Urban flood disaster comprehensive defense capability evaluation method based on weight initialization

Abstract

The invention discloses an urban flood disaster comprehensive defense capability evaluation method based on weight initialization, and the method comprises the steps: collecting historical data of a research region, and carrying out the preprocessing of the historical data; obtaining an evaluation partition based on the historical data, and screening the historical data according to the flood defense influence degree to obtain an evaluation index; constructing an evaluation index system based on the evaluation indexes, performing weight initialization on the evaluation index system by adopting He initialization, and obtaining a comprehensive flood defense grade according to the evaluation index system and the evaluation partitions; and establishing a flood defense evaluation model based on the comprehensive flood defense grade, inputting to-be-evaluated data into the flood defense evaluation model, and outputting an evaluation result. The method not only can improve the precision of urban space flood disaster comprehensive defense capability evaluation, but also has good interpretability, and can be directly applied to an urban space flood disaster comprehensive defense capability evaluation system.

Description

Technical Field

[0001] The present invention relates to the field of urban flood evaluation, and particularly to a method for evaluating the comprehensive flood defense ability of cities based on weight initialization. Background Art

[0002] In recent years, affected by climate change and rapid urbanization, the urban water cycle process has changed significantly. The frequency and intensity of floods in some cities have shown a significant increase. Urban flood events not only directly threaten human life safety but also cause significant social and economic losses. How to effectively defend against flood disasters and minimize the social and economic losses caused by disasters is an important research topic in the current urban flood field. Currently, affected by factors such as land use, terrain conditions, and water conservancy project construction, there are significant differences in the flood disaster defense capabilities of different urban spatial functional areas. Evaluating the flood disaster defense capabilities of different urban regions has become particularly important. In traditional flood defense ability evaluation methods, the defense ability of projects is usually considered, but insufficient consideration is given to factors such as spatial facility layout, emergency management ability, and resource allocation ability. There is an urgent need to propose a rapid and accurate method for evaluating the comprehensive flood defense ability of urban spaces. Summary of the Invention

[0003] The purpose of the present invention is to provide a method for evaluating the comprehensive flood defense ability of cities based on weight initialization.

[0004] To achieve the above purpose, the present invention is implemented according to the following technical solutions:

[0005] The present invention includes the following steps:

[0006] Collect historical data of the research area and preprocess the historical data;

[0007] Obtain evaluation sub-areas based on the historical data, and screen the historical data according to the flood defense influence degree to obtain evaluation indicators;

[0008] Construct an evaluation index system based on the evaluation indicators, use He initialization to initialize the weights of the evaluation index system, and obtain the comprehensive flood defense level according to the evaluation index system and the evaluation sub-areas;

[0009] Establish a flood defense evaluation model based on the comprehensive flood defense level, input the data to be evaluated into the flood defense evaluation model, and output the evaluation result.

[0010] Further, the method for collecting historical data of the research area includes:

[0011] Historical data includes basic geographic data, land use data, industrial distribution data, socioeconomic data, emergency management data, public awareness data, and a historical flood event dataset. The basic geographic data includes topographic maps, remote sensing images, and GIS vector data. The land use data includes land use type distribution maps and land use right ownership information. The industrial distribution data includes enterprise lists, industrial categories, and industrial scales.

[0012] Furthermore, a method for obtaining an evaluation partition based on the historical data includes:

[0013] Using the point data in the GIS vector data of the study area and its corresponding land use type, land use right ownership, and industrial distribution data to establish data points, and clustering the data points:

[0014] Neighborhood of data points:

[0015] N ∈ (p) = {q ∈ D | dist(p, q) ≤ ∈ ∧ q ≠ p ∧ w(q) ≥ w min}

[0016] where N ∈ (p) is a neighborhood with data point p as the neighborhood core, ∈ is the neighborhood radius, D is the set of data points, w(q) is the local density weight of data point q, and w min is the weight threshold of the neighborhood, and dist(p, q) represents the Euclidean distance between data point p and data point q.

[0017] Core point of the neighborhood:

[0018]

[0019] where |N ∈ (p)| is the number of data points in neighborhood N ∈ (p), MinPts is the minimum point number threshold for determining whether a data point can be a core point, ν(p) is the noise term of neighborhood N ∈ (p), and α is the adjustment parameter of the noise term.

[0020] Calculation formula for the cluster center:

[0021]

[0022] where c j is the center point of cluster j, x is the data point within cluster j, w(x) is the local density weight of data point x, C j is cluster j, and I j is the influence term of cluster j and is the variance of the data points within cluster j.

[0023] Objective function:

[0024]

[0025] Among them, β and γ are weight parameters, K is the number of clusters, and λ is the weight of the regularization term. is the compactness of cluster j, and N noise (C j ) is the number of noise points in cluster j, R is the global regularization term, and is the L1 norm.

[0026] Obtain the evaluation partition of the point data in the GIS vector data according to the clustering result.

[0027] Furthermore, a method for screening the historical data according to the flood prevention impact degree to obtain evaluation indicators includes:

[0028] Perform a preliminary screening on the historical data to obtain the first feature:

[0029]

[0030] Among them, u(h,y) is the information gain rate of feature y, n is the number of features of the historical data, h is the historical data, and h k is the k-th feature in the historical data, |h| is the number of historical data, and |h k | and |h m | are the numbers of h k and h m in the historical data respectively, h m is the feature subset of the m-th feature value in feature y, m is the index of the feature value in feature y, and E(h m ) is the conditional entropy of h m . Select features using the information gain rate of the historical data to obtain the first feature.

[0031] Convert the first feature into a feature vector with the same dimension, fill the missing values with 0, and use the hierarchical clustering algorithm to cluster the first feature to obtain similar features. The calculation formula for the distance between clusters is:

[0032]

[0033] Among them, C S (g d , g b ) is the similarity between the first features g d and g b , ||·|| represents the Euclidean distance, g d and g b are ξ-dimensional feature vectors composed of the d-th and b-th first features, and d i represents the component of vector d in the i-th dimension, and D U (I q , Is ) is I q and I s is the inter-cluster spacing of I, q and I s are different clusters,|I q |and|I s |is I q and I s is the number of the first features included;

[0034] Flood defense influence degree of similar features:

[0035]

[0036] where n 1 is the number of similar features, U o is the flood defense influence degree of the o-th similar feature, is flood disaster impact data, and the flood disaster impact data includes casualty data and economic loss data, I o and I r are the o-th and r-th similar features, D UZ is the sum of all the spacings of similar features, D U (I r ,I o ) is I r and I o is the inter-cluster spacing of, is I o and is the joint probability distribution of, p(I o ) is I o is the marginal probability distribution of, is is the marginal probability distribution of, C S (I r ,I o ) is I r and I o is the similarity degree of and, F is the identity matrix, and T represents transpose,

[0037] The similar features with a flood defense influence degree greater than 0.632 are used as evaluation indicators.

[0038] Furthermore, a method for constructing an evaluation index system based on the evaluation indicators includes:

[0039] The evaluation objective is the evaluation of the urban space flood defense ability. An evaluation index system is constructed by using the analytic hierarchy process based on the evaluation indicators, and it includes a hierarchical evaluation structure with three levels: the criterion layer, the sub-criterion layer, and the index layer;

[0040] Among them, the criterion layer includes five first-level indicators: infrastructure, natural environment, social economy, emergency management, and public awareness. The sub-criterion layer is the second-level indicator connecting the criterion layer and the indicator layer, and the indicator layer is the third-level indicator, which is composed of evaluation indicators.

[0041] Furthermore, the method for initializing the weights of the evaluation index system using He initialization includes:

[0042] Establish a He initialization model using the historical flood event dataset and its corresponding evaluation indicators. The He initialization model uses a neural network algorithm to predict flood disaster impact data based on the evaluation indicators.

[0043] Take the evaluation indicators as the input nodes, calculate the standard deviation of He initialization, and extract the initial weight values of the nodes from the Gaussian distribution:

[0044]

[0045] ω = N(0, Std 2 )

[0046] where std represents the standard deviation, sqrt(·) represents the square root, fan in is the number of input nodes, N(0, Std 2 ) represents a normal distribution with a mean of 0 and a variance of std 2 , and ω is the initial weight value of the node;

[0047] When the predicted loss is less than 1.43%, the He initialization model outputs the weights of the nodes as the weights of the evaluation indicators. The weight of the second-level indicator is the sum of the weights of the third-level indicators it contains, and the weight of the first-level indicator is the sum of the weights of the second-level indicators it contains.

[0048] Furthermore, the method for obtaining the comprehensive flood prevention and control level based on the evaluation index system and the evaluation partition includes:

[0049] The flood prevention and control ability index value of the evaluation partition:

[0050]

[0051] where W is the flood prevention and control ability index value, is the number of first-level indicators, is the number of second-level indicators contained in the first-level indicator , is the number of third-level indicators contained in the second-level indicator , is the th first-level indicator 's weight, is the Secondary indicators The weight of is the weight of the th tertiary indicator

[0052] Sort the flood defense ability index values from small to large to obtain the flood defense ability sequence. Divide the flood defense levels according to the flood defense ability sequence: low in the range of the first percentile (0, 10%], relatively low in the range from the first percentile to the third percentile (10%, 30%], medium in the range from the third percentile to the median (30%, 50%], high in the range from the median to the seventh percentile (50%, 70%], and extremely high in the range from the seventh percentile to the tenth percentile (70%, 100%); the comprehensive flood defense level of the research area is the median level of the flood defense levels of its evaluation sub-areas.

[0053] Furthermore, a method for establishing a flood defense evaluation model based on the comprehensive flood defense level includes:

[0054] Use the evaluation indicators and the comprehensive flood defense level to establish a data set, randomly sample and divide the data set into a training set and a validation set. Among them, the comprehensive flood defense level is the classification label of the evaluation indicators. Use the training set to learn the mapping relationship between the evaluation indicators and the comprehensive flood defense level based on the SVM algorithm, establish a flood defense evaluation model, and use the validation set to adjust the hyperparameters of the flood defense evaluation model.

[0055] In a second aspect, an embodiment of the present application further provides an electronic device, including:

[0056] A processor; and a memory arranged to store computer-executable instructions, and the executable instructions, when executed, cause the processor to execute the method steps described in the first aspect.

[0057] In a third aspect, an embodiment of the present application further provides a computer-readable storage medium, and the computer-readable storage medium stores one or more programs. When the one or more programs are executed by an electronic device including a plurality of application programs, the electronic device is caused to execute the method steps described in the first aspect.

[0058] The beneficial effects of the present invention are:

[0059] The present invention is a method for evaluating the comprehensive flood disaster defense ability of a city based on weight initialization. Compared with the prior art, the present invention has the following technical effects:

[0060] The present invention improves the accuracy and interpretability of flood disaster defense ability evaluation through weight initialization, comprehensively evaluates the urban defense ability by integrating multiple factors, quickly responds to emergencies, is applicable to different regions, and improves the efficiency of urban flood disaster management. Brief Description of the Drawings

[0061] Figure 1 It is a flowchart of the steps of a method for evaluating the comprehensive flood prevention and control ability of cities based on weight initialization according to the present invention;

[0062] Figure 2 It is a schematic structural diagram of an electronic device in an embodiment of the present specification. Detailed Embodiments

[0063] The present invention will be further described below through specific embodiments. The illustrative embodiments and descriptions of the present invention are used to explain the present invention, but not to limit the present invention.

[0064] A method for evaluating the comprehensive flood prevention and control ability of cities based on weight initialization according to the present invention includes the following steps:

[0065] As Figure 1 shown, in this embodiment, it includes the following steps:

[0066] Collect historical data of the research area and preprocess the historical data;

[0067] In actual evaluation, evaluate the urban flood prevention and control ability of Research Area A, clean the collected historical data, remove invalid data and error data, convert text data into data codes, and standardize the historical data;

[0068] Obtain evaluation sub - regions based on the historical data, and screen the historical data according to the flood prevention and control influence degree to obtain evaluation indicators;

[0069] In actual evaluation, divide Research Area A into the main urban area, slope - gully area, and near - reservoir area, and screen out 21 evaluation indicators;

[0070] Construct an evaluation index system based on the evaluation indicators, use He initialization to initialize the weights of the evaluation index system, and obtain the comprehensive flood prevention and control level according to the evaluation index system and the evaluation sub - regions;

[0071] In actual evaluation, the evaluation index system and the weights of the indicators:

[0072]

[0073]

[0074] Among them, the industrial structure includes industrial classification and the economic contribution ratio of each industry, the integrity of the emergency plan includes the coverage and detail of the plan, the publicity and education activities include the number of activities and the number of participants, the publicity and education materials include the number of distributed materials, and the self - rescue and mutual - rescue skills training includes the number of training sessions and the number of participants.

[0075] Flood control level:

[0076] The main urban area is at a medium level, the slope and gully area is at a relatively low level, and the area near the reservoir is at a low level. The comprehensive flood control capacity level of the urban space is obtained as relatively low;

[0077] Based on the comprehensive flood control level, a flood control evaluation model is established. The data to be evaluated is input into the flood control evaluation model, and the evaluation result is output.

[0078] In this embodiment, the method for collecting historical data of the research area includes:

[0079] The historical data includes basic geographical data, land use data, industrial distribution data, social and economic data, emergency management data, public awareness data, and historical flood event datasets. The basic geographical data includes topographic maps, remote sensing images, and GIS vector data. The land use data includes land use type distribution maps and land use right ownership materials. The industrial distribution data includes enterprise lists, industrial categories, and industrial scales.

[0080] In this embodiment, the method for obtaining the evaluation partition based on the historical data includes:

[0081] Using the point data in the GIS vector data of the research area and its corresponding land use type, land use right ownership, and industrial distribution data to establish data points, and clustering the data points:

[0082] Neighborhood of data points:

[0083] N ∈ (p) = {q ∈ D|dist(p, q) ≤ ∈ ∧ q ≠ p ∧ w(q) ≥ w min}

[0084] where N ∈ (p) is the neighborhood with data point p as the neighborhood core, ∈ is the neighborhood radius, D is the set of data points, w(q) is the local density weight of data point q, and w min is the weight threshold of the neighborhood, and dist(p, q) represents the Euclidean distance between data point p and data point q.

[0085] Core point of the neighborhood:

[0086]

[0087] where |N ∈ (p)| is the number of data points in the neighborhood N ∈ (p), MinPts is the minimum point number threshold used to determine whether a data point can become a core point, and ν(p) is the neighborhood N ∈(p)'s noise term, α is the adjustment parameter of the noise term,

[0088] Calculation formula for cluster center:

[0089]

[0090] where c j is the center point of cluster j, x is the data point within cluster j, w(x) is the local density weight of data point x, C j is cluster j, I j is the influence term of cluster j, which is the variance of the data points within cluster j,

[0091] Objective function:

[0092]

[0093] where β and γ are weight parameters, K is the number of clusters, λ is the weight of the regularization term, is the compactness of cluster j, N noise (C j ) is the number of noise points within cluster j, R is the global regularization term, which is the L1 norm,

[0094] Obtain the evaluation partition of the point data in the GIS vector data according to the clustering result.

[0095] In this embodiment, the method for screening the historical data to obtain the evaluation index according to the flood prevention influence degree includes:

[0096] Perform a preliminary screening on the historical data to obtain the first feature:

[0097]

[0098] where u(h,y) is the information gain rate of feature y, n is the number of features of the historical data, h is the historical data, h k is the k-th feature in the historical data, |h| is the number of historical data, |h k | and |h m | are the numbers of h k and h m in the historical data respectively, h m is the feature subset of the m-th feature value in feature y, m is the index of the feature value in feature y, E(h m ) is the conditional entropy of h m , and select the feature by using the information gain rate of the historical data to obtain the first feature,

[0099] Convert the first feature into a feature vector with the same dimension, fill the missing values with 0, and use the hierarchical clustering algorithm to cluster the first feature to obtain similar features. The calculation formula for the inter-cluster distance is:

[0100]

[0101] Among them, C S (g d , g b ) is the similarity between the first features g d and g b , ||·|| represents the Euclidean distance, g d and g b are the ξ-dimensional feature vectors composed of the d-th and b-th first features, d i represents the component of vector d in the i-th dimension, D U (I q , I s ) is the inter-cluster distance between I q and I s , I q and I s are different clusters, |I q | and |I s | are the number of first features included in I q and I s ;

[0102] Influence degree of flood prevention of similar features:

[0103]

[0104] Among them, n 1 is the number of similar features, U o is the influence degree of flood prevention of the o-th similar feature, is the flood disaster impact data, and the flood disaster impact data includes casualty data and economic loss data, I o and I r are the o-th and r-th similar features, D UZ is the sum of all distances of similar features, D U (I r , I o ) is the inter-cluster distance between I r and I o ; is I o and 's joint probability distribution, p(I o ) is I o 's marginal probability distribution, is 's marginal probability distribution, C S (I r , I o ) is I r and I oThe similarity, F is the identity matrix, T represents the transpose,

[0105] The similar features with a flood prevention and control impact degree greater than 0.632 are used as evaluation indicators.

[0106] In this embodiment, the method for constructing an evaluation index system based on the evaluation indicators includes:

[0107] The evaluation objective is the evaluation of the urban space flood prevention and control ability. Based on the evaluation indicators, the analytic hierarchy process is used to construct an evaluation index system, which includes a hierarchical evaluation structure with three levels: the criterion layer, the sub-criterion layer, and the index layer;

[0108] Among them, the criterion layer includes five first-level indicators: infrastructure, natural environment, social economy, emergency management, and public awareness. The sub-criterion layer is the second-level indicator connecting the criterion layer and the index layer, and the index layer is the third-level indicator, which is composed of evaluation indicators.

[0109] In this embodiment, the method for initializing the weights of the evaluation index system using He initialization includes:

[0110] Use the historical flood event dataset and its corresponding evaluation indicators to establish a He initialization model. The He initialization model uses a neural network algorithm to predict flood disaster impact data according to the evaluation indicators,

[0111] Take the evaluation indicators as the input nodes, calculate the standard deviation of He initialization, and extract the initial weight values of the nodes from the Gaussian distribution:

[0112]

[0113] ω = N(0, Std 2 )

[0114] where std represents the standard deviation, sqrt(·) represents the square root, fan in is the number of input nodes, N(0, Std 2 ) represents a normal distribution with a mean of 0 and a variance of std 2 , and ω is the initial weight value of the node;

[0115] When the predicted loss is less than 1.43%, the He initialization model outputs the weights of the nodes as the weights of the evaluation indicators. The weights of the second-level indicators are the sum of the weights of the third-level indicators they contain, and the weights of the first-level indicators are the sum of the weights of the second-level indicators they contain.

[0116] In this embodiment, the method for obtaining the comprehensive flood prevention and control level according to the evaluation index system and the evaluation partition includes:

[0117] The flood prevention and control ability index value of the evaluation partition:

[0118]

[0119] Among them, W is the index value of flood prevention and control ability, is the number of first-level indicators, is the first-level indicator is the number of second-level indicators included, is the second-level indicator is the number of third-level indicators included, is the weight of the i-th first-level indicator ; is the th second-level indicator ; is the th third-level indicator ;

[0120] Sort the flood prevention and control ability index values from small to large to obtain the flood prevention and control ability sequence, and divide the flood prevention and control levels according to the flood prevention and control ability sequence: the range within the first percentile (0, 10%] is low, the range from the first percentile to the third percentile (10%, 30%] is relatively low, the range from the third percentile to the median (30%, 50%] is medium, the range from the median to the seventh percentile (50%, 70%] is high, and the range from the seventh percentile to the tenth percentile (70%, 100%] is extremely high; the comprehensive flood prevention and control level of the research area is the median level of the flood prevention and control levels of its evaluation sub-areas.

[0121] In this embodiment, the method for establishing a flood prevention and control evaluation model based on the comprehensive flood prevention and control level includes:

[0122] Use the evaluation indicators and the comprehensive flood prevention and control level to establish a data set, randomly sample and divide the data set into a training set and a validation set. Among them, the comprehensive flood prevention and control level is the classification label of the evaluation indicators. Use the training set to learn the mapping relationship between the evaluation indicators and the comprehensive flood prevention and control level based on the SVM algorithm, establish a flood prevention and control evaluation model, and use the validation set to adjust the hyperparameters of the flood prevention and control evaluation model.

[0123] Figure 2 is a schematic structural diagram of an electronic device according to an embodiment of the present application. Please refer to Figure 2 , at the hardware level, the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and a memory. Among them, the memory may include a memory, such as a high-speed random access memory (Random-Access Memory, RAM), and may also include a non-volatile memory, such as at least one disk memory, etc. Of course, the electronic device may also include other hardware required for other services.

[0124] The processor, network interface, and memory can be interconnected through an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, an EISA (Extended Industry Standard Architecture) bus, or the like. The bus can be divided into an address bus, a data bus, a control bus, etc. For the sake of convenience in representation, Figure 2 only a bidirectional arrow is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0125] Memory, which is used to store programs. Specifically, the program can include program code, and the program code includes computer operation instructions. The memory can include a memory and a non-volatile memory, and provide instructions and data to the processor.

[0126] The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it, forming a comprehensive defense ability evaluation device for urban space flood disasters based on weight initialization at the logical level. The processor executes the program stored in the memory and is specifically used to execute any one of the aforementioned comprehensive defense ability evaluation methods for urban flood disasters based on weight initialization.

[0127] As described above in this application Figure 1A method for evaluating the comprehensive defense ability against urban flood disasters based on weight initialization disclosed in the illustrated embodiments can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the integrated logic circuit in the hardware of the processor or instructions in the form of software. The above-mentioned processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. It can implement or execute various methods, steps, and logic block diagrams disclosed in the embodiments of the present application. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in combination with the embodiments of the present application can be directly embodied as being executed and completed by the hardware decoding processor, or executed and completed by a combination of the hardware and software modules in the decoding processor. The software module may be located in a mature storage medium in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. This storage medium is located in the memory, and the processor reads the information in the memory and combines its hardware to complete the steps of the above method.

[0128] The electronic device can also execute Figure 1 a method for evaluating the comprehensive defense ability against urban flood disasters based on weight initialization, and implement Figure 1 the functions of the illustrated embodiments, which are not elaborated herein in the embodiments of the present application.

[0129] The embodiments of the present application also propose a computer-readable storage medium that stores one or more programs. The one or more programs include instructions that, when executed by an electronic device including multiple application programs, execute any of the foregoing methods for evaluating the comprehensive defense ability against urban flood disasters based on weight initialization.

[0130] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0131] The present application is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in Figure 1 one or more of the processes or Figure 1 blocks or multiple blocks.

[0132] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including instruction means, and the instruction means implements the functions specified in Figure 1 one or more of the processes or Figure 1 blocks or multiple blocks.

[0133] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one or more of the processes or Figure 1 blocks or multiple blocks.

[0134] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and memory.

[0135] The memory may include non-permanent memory in the computer-readable medium, random access memory (RAM), and / or non-volatile memory in the form of, for example, read-only memory (ROM) or flash memory (flashRAM). The memory is an example of a computer-readable medium.

[0136] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0137] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0138] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0139] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for evaluating the comprehensive defense capability of urban flood disasters based on weight initialization, characterized in that: The following steps are involved: Collect historical data of the study area and pre-process the historical data; Based on the historical data, evaluation zones are obtained, and the historical data are screened according to the flood defense impact to obtain evaluation indicators; An evaluation index system is constructed based on the evaluation index, the weight of the evaluation index system is initialized by using He initialization, and a comprehensive flood defense level is obtained according to the evaluation index system and the evaluation partition; A flood defense evaluation model is established based on the comprehensive flood defense level, the data to be evaluated is input into the flood defense evaluation model, and the evaluation result is output.

2. According to claim 1, a method for evaluating comprehensive urban flood disaster defense capabilities based on weight initialization is characterized in that: Methods for collecting historical data on the study area, including: Historical data include basic geographic data, land use data, industrial distribution data, socio-economic data, emergency management data, public awareness data and historical flood event data sets. Basic geographic data include topographic maps, remote sensing images and GIS vector data. Land use data include land use type distribution maps and land use rights information. Industrial distribution data include enterprise directories, industrial categories and industrial scales.

3. According to claim 1, a method for evaluating comprehensive urban flood disaster defense capabilities based on weight initialization is characterized in that: The method for obtaining the evaluation partition based on the historical data includes: The point data in the GIS vector data of the study area and its corresponding land use type, land use rights and industrial distribution data are used to establish data points and cluster the data points: Neighborhood of a data point: N ∈ (p)={q∈D|dist(p,q)≤∈∧q≠p∧w(q)≥w min } Among them, N ∈ (p) is the neighborhood with data point p as the neighborhood core, ∈ is the neighborhood radius, D is the set of data points, w(q) is the local density weight of data point q, w min is the weight threshold of the neighborhood, dist(p,q) represents the Euclidean distance between data point p and data point q, The core points of the neighborhood: Among them, |N ∈ (p)| is the neighborhood N ∈ (p) is the number of data points in the neighborhood, MinPts is the minimum point threshold used to determine whether a data point can become a core point, and ν(p) is the neighborhood N ∈ (p), α is the adjustment parameter of the noise term, The calculation formula of cluster center is: Among them, c j is the center point of cluster j, x is the data point in cluster j, w(x) is the local density weight of data point x, C j is cluster j, I j is the influence term of cluster j, is the variance of the data points in cluster j, Objective function: Among them, β and γ are weight parameters, K is the number of clusters, λ is the weight of the regularization term, is the compactness of cluster j, N noise (C j ) is the number of noise points in cluster j, R is the global regularization term, is the L1 norm, The evaluation partition of the point data in the GIS vector data is obtained according to the clustering results.

4. According to claim 1, a method for evaluating comprehensive urban flood disaster defense capabilities based on weight initialization is characterized in that: The method of obtaining the evaluation index by screening the historical data according to the flood defense impact includes: A preliminary screening of historical data yields the first feature: Where u(h,y) is the information gain rate of feature y, n is the number of features of historical data, h is the historical data, and h k is the kth feature in the historical data, |h| is the number of historical data, |h k | and |h m | are historical data respectively k and h m The number of m is the feature subset of the mth eigenvalue in feature y, where m is the index of the eigenvalue in feature y, E(h m ) is h m The conditional entropy of the historical data is used to select the first feature. The first feature is converted into a feature vector of the same dimension, and the missing values ​​are filled with 0. The hierarchical clustering algorithm is used to cluster the first feature to obtain similar features, where the calculation formula for the inter-cluster spacing is: Among them, C S (g d ,g b ) is the first feature g d and g b The similarity of ||·|| represents the Euclidean distance, g d and g b is the ξ-dimensional feature vector composed of the d-th and b-th first features, d i represents the component of vector d in the i-th dimension, D U (I q ,I s ) is I q and I s The inter-cluster spacing, I q and I s For different clusters, |I q |and|I s |For I q and I s The number of first features included; Flood protection impacts with similar characteristics: Where n1 is the number of similar features, U o is the flood defense impact of the oth similar feature, y is the flood disaster impact data, which includes casualties and economic losses, I o and I r is the oth and rth similarity feature, D UZ is the sum of all distances of similar features, D U (I r ,I o ) is I r and I o The inter-cluster spacing, p(I o ,y) is I o The joint probability distribution of and y, p(I o ) is I o The marginal probability distribution of is, p(y) is the marginal probability distribution of y, C S (I r ,I o ) is I r and I o The similarity of, F is the unit matrix, T represents the transpose, The similar characteristics with flood defense impact greater than 0.632 were used as evaluation indicators.

5. According to claim 1, a method for evaluating comprehensive urban flood disaster defense capabilities based on weight initialization is characterized in that: The method for constructing an evaluation index system based on the evaluation index comprises: The evaluation target is to evaluate the urban space flood defense capacity. Based on the evaluation index, the hierarchical analysis method is used to construct the evaluation index system, which includes a three-level hierarchical evaluation structure of criterion layer, sub-criteria layer and indicator layer. The criterion layer includes five first-level indicators: infrastructure, natural environment, social economy, emergency management and public awareness. The sub-criterion layer is the second-level indicator connecting the criterion layer and the indicator layer. The indicator layer is the third-level indicator composed of evaluation indicators.

6. According to claim 1, a method for evaluating comprehensive urban flood disaster defense capabilities based on weight initialization is characterized in that: The method of initializing the weights of the evaluation index system by using He initialization includes: The He initialization model is established using the historical flood event data set and its corresponding evaluation indicators. The He initialization model uses a neural network algorithm to predict the impact of flood disasters based on the evaluation indicators. Take the evaluation index as the input node, calculate the standard deviation of He initialization, and extract the initial weight value of the node from the Gaussian distribution: ω=N(0,Std 2 ) Where std represents standard deviation, sqrt(·) represents square root, and fan in is the number of input nodes, N(0,Std 2 ) means the mean is 0 and the variance is std 2 Normal distribution, ω is the initial weight value of the node; When the prediction loss is less than 1.43%, He initializes the weight of the model output node as the weight of the evaluation index. The weight of the secondary index is the sum of the weights of the tertiary indexes it contains, and the weight of the primary index is the sum of the weights of the secondary indexes it contains.

7. The method for evaluating comprehensive urban flood disaster defense capability based on weight initialization according to claim 1 is characterized in that: The method for obtaining the comprehensive flood defense level according to the evaluation index system and the evaluation zones includes: Flood defense capability index values ​​of the evaluation zones: Among them, W is the flood defense capability index value, z is the number of first-level indicators, and z i X is the first-level indicator i The number of secondary indicators included, z j is the secondary indicator Y j The number of third-level indicators included, p i is the i-th first-level index X i The weight, q j is the jth secondary index Y j The weight, r m is the mth third-level index Z m The weight of The flood defense capability sequence is obtained by sorting the flood defense capability index values ​​from small to large. The evaluation zones are divided into five flood defense levels according to the flood defense capability sequence: the first quartile [0, 10%] is low, the first quartile to the third quartile [10%, 30%] is low, the third quartile to the median [30%, 50%] is medium, the median to the seventh quartile [50%, 70%] is high, and the seventh quartile to the tenth quartile [70%, 100%] is extremely high. The comprehensive flood defense level of the study area is the median level of the flood defense levels of its evaluation zones.

8. The method for evaluating comprehensive urban flood disaster defense capability based on weight initialization according to claim 1 is characterized in that: The method for establishing a flood defense evaluation model based on the comprehensive flood defense level includes: A data set was established using evaluation indicators and comprehensive flood defense levels. The data set was randomly sampled and divided into a training set and a validation set, in which the comprehensive flood defense level was the classification label of the evaluation indicator. The training set was used to learn the mapping relationship between the evaluation indicator and the comprehensive flood defense level based on the SVM algorithm to establish a flood defense evaluation model. The validation set was used to adjust the hyperparameters of the flood defense evaluation model.

9. An electronic device, comprising: processor; as well as A memory arranged to store computer executable instructions, which when executed cause the processor to perform the method according to any one of claims 1 to 8.

10. A computer-readable storage medium storing one or more programs, which, when executed by an electronic device including a plurality of application programs, enables the electronic device to execute the method according to any one of claims 1 to 8.

Images