An intelligent urban and rural earthquake disaster risk assessment method and system

Abstract

The present invention discloses an intelligent urban and rural earthquake disaster risk assessment method and system. The method includes data construction, weight determination, factor correlation degree calculation, construction of an urban and rural earthquake disaster risk assessment model, and urban and rural earthquake disaster risk assessment. The present invention belongs to the field of data processing, and specifically refers to an intelligent urban and rural earthquake disaster risk assessment method and system. This solution introduces an environmental adaptability factor to adjust the regional differences of indicators; uses the coefficient of variation to balance data fluctuations; by introducing a non-linear attenuation factor of the building age, it ensures that the impact of the building age on risk assessment is not underestimated; introduces urban and rural loss weights to balance the urban and rural sample distributions; and suppresses the sensitivity of the model to extreme outliers through multiple extreme loss designs, so that it can still maintain a stable assessment ability on historical extremely large earthquake data; thereby improving the urban and rural earthquake disaster risk assessment effect.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and specifically refers to an intelligent urban and rural earthquake disaster risk assessment method and system. Background Art

[0002] The urban and rural earthquake disaster risk assessment method is a method system for evaluating the affected degree of urban and rural areas in earthquake disasters. This method predicts the disaster risk of different regions so that the government, disaster management agencies, and urban planners can formulate more effective disaster reduction measures. However, the general urban and rural earthquake disaster risk assessment method is affected by extreme values, resulting in unreasonable weight allocation, thus affecting the final disaster risk assessment. It ignores the non-linear impact of time factors, resulting in poor accuracy of the risk assessment of old buildings; the general urban and rural earthquake disaster risk assessment method has insufficient balance and lacks the ability of time dynamic adjustment, and cannot adapt to the trend of earthquake risk changing with time in real time, thus resulting in poor urban and rural earthquake disaster risk assessment effect. Summary of the Invention

[0003] In view of the above situation, in order to overcome the defects of the prior art, the present invention provides an intelligent urban and rural earthquake disaster risk assessment method and system. Aiming at the problem that the general urban and rural earthquake disaster risk assessment method is affected by extreme values, resulting in unreasonable weight allocation, thus affecting the final disaster risk assessment, and ignoring the non-linear impact of time factors, resulting in poor accuracy of the risk assessment of old buildings, this solution introduces an environmental adaptability factor to adjust the regional differences of indicators, so that the weight allocation is more stable and reasonable; the coefficient of variation is used to balance the data fluctuation, avoiding the instability of weight allocation; by introducing the non-linear attenuation factor of the building age, the impact of the building age on the risk is better quantified, ensuring that the impact of the building age on the risk assessment is not underestimated; thereby improving the accuracy of urban and rural earthquake disaster risk assessment; aiming at the problem that the general urban and rural earthquake disaster risk assessment method has insufficient balance, lacks the ability of time dynamic adjustment, and cannot adapt to the trend of earthquake risk changing with time in real time, thus resulting in poor urban and rural earthquake disaster risk assessment effect, this solution introduces urban and rural loss weights to balance the urban and rural sample distribution. The density weight factor reduces the bias of the model towards urban data by increasing the weight of the low-density rural area. The dynamic weight factor considers the characteristics of earthquake risk changing with time, enabling the model to adapt to different risk situations at different times and improving the timeliness; and suppressing the sensitivity of the model to extreme outliers through multiple extreme loss designs, so that it can still maintain a stable assessment ability on historical large earthquake data; thereby improving the urban and rural earthquake disaster risk assessment effect.

[0004] The technical solution adopted by the present invention is as follows: An intelligent urban and rural earthquake disaster risk assessment method provided by the present invention includes the following steps:

[0005] Step S1: Data construction;

[0006] Step S2: Weight determination;

[0007] Step S3: Measurement of factor correlation degree;

[0008] Step S4: Construction of urban and rural earthquake disaster risk assessment model;

[0009] Step S5: Urban and rural earthquake disaster risk assessment.

[0010] Furthermore, in step S1, the data construction is to obtain historical urban and rural earthquake disaster assessment data and disaster risk levels; the historical urban and rural earthquake disaster assessment data includes geological exploration data, social and economic data, building strength data, and environmental characteristic data; the disaster risk level is used as a data label; and missing value processing and feature standardization are performed.

[0011] Furthermore, in step S2, the weight determination is to calculate the environmental information entropy for each index of the historical urban and rural earthquake disaster assessment data, and introduce an environmental adaptability factor , and the environmental information entropy is expressed as: ; ; where is the environmental information entropy of the i-th index; n is the total number of samples; m is the sample index; is the value of the i-th index in the m-th sample; is the sensitivity adjustment parameter; and are respectively the maximum and minimum values of the i-th index in all samples; is the sample mean of the i-th index; using the stabilization formula, introduce the coefficient of variation , and the weight is expressed as: ; ; where is the weight of the i-th index; L is the total number of indexes; u is the index index; and are respectively the entropy values of the i-th index and the u-th index of the m-th sample; G is the average value of the information entropy of all evaluation indexes; and are respectively the standard deviation and mean of the i-th index.

[0012] Furthermore, in step S3, the measurement of factor correlation degree is to quantify the contribution of each risk factor to the overall risk; add a time decay factor , and the correlation coefficient is expressed as: ; ; The factor correlation degree is expressed as: ; ; where is the correlation coefficient of the i-th index in the m-th sample; is the reference sequence value; is the attenuation coefficient; t is the construction age of the sample; is the minimum absolute difference between all factors and the reference sequence; is the maximum absolute difference between all factors and the reference sequence; is the resolution coefficient; set the correlation threshold, and select the indicators with the average correlation coefficient higher than the correlation threshold in all samples as the finally selected features, so as to construct the urban and rural earthquake disaster dataset.

[0013] Furthermore, in step S4, the construction of the urban and rural earthquake disaster risk assessment model specifically includes the following steps:

[0014] Step S41: Architecture design; use an S-layer fully connected neural network as the basic architecture; the input layer directly inputs the feature vector constructed based on the urban and rural earthquake disaster dataset; the hidden layer designs D fully connected layers, and the number of nodes in each layer gradually decreases to capture the internal non-linear relationship of the data, and the ReLU activation function is used; Dropout is added after some hidden layers; the output layer uses the Sigmoid activation function to output the predicted disaster risk level;

[0015] Step S42: Design the urban and rural loss; increase the density weight factor Ws and introduce the dynamic weight factor , and the urban and rural loss FC is expressed as: ; ; ; where, W and are the current weight and the initial weight respectively; P is the classification parameter; and A are the sample prediction label and the true label respectively; , and are the current time, the previous evaluation time and the evaluation time interval respectively; is the density adjustment parameter; is the sample density of the sample's area; is the maximum sample density;

[0016] Step S43: Design of multiple extreme losses; the multiple extreme loss FJ is expressed as: ; where, and are the bandwidth parameters;

[0017] Step S44: Phased strategy; use FC in the 0th to Uth rounds of training; use FJ in the Uth to Eth rounds of training.

[0018] Further, in step S5, the urban and rural earthquake disaster risk assessment system collects real-time urban and rural earthquake disaster assessment data, which is input into the urban and rural earthquake disaster risk assessment model after feature engineering processing, and the disaster risk level output by the model is used as the urban and rural earthquake disaster risk assessment result.

[0019] An intelligent urban and rural earthquake disaster risk assessment system provided by the present invention includes a data construction module, a weight determination module, a factor correlation degree calculation module, an urban and rural earthquake disaster risk assessment model construction module, and an urban and rural earthquake disaster risk assessment module.

[0020] The data construction module obtains historical urban and rural earthquake disaster assessment data and disaster risk levels, and uses the disaster risk levels as data labels.

[0021] The weight determination module determines the weights of various indicators in the historical urban and rural earthquake disaster assessment data by introducing an environmental adaptability factor and a coefficient of variation; the indicators are specific dimensions of the historical urban and rural earthquake disaster assessment data.

[0022] The factor correlation degree calculation module calculates the factor correlation degree by adding a time decay factor; quantifies the contribution of each indicator to the overall risk.

[0023] The urban and rural earthquake disaster risk assessment model construction module constructs a fully connected neural network model, designs an urban and rural loss function and a multiple extreme loss function; uses the urban and rural earthquake disaster assessment data processed by the data construction module and the weight determination module as input, and further constructs an urban and rural earthquake disaster risk assessment model.

[0024] The urban and rural earthquake disaster risk assessment module conducts urban and rural earthquake disaster risk assessment on the real-time collected urban and rural earthquake disaster assessment data based on the urban and rural earthquake disaster risk assessment model.

[0025] The beneficial effects achieved by the present invention using the above solution are as follows:

[0026] (1) Aiming at the problems existing in the general urban and rural earthquake disaster risk assessment method, such as being affected by extreme values, resulting in unreasonable weight distribution, thus affecting the final disaster risk assessment, and ignoring the non-linear influence of time factors, resulting in poor accuracy of the risk assessment of old buildings. This solution introduces an environmental adaptability factor to adjust the regional differences of indicators, thereby making the weight distribution more stable and reasonable; uses the coefficient of variation to balance data fluctuations, avoiding the instability of weight distribution; by introducing a non-linear attenuation factor of the building age, better quantifies the impact of the building age on the risk, ensuring that the impact of the building age on the risk assessment is not underestimated; thereby improving the accuracy of urban and rural earthquake disaster risk assessment.

[0027] (2) Aiming at the problems existing in the general urban and rural earthquake disaster risk assessment method, such as insufficient balance, lack of time dynamic adjustment ability, and inability to adapt to the trend of earthquake risk changing with time in real time, which leads to poor evaluation effect of urban and rural earthquake disaster risk, this solution introduces urban and rural loss weights to balance the urban and rural sample distribution. The density weight factor reduces the bias of the model towards urban data by increasing the weight of the low-density rural areas. The dynamic weight factor takes into account the characteristics of earthquake risk changing with time, enabling the model to adapt to different risk situations in different periods and improving timeliness. In addition, the sensitivity of the model to extreme outliers is suppressed through multiple extreme loss designs, enabling the model to maintain stable evaluation ability even for historical data of extremely large earthquakes. As a result, the evaluation effect of urban and rural earthquake disaster risk is improved. Description of the Drawings

[0028] Figure 1 It is a schematic flowchart of an intelligent urban and rural earthquake disaster risk assessment method provided by the present invention;

[0029] Figure 2 It is a schematic diagram of an intelligent urban and rural earthquake disaster risk assessment system provided by the present invention;

[0030] Figure 3 It is a schematic flowchart of step S4.

[0031] The drawings are used to provide a further understanding of the present invention and constitute a part of the specification. They are used together with the embodiments of the present invention to explain the present invention, but do not constitute a limitation to the present invention. Detailed Embodiments

[0032] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0033] In the description of the present invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc. indicating the orientation or position relationship are based on the orientation or position relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the system or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as a limitation to the present invention.

[0034] Example 1. Refer to Figure 1 , an intelligent urban and rural earthquake disaster risk assessment method provided by the present invention, which includes the following steps:

[0035] Step S1: Data construction; Obtain historical urban and rural earthquake disaster assessment data and disaster risk levels, and use the disaster risk levels as data labels;

[0036] Step S2: Weight determination; Determine the weights of each index in the historical urban and rural earthquake disaster assessment data by introducing environmental adaptability factors and variation coefficients; The index is the specific dimension of the historical urban and rural earthquake disaster assessment data;

[0037] Step S3: Factor correlation measurement; Add a time decay factor for factor correlation measurement; Quantify the contribution of each index to the overall risk;

[0038] Step S4: Construct an urban and rural earthquake disaster risk assessment model; Construct a fully connected neural network model, design an urban and rural loss function and a multiple extreme loss function; Use the urban and rural earthquake disaster assessment data processed in Step S2 and Step S3 as inputs, and then construct an urban and rural earthquake disaster risk assessment model;

[0039] Step S5: Urban and rural earthquake disaster risk assessment; Based on the urban and rural earthquake disaster risk assessment model, conduct urban and rural earthquake disaster risk assessment on the urban and rural earthquake disaster assessment data collected in real time.

[0040] Example 2, refer to Figure 1 , This example is based on the above example. In Step S1, the data construction is to obtain historical urban and rural earthquake disaster assessment data and disaster risk levels; The historical urban and rural earthquake disaster assessment data includes geological exploration data, social and economic data, building strength data, and environmental characteristic data; Use the disaster risk level as the data label; The geological exploration data includes seismic intensity, geological structure, soil liquefaction index, and soil type; The social and economic data includes population density, urbanization level, economic development level, education level, and distribution of social service facilities; The building strength data includes building type, building age, building height, seismic design standard, and seismic performance of buildings; The environmental characteristic data includes terrain, hydrological conditions, and vegetation cover; The disaster risk levels include high risk and low risk; And perform missing value processing and feature standardization; For the continuous variables of seismic intensity and population density, use K-nearest neighbor interpolation to fill based on weighted average of neighboring samples, and perform Z-score standardization processing; For the discrete variables of soil type and building type, use the mode to fill, and convert the category to a numerical form based on target encoding.

[0041] Example 3, refer to Figure 1, this embodiment is based on the above embodiment. In step S2, when determining the weights, the earthquake assessment involves geological exploration data, socio-economic data, and building strength data, which often have different scales, sources, and high uncertainties. By objectively measuring the information content of each index and avoiding abnormal weights caused by high entropy values, a more stable and reasonable weight allocation is made for the indexes with extremely high or low information entropy of the seismic performance of buildings and population density. For each index of the historical urban and rural earthquake disaster assessment data, the environmental information entropy is calculated. When the index fluctuates greatly in a region, it indicates that the index is significantly affected by regional characteristics. Therefore, the influence should be appropriately smoothed in the information entropy calculation, and thus an environmental adaptability factor is introduced. , the environmental information entropy is expressed as: ; ; where is the environmental information entropy of the i-th index; n is the total number of samples; m is the sample index; is the value of the i-th index in the m-th sample; is the sensitivity adjustment parameter; and are the maximum and minimum values of the i-th index in all samples, respectively; is the sample mean of the i-th index; A stabilization formula is adopted to avoid the distortion of weights caused by high entropy values. In the urban and rural earthquake disaster assessment, the information entropy of indexes such as the seismic performance of buildings and population density may be extremely high or low, and weight distortion may occur due to extreme values. Therefore, the coefficient of variation is introduced to make the weight allocation more stable; the weight is expressed as: ; ; where is the weight of the i-th index; L is the total number of indexes; u is the index index; and are the entropy values of the i-th index and the u-th index of the m-th sample, respectively; G is the average value of the information entropy of all evaluation indexes; and are the standard deviation and mean of the i-th index, respectively.

[0042] Example 4, refer to Figure 1 , this embodiment is based on the above embodiment. In step S3, when calculating the factor correlation degree, it is necessary to quantify the contribution of each risk factor to the overall risk; considering that the influence of the building age on the sample changes non-linearly with time, a time decay factor is added, and the correlation coefficient is expressed as: ; ; The factor correlation degree is expressed as: ; ; where is the correlation coefficient of the i-th index in the m-th sample; is the reference sequence value; is the attenuation coefficient; t is the construction age of the sample; is the minimum absolute difference between all factors and the reference sequence; is the maximum absolute difference between all factors and the reference sequence; is the resolution coefficient; set the correlation threshold, and select the indicators with the average correlation coefficient higher than the correlation threshold among all samples as the finally selected features, so as to construct the urban-rural earthquake disaster dataset.

[0043] By performing the above operations, for the problems existing in the general urban-rural earthquake disaster risk assessment method, such as being affected by extreme values, resulting in unreasonable weight distribution, thus affecting the final disaster risk assessment, and ignoring the non-linear influence of time factors, resulting in poor accuracy of the risk assessment of old buildings, this solution introduces an environmental adaptability factor to adjust the regional differences of indicators, so that the weight distribution is more stable and reasonable; uses the coefficient of variation to balance data fluctuations, avoiding the instability of weight distribution; by introducing the non-linear attenuation factor of the construction age, better quantifies the impact of the building age on the risk, ensuring that the impact of the construction age on the risk assessment is not underestimated; and then improves the accuracy of the urban-rural earthquake disaster risk assessment.

[0044] Example 5, refer to Figure 1 and Figure 3 Based on the above example, in step S4, the urban-rural earthquake disaster risk assessment model is constructed to process the complex data in the urban-rural earthquake disaster risk assessment. A multi-layer fully connected neural network is used, and the number of nodes in each layer gradually decreases, aiming to capture the non-linear relationship of the data; construct the urban-rural earthquake disaster risk assessment model based on the urban-rural earthquake disaster dataset; specifically include the following steps:

[0045] Step S41: Architecture design; use an S-layer fully connected neural network as the basic architecture; the input layer directly inputs the feature vector constructed based on the urban-rural earthquake disaster dataset; the hidden layer designs D fully connected layers, and the number of nodes in each layer gradually decreases to capture the internal non-linear relationship of the data, and the ReLU activation function is used; add Dropout after some hidden layers; the output layer uses the Sigmoid activation function to output the predicted disaster risk level;

[0046] Step S42: Design the urban-rural loss; aiming at the difference in the number of urban and rural samples, increase the density weight factor Ws, and give higher weights to the low-density areas in the countryside to avoid the model being biased towards urban data; considering that the earthquake risk assessment in different regions may change over time, introduce the dynamic weight factor , the urban-rural loss FC is expressed as: ; ; ; where, W and are the current weight and the initial weight respectively; P is the classification parameter; A and are the sample prediction label and the true label respectively; 、 and are the current time, the previous evaluation time, and the evaluation time interval respectively; is the density adjustment parameter; is the sample density of the area where the sample belongs; is the maximum sample density;

[0047] Step S43: Multiple extreme loss design to enhance the adaptability to extreme outliers of historical great earthquakes; The multiple extreme loss FJ is expressed as: ; where and are the bandwidth parameters;

[0048] Step S44: Phased strategy; FC is adopted in the 0th to Uth rounds of training to mainly solve the problem of sample imbalance; FJ is adopted in the Uth to Eth rounds of training to improve the adaptability to outliers.

[0049] By performing the above operations, aiming at the problems of insufficient balance, lack of time dynamic adjustment ability, and inability to adapt to the trend of earthquake risk changing with time in the general urban and rural earthquake disaster risk assessment method, which leads to poor evaluation effect of urban and rural earthquake disaster risk, this solution introduces urban and rural loss weights to balance the urban and rural sample distribution. The density weight factor reduces the bias of the model towards urban data by increasing the weight of low-density rural areas. The dynamic weight factor considers the characteristics of earthquake risk changing with time, enabling the model to adapt to different risk situations at different times and improving timeliness; and suppresses the sensitivity of the model to extreme outliers through multiple extreme loss design, so that it can still maintain a stable evaluation ability on historical great earthquake data; thereby improving the evaluation effect of urban and rural earthquake disaster risk.

[0050] Example Six, refer to Figure 1 In this example, based on the above example, in step S5, the urban and rural earthquake disaster risk assessment system collects urban and rural earthquake disaster assessment data in real time. After being processed by feature engineering, the data is input into the urban and rural earthquake disaster risk assessment model, and the disaster risk level output by the model is used as the urban and rural earthquake disaster risk assessment result; when the output disaster risk level is high risk, a warning is issued to the management personnel.

[0051] Example Seven, refer to Figure 2 In this example, based on the above example, an intelligent urban and rural earthquake disaster risk assessment system provided by the present invention includes a data construction module, a weight determination module, a factor correlation degree calculation module, an urban and rural earthquake disaster risk assessment model construction module, and an urban and rural earthquake disaster risk assessment module;

[0052] The data construction module obtains historical urban and rural earthquake disaster assessment data and disaster risk levels, and uses the disaster risk level as a data label;

[0053] The weight determination module determines the weights of each index by introducing an environmental adaptability factor and a coefficient of variation; the index is a specific dimension of the historical urban and rural earthquake disaster assessment data;

[0054] The factor correlation degree calculation module calculates the factor correlation degree by adding a time decay factor; quantifies the contribution of each index to the overall risk;

[0055] The urban and rural earthquake disaster risk assessment model construction module constructs a fully connected neural network model, designs an urban and rural loss function and a multiple extreme loss function; uses the urban and rural earthquake disaster assessment data processed by the weight determination module and the factor correlation degree calculation module as input, and then constructs an urban and rural earthquake disaster risk assessment model;

[0056] The urban and rural earthquake disaster risk assessment module conducts an urban and rural earthquake disaster risk assessment on the urban and rural earthquake disaster assessment data collected in real time based on the urban and rural earthquake disaster risk assessment model.

[0057] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device.

[0058] Although the embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made in these embodiments without departing from the principles and spirit of the present invention.

[0059] The above describes the present invention and its implementation manners. Such description is not restrictive. What is shown in the drawings is only one of the implementation manners of the present invention, and the actual structure is not limited thereto. Generally speaking, if those of ordinary skill in the art are inspired by it and design similar structural manners and embodiments without creative efforts without departing from the purpose of the present invention, they shall fall within the protection scope of the present invention.

Claims

1. An intelligent urban and rural earthquake disaster risk assessment method, characterized by: The method comprises the following steps: Step S1: Data construction: Obtain historical urban and rural earthquake disaster assessment data and disaster risk levels, and use the disaster risk level as a data label; Step S2: weight determination; by introducing environmental adaptability factors and coefficients of variation, the weights of various indicators in the historical urban and rural earthquake disaster assessment data are determined; the indicators are specific dimensions of the historical urban and rural earthquake disaster assessment data; Step S3: Calculate the correlation of factors; add the time decay factor to calculate the correlation of factors; quantify the contribution of each indicator to the overall risk; Step S4: constructing an urban and rural earthquake disaster risk assessment model; constructing a fully connected neural network model, designing an urban and rural loss function and a multiple extreme loss function; taking the urban and rural earthquake disaster assessment data processed by steps S2 and S3 as input, and then constructing an urban and rural earthquake disaster risk assessment model; Step S5: urban and rural earthquake disaster risk assessment: Based on the urban and rural earthquake disaster risk assessment model, the urban and rural earthquake disaster risk assessment is performed on the urban and rural earthquake disaster assessment data collected in real time; In step S2, the weight determination is to calculate the environmental information entropy of each indicator and introduce the environmental adaptability factor , the environmental information entropy is expressed as: ; ;in, is the environmental information entropy of the ith indicator; n is the total number of samples; m is the sample index; is the sample value of the ith indicator at the mth sample; is the sensitivity adjustment parameter; and are the maximum and minimum values ​​of the i-th indicator in all samples respectively; is the sample mean of the i-th indicator; using the stabilization formula, the coefficient of variation is introduced , the weight is expressed as: ; ;in is the weight of the i-th indicator; L is the total number of indicators; u is the indicator index; and are the entropy values ​​of the i-th indicator and the u-th indicator of the m-th sample respectively; G is the average information entropy of all evaluation indicators; and are the standard deviation and mean of the ith indicator respectively.

2. The intelligent urban and rural earthquake disaster risk assessment method according to claim 1 is characterized by: In step S3, the factor correlation degree measurement is to quantify the contribution of each risk factor to the overall risk; adding the time decay factor , the correlation coefficient is expressed as: ; ; The factor correlation is expressed as: ; ;in, is the correlation coefficient of the ith indicator in the mth sample; is the reference sequence value; is the attenuation coefficient; t is the building age of the sample; is the minimum absolute difference between all factors and the reference sequence; is the maximum absolute difference between all factors and the reference sequence; is the resolution coefficient; set the correlation threshold, and select the indicators whose average correlation coefficient of all samples is higher than the correlation threshold as the final selected features, so as to construct the urban and rural earthquake disaster data set.

3. The intelligent urban and rural earthquake disaster risk assessment method according to claim 2 is characterized by: In step S4, the construction of the urban and rural earthquake disaster risk assessment model specifically includes the following steps: Step S41: Architecture design; an S-layer fully connected neural network is used as the basic architecture; the input layer directly inputs the feature vector constructed based on the urban and rural earthquake disaster data set; the hidden layer designs D fully connected layers, the number of nodes in each layer gradually decreases to capture the nonlinear relationship within the data, and the ReLU activation function is used; Dropout is added after some hidden layers; the output layer uses the Sigmoid activation function to output the predicted disaster risk level; Step S42: Design urban and rural losses; increase density weight factor Ws and introduce dynamic weight factor , the urban and rural loss FC is expressed as: ; ; ; Among them, W and are the current weight and the initial weight respectively; P is the classification parameter; and A are the sample prediction labels and true labels respectively; , and They are the current time, the last evaluation time and the evaluation time interval; is the density adjustment parameter; is the sample density of the region to which the sample belongs; is the maximum sample density; Step S43: multiple extreme loss design; multiple extreme loss FJ is expressed as: ;in, and is the bandwidth parameter; Step S44: a phased strategy; FC is used in the 0th to Uth rounds of training; and FJ is used in the Uth to Eth rounds of training.

4. The intelligent urban and rural earthquake disaster risk assessment method according to claim 3 is characterized by: In step S1, the data construction is to obtain historical urban and rural earthquake disaster assessment data and disaster risk levels; the historical urban and rural earthquake disaster assessment data includes geological exploration data, socio-economic data, building strength data and environmental characteristic data; the disaster risk level is used as a data label; and missing value processing and feature standardization are performed.

5. The intelligent urban and rural earthquake disaster risk assessment method according to claim 4 is characterized by: In step S5, the urban and rural earthquake disaster risk assessment collects urban and rural earthquake disaster assessment data in real time, inputs it into the urban and rural earthquake disaster risk assessment model after feature engineering processing, and uses the disaster risk level output by the model as the urban and rural earthquake disaster risk assessment result.

6. An intelligent urban and rural earthquake disaster risk assessment system, used to implement an intelligent urban and rural earthquake disaster risk assessment method as described in any one of claims 1 to 5, characterized in that: It includes a data construction module, a weight determination module, a factor correlation calculation module, an urban and rural earthquake disaster risk assessment model construction module and an urban and rural earthquake disaster risk assessment module; The data construction module obtains historical urban and rural earthquake disaster assessment data and disaster risk levels, and uses the disaster risk level as a data label; The weight determination module determines the weight of each indicator in the historical urban and rural earthquake disaster assessment data by introducing environmental adaptability factors and coefficients of variation; the indicator is a specific dimension of the historical urban and rural earthquake disaster assessment data; The factor correlation calculation module adds a time decay factor to calculate the factor correlation; quantifies the contribution of each indicator to the overall risk; The urban and rural earthquake disaster risk assessment model construction module constructs a fully connected neural network model, designs an urban and rural loss function and a multiple extreme loss function; takes the urban and rural earthquake disaster assessment data processed by the data construction module and the weight determination module as input, and then constructs an urban and rural earthquake disaster risk assessment model; The urban and rural earthquake disaster risk assessment module performs an urban and rural earthquake disaster risk assessment on the urban and rural earthquake disaster assessment data collected in real time based on the urban and rural earthquake disaster risk assessment model.

Images