and layout can optimize residents' living experience and facilitate the use and maintenance of urban resources. The area and density of urban roads reflect the abundance of a city's transportation resources, the degree of traffic congestion, and the ability to evacuate people and vehicles during disasters. The density of water supply and drainage pipelines reflects a city's water storage, drainage, and post-disaster water supply capabilities, demonstrating the level of its flood control and drainage infrastructure. The presence of fire stations and sewage treatment plants ensures pre-disaster risk prevention and indirectly reflects the recycling of urban water resources. The prevalence of gas reflects the safety of a city's gas system facilities and its ability to quickly recover from disasters.

[0082] (2) Economic resilience.

[0083] A stable economy is a crucial factor in urban development and a solid foundation for achieving higher levels of development. A sound economic development model is reflected not only in the composition of economic output but also in people's livelihoods and economic development potential. Growth in public budget expenditures and per capita disposable income reflects the intensity of public investment in public services, ensuring improvements in people's economic and living standards. Encouraging research and development (R&D) funding and invention patents reflects economic innovation capacity and the productive potential of economic development. The registered urban unemployment rate is an unemployment statistic that directly reflects economic stability. Per capita GDP, to a certain extent, better reflects the level of prosperity and economic development of a country or region than total GDP. The primary industry's share of GDP reflects economic stability and its reliance on basic industries, while the tertiary industry's share of GDP reflects the quality of the national economy and the optimization of its industrial structure.

[0084] (3) Community resilience.

[0085] Comfortable community development is directly linked to residents' lives and reflects a city's resilience. The establishment of cultural centers and service centers reflects a city's ability to communicate public information. Average power outage duration, high school enrollment rates, and the proportion of elderly residents ensure a city's ability to provide power and rebuild after a disaster, reflecting the city's educational and community development and resource carrying capacity. Basic medical insurance coverage and the establishment of community health service institutions ensure residents' health needs and alleviate their stress, reflecting access to healthcare, community health needs, and post-disaster disease prevention and control capabilities. Population density reflects a city's carrying capacity and resource allocation. The number of emergency shelters reflects its ability to prevent and handle disasters before and during disasters, demonstrating the effectiveness of urban safety initiatives and regional planning.

[0086] (4) Ecological resilience.

[0087] A high-quality ecological environment maintains the health of urban ecosystems, reflecting not only urban green construction planning but also the city's resource allocation and processing capacity. Energy consumption per unit of GDP and the harmless disposal rate of domestic waste reflect a city's resource processing capacity and energy consumption level. Public coverage of meteorological disaster monitoring, forecasting, and early warning information reflects the effectiveness of urban oversight and the timeliness of pre-disaster prevention and response strategies. The number of urban parks, greenery coverage, and per capita park green space reflect the ability of urban green spaces to absorb, recover, and mitigate stormwater and flooding, reflecting the living environment and quality, and demonstrating ecologically livable spaces. The annual ambient air quality compliance rate indicates the ability to prevent post-disaster epidemics, and the daily sewage treatment capacity reflects the pressure sewage places on the ecosystem.

[0088] S2. Urban resilience assessment.

[0089] By analyzing the fundamental elements of urban resilience regulation, a comprehensive urban resilience measurement and assessment system that meets the requirements of safe and sustainable urban development is constructed, providing a scientific, quantitative foundation for urban resilience regulation. To further explore urban resilience regulation methods, it is necessary to first select specific case studies. By analyzing the characteristics of their resilience levels and the inherent influencing patterns, the necessary components, key links, and implementation paths of urban resilience regulation methods are studied, thereby formulating a general urban resilience regulation approach. Because the various characteristics of urban resilience require scientific assessment to effectively identify, urban resilience assessment is a necessary prerequisite and foundational work for regulatory research.

[0090] This example conducts toughness assessment on the research object, specifically as follows:

[0091] First, collect original data on factors influencing urban resilience in the urban areas under study from official sources such as city statistical yearbooks and urban construction status bulletins to ensure data openness and integrity. This original data must first clarify the timeframe of the urban resilience regulation study and secondly, include the research object, such as a specific city's urban planning, which is time-dependent. Thirdly, select factors influencing resilience as needed and obtain specific values ​​directly from publicly available data or calculate them using directly obtained data.

[0092] Secondly, the data is standardized and normalized, and according to the actual application scenarios, a more objective method is selected to assign weight values ​​to each influencing factor, such as entropy method, coefficient of variation method, principal component analysis method, etc.

[0093] Taking the entropy method as an example, it directly uses information entropy to assign weights to factors. This method can reflect the differences and importance of different indicator factors, avoiding the influence of subjective factors. It is a more objective and accurate weighting method. By determining the weights of each indicator, we can roughly determine the extent of each indicator's impact on the level of urban resilience. By assigning weights to each indicator, we can assess urban resilience and establish an algorithmic foundation for research and analysis of urban resilience regulation. This method consists of three steps: dimensionless processing, calculating information entropy, and calculating weights. It is highly objective and accurate and is suitable for weighting urban resilience indicators with different dimensions.

[0094] S21. Given that the various toughness indicators have different dimensions and positive and negative orientations, resulting in large differences in the size of the indicator data, it is not convenient to perform evaluation calculations. Therefore, it is necessary to first perform dimensionless processing on each indicator. Currently, the commonly used dimensionless processing methods include standardization, centering, normalization, averaging, etc. Considering that each indicator has positive and negative orientation requirements, and combining the application characteristics of the entropy method. Therefore, this embodiment uses normalization to perform dimensionless processing on each toughness evaluation value, as shown in the following formula:

[0095] Positive indicators:

[0096] Negative indicators:

[0097] Where i represents the i-th resilience subsystem, j represents the j-th resilience index, and R' ij is the corresponding X ij Indicator data; T ij represents the normalized value of the jth index of the i-th resilience subsystem, min(R' ij ) and max(R' ij ) represent R' ij The minimum and maximum values ​​of

[0098] S22. Calculate the information entropy value e of each indicator ij , as follows:

[0099]

[0100] Among them, n represents the number of indicators, q ij represents the proportion of the jth indicator in the i-th resilience subsystem;

[0101] S23. Calculate the weight W of each indicator ij , as follows:

[0102] According to the above formula, the weight can be directly calculated using statistical analysis software such as SPSS.

[0103] Next, the current regional resilience level is calculated based on the resilience assessment value and its weight, and the results are ranked and classified into urban resilience levels, and areas with low urban resilience levels are identified.

[0104] The current regional resilience level is calculated based on the resilience assessment value and its weight. Specifically, the weighted sum of the resilience assessment values ​​is calculated to obtain the current regional resilience level representation, as shown in the following formula:

[0105]

[0106] Among them, W ij is the weight of each indicator, T ij represents the normalized value of the jth index of the i-th toughness criterion, W ij T ij Used to calculate the resilience evaluation value of the positive indicator, W ij (1-T ij ) is used to calculate the resilience assessment value of negative indicators.

[0107] Then, compare and sort the above URLs to identify areas with low urban resilience levels.

[0108] S3. Analysis of the intrinsic characteristics of basic urban regulatory elements.

[0109] The heterogeneity of urban resilience distribution is determined by the disparities in resilience-influencing factors across cities and their impact on urban subsystems. Resilience factors themselves possess distinct characteristics, and certain connections and interactions exist between these factors. Together with the interdependencies between urban subsystems, these factors directly influence the resilience of urban subsystems. This example analyzes the inherent characteristics of factors influencing urban resilience from the perspectives of correlation, differentiation, and sensitivity, exploring the causes of uneven urban resilience development and laying the foundation for subsequent analysis of key factors influencing urban resilience.

[0110] It should be noted that the basic regulatory elements of a city include three levels from low to high and from inside to outside, namely the indicator level - factors affecting urban resilience, the criterion level - urban subsystems and regional resilience. Considering the research characteristics of the three levels, the relevant characteristics between the levels are studied separately. The regional resilience includes the resilience of multiple urban subsystems, and the resilience evaluation of each urban subsystem has multiple resilience indicators, among which the urban subsystem is divided into urban facilities subsystem, economic subsystem, community subsystem and ecological subsystem.

[0111] 1. Correlation analysis.

[0112] (1) Correlation analysis between resilience indicators and urban subsystems: The Spearman correlation coefficient was used to measure the correlation between resilience influencing factors within the urban area to be studied, as shown in the following formula:

[0113]

[0114] Among them, d i is the difference in the rank values ​​of the i-th data pair, n is the total number of observed samples, and the correlation coefficient ρ is [-1,1]. When it is greater than 0, it indicates that the variables are positively correlated, and when it is less than 0, it indicates that the variables are negatively correlated.

[0115] The correlation between factors influencing urban resilience and the resilience of urban subsystems demonstrates the closeness of the relationship between each factor and the resilience of the corresponding subsystem, indirectly reflecting the importance of resilience factors. Considering that factors influencing urban resilience are nonlinear and fall under the category of nonparametric statistics, this example directly measures the correlation between each urban subsystem under study and its internal resilience factors using the Spearman correlation coefficient in SPSS PRO. Heat maps are then drawn to analyze the correlation between factors influencing urban resilience and the resilience of urban subsystems.

[0116] (2) Correlation between urban subsystems: The coupling correlation between the urban subsystems to be studied is calculated using the coupling coordination model, as shown in the following formula:

[0117]

[0118] Where k is the number of urban subsystems, R i is the toughness evaluation value of each criterion layer.

[0119] The resilience correlation between urban subsystems represents the degree of interaction and mutual influence between these systems, reflected in the coordinated development of these systems. The coupling coordination degree is used to analyze the coordinated development level and the quality of coordination within a given entity. It reflects the dynamic relationship between the interactions and coordinated development between systems and is widely used in empirical research on the coupling development level between various systems, including the economy, urbanization, agriculture, industry, and population. This paper uses a modified coupling coordination degree model to measure and analyze the resilience correlation between urban subsystems.

[0120] (3) Correlation between urban subsystems and regional resilience: The spatial correlation between the urban subsystems to be studied and the regional resilience was calculated using the local Moran's I index in the spatial autocorrelation using GeoDa software.

[0121] It needs to be explained that the study of resilience in urban areas is inevitably affected by geographical location. Analyzing the resilience-related characteristics of the two from a spatial perspective can better illustrate the regional differences in subsystem resilience levels and the correlation between subsystem resilience in different regions, thereby obtaining spatial distribution characteristics.

[0122] 2. Difference analysis.

[0123] Differences are the primary cause of disparities in resilience levels between urban regions, directly reflecting the spatiotemporal heterogeneity and uneven distribution of urban resilience. The contribution rate of each resilience indicator to the resilience of an urban subsystem determines the extent to which differences in these factors influence regional resilience. Therefore, this example characterizes the differential characteristics of the factors influencing resilience across cities by calculating the contribution rate of these factors to their corresponding urban subsystems, as shown in the following formula:

[0124]

[0125] Among them, R ij is the jth resilience evaluation value of the i-th criterion layer, R i is the comprehensive evaluation value of each criterion layer.

[0126] 3. Sensitivity analysis.

[0127] Sensitivity measures the sensitivity of changes in resilience factors to changes in the resilience of urban subsystems or regional resilience levels. It can be used to explore the robustness of analysis results and the uncertainty of inputs affecting outputs. Sensitivity analysis encompasses both local and global sensitivities. Local sensitivity is primarily achieved by solving partial derivatives, while global sensitivity considers changes across the entire parameter space and assesses the overall impact of input parameters. The impact of urban resilience modulating factors on urban resilience does not exhibit a single linear relationship, making a standard mathematical model impractical. Therefore, it is only applicable to measuring global sensitivity. Among the various global sensitivity analysis methods, most rely on subjective, human-defined ranges for input variables, resulting in a high reliance on insensitive parameters. Compared to these methods, the Random Forest algorithm calculates global sensitivity based on data variance, minimizing the influence of human factors while simultaneously identifying the contributions of key and insensitive parameters. It offers excellent interpretability and robustness, and is widely used in machine learning and data mining. The Random Forest algorithm is often used in multi-criteria decision-making. Without performing feature selection, it assesses the importance of input parameters and predicts the significance of key variables, thereby characterizing the sensitivity of a factor to the dependent variable. Two main approaches are available: Monte Carlo-based random forests and Bootstrap-based random forests. Datasets of factors influencing urban resilience are relatively small, generally lack high linear correlations, and lack a clear overall distribution trend. Therefore, the Bootstrap-based Random Forest algorithm is more suitable for sensitivity analysis.

[0128] Specifically, the sensitivity analysis of this embodiment is specifically: performing global sensitivity calculation using the Bootstrap-based Random Forest algorithm, including:

[0129] (1) Data preparation: Define input variable X = {R i1, Ri2 ,..,R ij} and output variable R i , generate data set D;

[0130] (2) Construct Bootstrap samples: Extract B Bootstrap samples D from the original dataset D with replacement b (b=1,2,..,B), each sample size is n;

[0131] (3) Training random forest model: For each Bootstrap sample D b Randomly select m features for node splitting to generate an unpruned decision tree T b , and get the random forest model

[0132] (4) Variance-based sensitivity: Calculate R for each variable ij The mean of the MSE reductions caused by splitting nodes in all trees;

[0133] (5) Significance test: The statistical significance of the importance score was evaluated by Bootstrap or permutation test.

[0134] S4. Identify the key factors affecting urban resilience.

[0135] By analyzing the inherent characteristics of basic urban regulatory factors, a method for calculating the criticality of factors influencing urban resilience was constructed. The criticality of basic urban regulatory factors is defined as consisting of three dimensions: relevance, difference, and sensitivity. Calculation equations were established for each dimension to comprehensively measure the criticality of factors influencing urban resilience, as follows:

[0136] 1. Calculation of the correlation between basic urban regulatory factors.

[0137] The correlations between urban resilience indicators were calculated using the Spearman correlation coefficients, and the average value was recorded as ρ. xij ; In the correlation between the urban resilience criteria layers, the resilience coupling between each two is calculated and the average value is recorded as In the correlation between the urban resilience criterion layer and regional resilience, the local Moran's I index of each criterion layer's resilience to each region's resilience level is calculated, denoted as The established calculation equation for the correlation CM of factors affecting urban resilience is as follows:

[0138] When ρ xij >0,

[0139] When ρ xij <0,

[0140] Where i is the resilience subsystem of the i-th city, k is the resilience subsystem of the city other than the i-th, j is the resilience index of the j-th city, n is the year, R ij is the evaluation value of each influencing factor, R i is the resilience assessment value of each urban subsystem.

[0141] 2. Calculation of differences in basic urban regulatory factors.

[0142] In the differential analysis of factors influencing urban resilience, the contribution of each influencing factor to the resilience of the urban subsystem to which it belongs is analyzed. xij , combined with the evaluation and assignment of each influencing factor, the following DM calculation equation for the difference degree of factors affecting urban resilience is established:

[0143] For positive indicator factors, DM=R ij I xij ;

[0144] For negative indicator factors, DM=(1-R ij )I xij .

[0145] 3. Calculation of sensitivity of basic urban regulatory factors.

[0146] The sensitivity of urban resilience factors is the degree to which changes in resilience factors affect the resilience of urban subsystems. This example uses the Random Forest algorithm to perform a sensitivity analysis on factors affecting urban resilience and calculates the sensitivity of each resilience factor to the urban resilience criterion layer to which it belongs, denoted as S. xij , combined with the evaluation and assignment of each influencing factor, the following calculation equation for the sensitivity SM of urban resilience influencing factors is established:

[0147] For positive indicator factors, SM=R ij S xij ;

[0148] For negative indicator factors, SM=(1-R ij )S xij .

[0149] Based on the correlation, difference and sensitivity calculations of the above basic urban regulatory factors, a general method for calculating the criticality of factors affecting urban resilience is constructed as follows: KD = CM + DM + SM.

[0150] Based on this, the criticality of all factors affecting urban resilience is calculated, and a certain number of key factors affecting urban resilience are selected according to actual needs.

[0151] S4. Threshold effect of key influencing factors of urban resilience and urban resilience regulation.

[0152] First, for the key influencing factors of urban resilience that have been solved and selected, the critical values ​​(i.e., threshold points) of these key influencing factors are determined through threshold effect analysis when the urban area's resilience level achieves a phased leap. The non-parametric smooth curves of the key influencing factors of urban resilience are drawn using the Cubic Spline Interpolation algorithm to quantitatively characterize their impact trends on urban subsystems; the Gaussian kernel density algorithm is used to draw the distribution density characteristic curves of the key influencing factors under different resilience levels, depicting the factor agglomeration characteristics under different resilience levels; the comprehensive scatter-influence trend diagram (hereinafter referred to as the influence trend diagram) and the distribution characteristic curve are used to jointly analyze the threshold points where each key influencing factor plays a key role, such as Figure 2 shown.

[0153] The specific steps are as follows:

[0154] Using the cubic spline interpolation algorithm, based on the resilience assessment values ​​of key influencing factors of urban resilience in different regions arranged in ascending order, we obtained an impact trend map, including a scatter plot and smooth curve of the distribution of key influencing factors;

[0155] Using the Gaussian kernel density algorithm, we established a kernel density estimation function based on the distribution of key influencing factors of urban resilience at different levels of urban resilience, and selected an appropriate bandwidth to obtain a smooth density curve.

[0156] The critical transition point of the influencing mechanism is determined by analyzing the scatter distribution and inflection points of the impact trend diagram, and the significance of the threshold interval is verified by the kurtosis characteristics of the density curve. Combined with the analysis of the critical transition point and threshold interval, the threshold point of the key influencing factors of urban resilience is obtained.

[0157] Furthermore, the threshold points of key influencing factors of urban resilience are divided into urban facility resilience threshold, economic resilience threshold, community resilience threshold and ecological resilience threshold.

[0158] (1) In urban facility resilience Figure 2 ), it can be seen from the impact trend diagram on the left that X 16 The values ​​of the factors are mainly concentrated between 0 and 0.1, and the change range of their influence trend is small, indicating that the impact of this factor on the resilience of urban facilities is relatively stable. 17 The influence trend of the factors remained basically consistent over the three years. When the value was between 0.035 and 0.04, the urban facility resilience was at a high or high level, indicating that X 17 The factors have significant stage characteristics in improving the resilience of urban facilities. The distribution characteristics of the two at different resilience levels show that X 16The peaks of factor distribution are relatively concentrated, indicating that their influence has a strong regularity; 17 The distribution characteristics of the factors show significant differences, reflecting their complex role in the resilience of urban facilities. When the resilience of urban facilities is at a medium to high level, X 16 The value is about 0.003, and X 17 The value of the factor is between 0.01 and 0.03. Combining the influence trend and distribution characteristics of the two, we can approximately set 0.003 as X 16 The threshold point at which the resilience of urban facilities reaches a high level is 0.035, which is the threshold point at which the resilience of urban facilities reaches a high level. 17 The threshold point for the feature.

[0159] (2) Figure 3 In the comparison of the two pictures above and below, we can see that X 25 The impact of factors on economic resilience is relatively flat, indicating that their contribution to economic resilience is relatively stable and sustained. 27 The impact span of the factors is more significant, reflecting their dynamic role in economic resilience. 25 The value of the factor is between 0.003 and 0.005, and X 27 When the value of the factor is between 0.02 and 0.04, the level of economic resilience gradually increases from a low stage to a high stage. However, when economic resilience reaches a high level, X 25 The value of the factor is relatively small, indicating that its role in the high toughness stage is weakened. According to the distribution characteristics under different toughness levels, when the toughness level transitions from medium to high stage, X 27 The value is between 0.01-0.025 and remains in this range in the high toughness stage, indicating that X 27 Factors play a continuous and critical role in improving economic resilience. Based on this, we can approximately determine that 0.004 is the value of X 25 The critical value of the factor that produces a key effect is 0.015. 27 The critical value of the feature.

[0160] (3) Figure 4 In the impact trend of community resilience, X 34 The influence trend of the factor shows a high degree of concentration. When its value increases significantly to about 0.015, the community resilience level transitions from the medium stage to the high stage and then tends to be flat, indicating that this factor has significant stage characteristics and its contribution tends to be stable after reaching a certain threshold. 36The influence trend of the factor shows greater volatility. When its value increases to about 0.02, the community resilience level transitions from the medium stage to the high stage. Although there is a short fault afterwards, it can still reach the high resilience level at around 0.02. In the distribution characteristics, when the community resilience level transitions from the medium stage to the high stage, X 34 and X 36 The values ​​are between 0.01 and 0.02. At higher toughness and high toughness levels, X 34 The values ​​of are roughly the same, which is consistent with the analysis of the impact trend, further verifying that X 34 The stability role of factors in community resilience. Comprehensive analysis can determine that 0.015 is X 34 The threshold point of the feature, 0.02 is X 36 The threshold point for the feature.

[0161] (4) By Figure 5 It can be known that X 41 The impact of factors on ecological resilience shows significant fluctuations, and X 44 The influence of factors is relatively gentle and concentrated. Specifically, when X 41 The factor value is about 0.02, X 44 When the factor value is around 0.01, the ecological resilience is at a medium to high level, which indicates that the two factors have different modes of action in the process of improving ecological resilience. In the distribution characteristics of the different resilience levels of the two, when the ecological resilience achieves a transition from a medium to a high level, X 41 The value is around 0.025, and at a high level, X 41 The values ​​show a bimodal distribution characteristic, that is, concentrated between 0.01 and 0.02, indicating that high ecological resilience can be achieved through the two X 41 In contrast, X 44 The value is relatively stable at different toughness levels, ranging from about 0.01 in the medium to high stages, and drops to about 0.005 in the high toughness stage. Based on the above analysis, 0.025 can be approximately set as X 41 The threshold point of the feature, 0.01 is X 44 The threshold point for the feature.

[0162] As shown in Table 2, based on the thresholds for key regulatory factors for urban resilience determined through the above analysis, a reverse engineering approach was used to determine the original values ​​of these factors before they were assessed and assigned. In urban development, these values ​​can be used to plan the distribution of facilities, industrial restructuring, and land use related to these key regulatory factors. The thresholds for these key regulatory factors provide a reference basis for resilient urban development, facilitating urban planners' control over the configuration of certain urban elements while also meeting the requirements of safe development and resilience.

[0163] Table 2 Threshold points of key regulatory factors for urban resilience

[0164]

[0165] Finally, for the low-resilience regions identified in the urban resilience assessment, a threshold effect analysis was conducted on the key influencing factors of urban resilience, deriving threshold points for these key influencing factors. By regulating resource allocation and capital investment in low- and high-resilience regions, the key influencing factors in low-resilience regions were brought within the threshold range, thereby improving the resilience of low-resilience regions, narrowing the resilience development gap between regions, and promoting balanced urban resilience development.

[0166] The goal of regulating regional heterogeneity in urban resilience is to narrow the gap in resilience development between regions, achieve a spatial structure of urban resilience in which "central regions lead high-resilience development, while surrounding regions collaboratively enhance resilience," and enhance the overall resilience of each region by strengthening the resilience of its subsystems. However, general qualitative regional heterogeneity regulation strategies can only provide guiding advice for urban resilience planning and construction. More specific regulatory pathways are needed, based on the criticality of each strategy, the threshold effects of key influencing factors, and the inherent influencing mechanisms. Ultimately, the following comprehensive and universally applicable urban resilience regulation approach is developed from the four dimensions of urban subsystems.

[0167] (1) Urban facility resilience control methods

[0168] Based on the heterogeneous regulation strategy of urban facility resilience levels and combined with the analysis of the threshold effect and inherent influencing mechanism of key regulatory factors of urban facility resilience, a general comprehensive regulation method for urban facility resilience is proposed.

[0169] Main regulatory targets: areas with low resilience of urban facilities, such as Tianhe District, Yuexiu District, Haizhu District and surrounding areas.

[0170] Key control measures include: 1) increasing multi-route and multi-modal travel options, increasing the proportion of personal and public transportation, and expanding parking facilities; 2) expanding the functions of urban road trunk lines, strengthening inter-regional interconnection networks, and improving road greening; 3) increasing investment in road expansion, maintenance, and upgrades to reduce road wear. Based on these three approaches, the road network density in urban areas will be set to 7.1 km / km². 4) grading underground pipeline corridors based on actual regional pipeline transportation needs, prioritizing the planning of high-grade pipelines; 5) increasing the number, length, and diameter of drainage pipes in areas with dense above-ground buildings to expand natural drainage channels; 6) encouraging businesses, universities, and other institutions to optimize and innovate methods to improve drainage efficiency, increase investment in drainage construction, and improve the quality of drainage pipe materials; 7) in areas with unique geographical environments, prioritizing natural drainage, supplemented by artificial drainage, to alleviate drainage pressure in activity areas. Based on these four approaches, the drainage pipe density in urban areas will be optimized to 12.32 km / km².

[0171] Additional control methods: 1) Optimize the layout of fire-fighting facilities and emergency facilities based on the regional population density distribution, and strengthen risk monitoring in special geographical environments; 2) Regularly inspect the composition and wear of underground pipelines in densely populated areas, and conduct ground inspections of long-distance deep-buried pipelines based on time, year, transportation medium, and frequency of use.

[0172] (2) Economic resilience regulation methods

[0173] Based on the heterogeneous regulation strategy of urban economic resilience levels and combined with the analysis of the threshold effect and internal influencing mechanism of key regulatory factors of economic resilience, a general comprehensive regulation method for economic resilience is proposed.

[0174] Main regulatory targets: regions with low economic resilience.

[0175] Key regulatory measures include: 1) promoting digital transformation to adjust the employment structure, adjusting the proportion of fixed asset investment to increase the number of jobs; 2) controlling population growth and the rate of population aging, and strengthening vocational training and skills development; 3) expanding cross-regional and cross-industry employment, improving benefits such as resource allocation, children's schooling, and vocational training; 4) increasing policy investment in employment subsidies and entrepreneurship support, strengthening social security for the unemployed, and encouraging them to actively seek employment. Based on these four approaches, the registered unemployment rate in urban areas will be reduced to 2.13%. 5) rationally allocating land resources, improving land use efficiency, enhancing soil quality and climate suitability, and controlling the occurrence and minimizing the impact of extreme climate events; 6) increasing investment in the primary industry, accelerating technological innovation in the primary industry, and improving production efficiency; 7) increasing funding for supportive agricultural policies and rationally planning the proportion of land used for greening, industry, commerce, and agriculture. Based on these three approaches, the primary industry-to-GDP ratio in urban areas will be maintained at 2.76%.

[0176] Additional regulatory methods: 1) In combination with regional characteristics, develop a localized and diversified industrial economy and promote the construction of characteristic industrial clusters; 2) Encourage and support the development of scientific and technological innovation, increase regional cooperation, and strengthen economic connectivity among regions.

[0177] (3) Community resilience regulation methods

[0178] Based on the heterogeneous regulation strategy of urban community resilience levels and combined with the analysis of the threshold effect and internal influencing mechanism of key regulatory factors of community resilience, a general comprehensive regulation method for community resilience is proposed.

[0179] Main regulatory targets: areas with low community resilience, such as surrounding areas.

[0180] Key regulatory approaches include: 1) adjusting the number, size, and type of emergency shelters based on the disaster risk level of specific geographic areas; 2) reducing the number or size of emergency shelters in areas with low population density and well-developed transportation networks, while increasing the number and service area of ​​emergency shelters in high-activity areas; 3) improving the seismic and wind resistance of buildings and establishing temporary emergency shelters. Based on these three approaches, the number of emergency shelters in urban areas will be set at 130. 4) coordinating and controlling the proportion of aging and younger residents in communities, encouraging community members to improve their education; 5) increasing medical insurance reimbursement rates for medical treatment, lowering the minimum reimbursement threshold, and encouraging residents to undergo annual health checkups and actively participate in insurance; 6) improving regional medical facilities, increasing community medical consultation service points, and improving the quality of medical services. Based on these three approaches, the basic medical insurance coverage rate for urban residents will be increased to 98.3%.

[0181] Additional regulatory measures: 1) Expand regional employment opportunities, improve medical care, education, housing and other welfare levels, and attract more talents to work and live there; 2) Add community cultural stations, increase community cultural publicity, and optimize the allocation of grassroots cadres; 3) Plan the use of community public spaces, establish multi-functional convenience service stations, and improve supply chain and logistics management systems; 4) Promote the construction of smart communities, strengthen community risk perception, emergency response and other functions, and broaden the channels for publicity and announcement of emergency information.

[0182] (4) Ecological resilience regulation methods

[0183] Based on the heterogeneous regulation strategy of urban ecological resilience levels and combined with the analysis of the threshold effect and inherent influencing mechanism of key regulatory factors of ecological resilience, a general comprehensive regulation method of ecological resilience is proposed.

[0184] Main regulatory targets: areas with low ecological resilience.

[0185] Key regulatory approaches include: 1) promoting interregional energy resource exchange, controlling energy prices and consumption levels, and fostering frugal energy consumption habits; 2) developing advanced energy technologies such as smart grids, accelerating technological innovation in energy-saving technologies, and optimizing energy management; 3) strictly adhering to environmental policies such as carbon emission limits to improve energy efficiency. Based on these three pathways, urban energy consumption per unit of GDP will be controlled at 0.3 tons of standard coal per 10,000 yuan. 4) rationally planning the planting layout of different tree species based on regional climate conditions and soil quality; 5) increasing funding for greening projects and encouraging businesses, universities, and other institutions to innovate and improve irrigation, planting, and greening management technologies; 6) reducing the occupation of green space for construction, adhering to land use and environmental protection policies, and creating ecologically livable spaces in accordance with greening standards; 7) reducing domestic and industrial pollution levels, increasing public awareness of environmental protection, and boosting resident participation in environmental protection efforts. Based on these four pathways, the green coverage rate in urban areas will be raised to 45.8%.

[0186] Additional regulatory measures: 1) Divide urban areas into functional zones, increase the number of street trees and shrubs, and give priority to the construction of small parks on newly demolished land; 2) Promote a green lifestyle, implement electric vehicle pricing policies, and reduce exhaust emissions from traditional fuel vehicles; 3) Promote the research and development and application of clean energy such as hydrogen energy to reduce traditional energy consumption.

[0187] It should be noted that, for the sake of convenience, the aforementioned method embodiments are all expressed as a series of action combinations, but those skilled in the art should know that the present invention is not limited to the described order of actions, because according to the present invention, certain steps can be performed in other orders or simultaneously.

[0188] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0189] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be considered as equivalent replacement methods and are included in the scope of protection of the present invention.

Claims

1. A comprehensive urban resilience control method based on multi-source data mining and threshold effect, characterized by: The steps include: Constructing an urban resilience assessment system, including a target layer, a criterion layer, and an indicator layer. The criterion layer includes multiple resilience criteria, and the indicator layer includes multiple resilience assessment values. Each resilience criterion corresponds to several resilience assessment values. Collect and process raw urban data, conduct resilience assessments of various urban areas based on resilience-influencing factors, obtain weights for each resilience assessment value, calculate the current regional resilience level based on the resilience assessment values ​​and their weights, rank and rank the cities based on the calculated results, and identify areas with low resilience levels. The raw urban data includes time, urban area planning, and specific values ​​of resilience-influencing factors. Analyze the inherent characteristics of basic urban regulatory elements in multiple dimensions to obtain the relevant characteristics between each level. Determine the criticality of basic urban regulatory elements based on the relevant characteristics between each level. Calculate and analyze the criticality to obtain the key influencing factors of resilience. The basic urban regulatory elements include multiple levels. The threshold effect analysis method is used to analyze the key influencing factors of resilience, obtain the critical values ​​of the key influencing factors of resilience, and implement regulation according to the regional resilience level and the critical values ​​of the key influencing factors of resilience.

2. The comprehensive urban resilience control method based on multi-source data mining and threshold effect according to claim 1 is characterized by: The criterion layer includes urban facility resilience, economic resilience, community resilience and ecological resilience. Each criterion layer has multiple representation levels, including absorptive capacity, adaptive capacity and recovery capacity; the resilience assessment value selects a resilience assessment value according to each representation level.

3. The comprehensive urban resilience control method based on multi-source data mining and threshold effect according to claim 1 is characterized by: The above mentioned process involves collecting and processing raw urban data and conducting resilience assessments of various urban areas based on resilience influencing factors, including: Standardize and normalize the original urban data to obtain standard data; Based on the standard data corresponding to the resilience influencing factors, the entropy method is used to evaluate the resilience of various urban areas. Specifically: S21. Use normalization to make each toughness evaluation value dimensionless, as shown in the following formula: Positive indicators: Negative indicators: Where i represents the i-th resilience subsystem, j represents the j-th resilience index, and R' ij is the corresponding X ij Indicator data; T ij represents the normalized value of the jth index of the i-th resilience subsystem, min(R' ij ) and max(R' ij ) represent R' ij The minimum and maximum values ​​of S22. Calculate the information entropy value e of each indicator ij , as follows: Among them, n represents the number of indicators, q ij represents the proportion of the jth indicator in the i-th resilience subsystem; S23. Calculate the weight W of each indicator ij , as follows:

4. The comprehensive urban resilience control method based on multi-source data mining and threshold effect according to claim 1 is characterized by: The current regional resilience level is calculated based on the resilience assessment value and its weight, specifically: the weighted sum of the resilience assessment values ​​is calculated to obtain the current regional resilience level representation, as shown in the following formula: Among them, W ij is the weight of each indicator, T ij represents the normalized value of the jth index of the i-th toughness criterion, W ij T ij Used to calculate the resilience evaluation value of the positive indicator, W ij (1-T ij ) is used to calculate the resilience assessment value of negative indicators.

5. The comprehensive urban resilience control method based on multi-source data mining and threshold effect according to claim 1 is characterized by: The multi-dimensional analysis of the intrinsic characteristics of the basic regulatory elements of the city includes correlation analysis, difference analysis and sensitivity analysis; the basic regulatory elements of the city include resilience indicators, urban subsystems and regional resilience, the regional resilience includes the resilience of multiple urban subsystems, and the resilience evaluation of each urban subsystem has multiple resilience indicators.

6. The comprehensive urban resilience control method based on multi-source data mining and threshold effect according to claim 5 is characterized by: The correlation analysis includes: S31. Correlation analysis between resilience indicators and urban subsystems: The Spearman correlation coefficient was used to measure the correlation between resilience influencing factors within the urban area to be studied, as shown in the following formula: Among them, d i is the difference in the rank value of the i-th data pair, n is the total number of observation samples, and the correlation coefficient ρ is [-1,1]; S32. Correlation between urban subsystems: The coupling correlation between the urban subsystems to be studied is calculated using the coupling coordination model, as shown in the following formula: Where k is the number of urban subsystems, R i is the toughness evaluation value of each criterion layer; S33. Correlation between urban subsystems and regional resilience: Using GeoDa software, the local Moran's I index in spatial autocorrelation was used to measure the spatial correlation between the urban subsystems to be studied and the regional resilience.

7. The comprehensive urban resilience control method based on multi-source data mining and threshold effect according to claim 5 is characterized by: The difference analysis includes calculating the contribution rate of the resilience index to the corresponding urban subsystem as follows: Among them, R ij is the jth resilience evaluation value of the i-th criterion layer, R i is the comprehensive evaluation value of each criterion layer.

8. The comprehensive urban resilience control method based on multi-source data mining and threshold effect according to claim 5 is characterized by: The sensitivity analysis is specifically: performing global sensitivity calculation using the Bootstrap-based Random Forest algorithm, including: Define input variables and output variables and obtain the original data set; Extract several Bootstrap samples from the original data set with replacement and set the sample size; For each Bootstrap sample, a set number of features are randomly selected for node splitting to generate an unpruned decision tree to obtain a random forest model; Calculate the mean MSE reduction resulting from splitting nodes in all trees for each input variable; Importance scores were calculated through Bootstrap or permutation test, and the importance scores of each item were counted.

9. The comprehensive urban resilience control method based on multi-source data mining and threshold effect according to claim 1 is characterized by: The above-mentioned determination of the criticality of basic urban regulatory elements based on the relevant characteristics between various levels includes: Through numerical analysis of the correlation characteristics between each level, the criticality (KD) of resilience influencing factors is constructed as follows: KD = CM + DM + SM, where CM is the correlation of basic urban regulatory factors, DM is the difference of basic urban regulatory factors, and SM is the sensitivity of basic urban regulatory factors. Basic urban regulatory factors include resilience indicators, urban subsystems, and regional resilience. Regional resilience includes the resilience of multiple urban subsystems, and the resilience evaluation of each urban subsystem has multiple resilience indicators. The numerical analysis of the relevant features between the various levels includes: The CM analysis of the correlation between the basic urban control factors is as follows: the correlation between all urban subsystems and resilience indicators is calculated and the average is taken to obtain the mean correlation coefficient ρ xij ; Calculate the coupling degree between each urban subsystem and take the average value to obtain the mean resilience coupling degree Calculating spatial correlations between urban subsystems and regional resilience Establish the correlation degree CM of basic urban control elements as follows: Where i is the resilience subsystem of the i-th city, k is the city subsystem other than the i-th city, j is the resilience index of the j-th city, n is the year, R ij is the evaluation value of each toughness index, R i is the resilience assessment value of each urban subsystem; The DM analysis of the differences between the various levels includes: the contribution of each resilience indicator to the resilience of the urban subsystem to which it belongs. xij , analyze the differences in basic urban regulatory factors, combine the evaluation and assignment of various resilience indicators, and establish the difference degree DM of basic urban regulatory factors, as follows: For positive indicator factors, DM=R ij I xij ; For negative indicator factors, DM=(1-R ij )I xij ; The sensitivity SM between each level includes: performing sensitivity analysis on the basic urban regulatory factors through the Random Forest algorithm, calculating the sensitivity of each resilience indicator to the resilience of the urban subsystem to which it belongs, denoted as S xij , and combined with the evaluation and assignment of each factor, the sensitivity SM of the factors affecting urban resilience is established as follows: For positive indicator factors, SM=R ij S xij ; For negative indicator factors, SM=(1-R ij )S xij .

10. The comprehensive urban resilience control method based on multi-source data mining and threshold effect according to claim 1 is characterized in that: The threshold effect analysis method is used to analyze the key factors affecting urban resilience, including: Using the cubic spline interpolation algorithm, based on the resilience assessment values ​​of key influencing factors of urban resilience in different regions arranged in ascending order, we obtained an impact trend map, including a scatter plot and smooth curve of the distribution of key influencing factors; Using the Gaussian kernel density algorithm, we established a kernel density estimation function based on the distribution of key influencing factors of urban resilience at different levels of urban resilience, and selected an appropriate bandwidth to obtain a smooth density curve. The critical transition point of the influencing mechanism is determined by analyzing the scatter distribution and inflection points of the impact trend diagram, and the significance of the threshold interval is verified by the kurtosis characteristics of the density curve. Combined with the analysis of the critical transition point and threshold interval, the threshold point of the key influencing factors of urban resilience is obtained.

Images