Scenic area geological disaster monitoring method and system based on big data

Abstract

The invention provides a scenic spot geological disaster monitoring method and system based on big data, relates to the technical field of geological disaster early warning, and aims to establish a more reliable corresponding relationship between data types and geological disaster types and realize more accurate geological disaster monitoring. Comprising the steps of setting various geological disaster monitoring data and various to-be-predicted geological disaster types; according to the action range of the geological disaster monitoring data obtained based on the big data of the scenic spot and the influence degrees on the occurrence trend of the geological disaster types obtained based on the big data of the multiple regions, establishing a mapping relation table of the geological disaster monitoring data corresponding to one geological disaster type; and predicting each geological disaster type to be predicted through the disaster prediction model. The method has the advantages of high pertinence, high flexibility and high accuracy and reliability in related analysis of geological disasters.

Description

Technical Field

[0001] The present invention relates to the field of geological disaster early warning technology, and in particular to a scenic area geological disaster monitoring method and system based on big data. Background Art

[0002] Geological disaster monitoring is an important means to ensure outdoor safety and reduce accident risks today.

[0003] Especially in scenic spots, where there are usually dense crowds of people, there is a greater need to monitor geological disasters. However, geological disaster monitoring in scenic spots faces some challenges. Geological disaster monitoring usually requires the collection of multiple data, and different types of geological disasters require different data selections. When determining the type of data that needs to be collected for the type of geological disaster, a large amount of data is needed to capture the relationship between the data. However, collecting data for a specific scenic spot to capture data relationships may result in insufficient data volume and inaccurate data analysis results due to the limited size of the scenic spot. On the other hand, if large-scale data collection is not limited to the scenic spot, the limitation of data volume can be solved, but the pertinence to the characteristics of the scenic spot itself will be weakened, which will also reduce the reliability of data analysis. The above defects will affect the accuracy of disaster prediction.

[0004] Therefore, it is necessary to optimize the geological disaster monitoring in scenic areas, establish a more reliable correspondence between data types and geological disaster types, and achieve more accurate geological disaster monitoring. Summary of the invention

[0005] The purpose of the present invention is to provide a scenic area geological disaster monitoring method and system based on big data, which can establish a more reliable correspondence between data types and geological disaster types and achieve more accurate geological disaster monitoring.

[0006] The present invention is achieved through the following technical solutions: A method for monitoring geological disasters in scenic areas based on big data comprises the following steps: Set up a variety of geological disaster monitoring data and a variety of geological disaster types to be predicted; According to the scope of the geological disaster monitoring data obtained based on the big data of the scenic area and the degree of influence on the occurrence trend of geological disaster types obtained based on the big data of multiple regions, the influencing factors are obtained, and a mapping relationship table corresponding to multiple geological disaster monitoring data and one geological disaster type is established; The disaster prediction model is used to make predictions for each type of geological disaster to be predicted.

[0007] Preferably, the method for establishing a mapping relationship between a plurality of geological disaster monitoring data and a geological disaster type comprises the following steps: Respectively obtain the influence factors of each type of geological disaster monitoring data on the geological disaster type, wherein the influence factors include occurrence trend influence factors and action range influence factors, and the larger the value of the occurrence trend influence factor, the greater the influence degree on the occurrence trend, and the larger the value of the action range influence factor, the greater the influence degree on the action range; The mapping relationship is determined based on the influencing factors of each type of geological disaster monitoring data and the geological disaster type.

[0008] Preferably, the impact factor is the product of the occurrence trend impact factor and the scope impact factor: ; in, Representative The geological disaster monitoring data mentioned above is The influencing factors of the above mentioned geological disaster types, Representative The geological disaster monitoring data mentioned above is The occurrence trend influencing factors of the above-mentioned geological disaster types, Representative The geological disaster monitoring data mentioned above is The influencing factors of the scope of action of the above-mentioned geological disaster types.

[0009] Preferably, the method for obtaining the occurrence trend influencing factor is: ; ; in, For the first The geological disaster monitoring data and The correlation coefficient of the occurrence of the above-mentioned geological disaster types is is the total number of types of geological disaster monitoring data, is the natural logarithm function, To find the function of the Pearson correlation coefficient of two parameters, For the first The time series data of the geological disaster monitoring data is as follows: For the first A time series data sequence of the geological disaster type; The method for obtaining the scope influence factor is: ; in, is a natural constant, is the total area of ​​the scenic area, For the The area where the geological disaster monitoring data mentioned above is effective within the scenic area.

[0010] Preferably, the method for determining the mapping relationship table based on the influencing factors of the geological disaster type for each type of geological disaster monitoring data is: Preset impact factor thresholds; Jordi The geological disaster monitoring data mentioned above is If the impact factor of the geological disaster type is greater than the impact factor threshold, the The geological disaster monitoring data is placed in the A mapping relationship table of the types of geological disasters.

[0011] Preferably, the disaster prediction model includes an input layer, a data cleaning layer, a semantic matching layer, a feature extraction layer and a recognition layer; The input layer is used to receive the type of geological disaster to be predicted, multiple geological disaster monitoring data for N consecutive days and the pre-stored mapping relationship table; The function of the data cleaning layer is: if any geological disaster occurs in N consecutive days, the multiple geological disaster monitoring data of the corresponding date will be cleared and the cleared data will be supplemented through the Kalman filter; The functions of the semantic matching layer are: The name of each of the input geological disaster monitoring data is converted into a word vector by using the Word2Vec function; Obtaining the word vector of the geological disaster monitoring data required for prediction according to the type of geological disaster to be predicted through the mapping relationship table; Performing semantic matching through word vectors, selecting the geological disaster monitoring data required for prediction from the input geological disaster monitoring data; The feature extraction layer is used to extract features based on the geological disaster monitoring data required for the prediction to obtain multiple feature values; The recognition layer is used to output the prediction result of the geological disaster based on the activation function after nonlinearizing the characteristic value.

[0012] Preferably, the method for extracting features from the geological disaster monitoring data required for the prediction is: ; ; ; in, for and The eigenvalue array formed is, and The prediction The first of the geological disaster monitoring data required by the geological disaster type a first characteristic value and a second characteristic value of the geological disaster monitoring data; To find the standard deviation function, For N consecutive days The first of the geological disaster monitoring data required by the geological disaster type An array of geological disaster monitoring data, and They are Middle and elements.

[0013] Preferably, the activation function is: ; ; in, For the The output of the activation function of the identification layer of the geological disaster type, Represents the input parameters of the activation function after nonlinearization, Represents the geological disaster monitoring data required for prediction The first The nonlinear output of the eigenvalues, , The total amount of geological disaster monitoring data required for prediction, is a natural constant, and Respectively represent and The geological disaster monitoring data mentioned above is The influencing factors of the types of geological disasters, , and They are the first training weight, the first training bias, and the second training weight respectively.

[0014] Preferably, the first training weight, the first training bias and the second training weight are determined by minimizing The loss function of the geological disaster type Sure: ; ; ; ; in, To determine the threshold, To find the function of the parameter that minimizes the function value, is the total number of training data sets, is the intermediate parameter, Representative The training data is calculated from The value of Representative The training data is calculated from The value of is the truth judgment function, Representative The first set of training data The true value corresponding to the prediction result of the above-mentioned geological disaster type.

[0015] The present invention also provides a scenic area geological disaster monitoring system based on big data, which is applied to the above-mentioned scenic area geological disaster monitoring method based on big data, including: An initialization module is used to set various geological disaster monitoring data and various types of geological disasters to be predicted; A mapping establishment module is used to establish a mapping relationship table corresponding to one geological disaster type based on the scope of the geological disaster monitoring data obtained based on the big data of the scenic area and the degree of influence on the occurrence trend of geological disaster types obtained based on the big data of multiple regions; The prediction module is used to predict each type of geological disaster to be predicted through a disaster prediction model.

[0016] The technical solution of the present invention has at least the following advantages and beneficial effects: The present invention can collect geological disasters and related monitoring data from a large number of regions, and establish a mapping relationship between the geological disasters to be predicted and the corresponding monitoring data in combination with the regional characteristics of the target scenic spot itself. While ensuring the amount of data, it also ensures that the data characteristics are consistent with the characteristics of the scenic spot itself, which helps to improve the accuracy of geological disaster monitoring for specific scenic spots; When establishing a mapping relationship table, the present invention establishes a data trend relationship through a large amount of regional data, and establishes a local data influence degree relationship through the area characteristics of the scenic area. The combination of the two can correct the relationship establishment of a large amount of data under the condition of simple calculation and easy implementation, and has high cost performance. When the present invention predicts geological disasters through the model, different weights are assigned to different types of monitoring data according to different types of geological disasters based on the characteristics of the scenic area itself, thereby improving the pertinence of geological disaster monitoring and thus improving monitoring reliability; When the present invention trains the model, the output model established and the loss function adopted can further optimize the specific coefficients of the weights, which helps to further improve the accuracy of the prediction; The invention is reasonably designed, easy to implement, has strong local area specificity, and is easy to promote and implement in various scenic spots. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A schematic flow chart of a method for monitoring geological disasters in scenic areas based on big data provided in Example 1 of the present invention. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, 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 part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0019] Example 1 This embodiment provides a method for monitoring geological disasters in scenic areas based on big data. Figure 1 , including the following steps: Set up a variety of geological disaster monitoring data and a variety of geological disaster types to be predicted; According to the scope of the geological disaster monitoring data obtained based on the big data of the scenic area and the degree of influence on the occurrence trend of geological disaster types obtained based on the big data of multiple regions, the influencing factors are obtained, and a mapping relationship table corresponding to multiple geological disaster monitoring data and one geological disaster type is established; The disaster prediction model is used to make predictions for each type of geological disaster to be predicted.

[0020] Based on the technical solution of this embodiment, geological disasters and related monitoring data in multiple regions can be widely collected, and the mapping relationship between the type of geological disaster to be predicted and the corresponding geological disaster monitoring data can be established in combination with the specific regional characteristics of the target scenic spot. This process not only ensures the sufficiency of the sample data used to confirm the relationship, but also ensures that the data characteristics are highly consistent with the characteristics of the scenic spot itself, thereby significantly improving the accuracy of geological disaster monitoring for specific scenic spots, helping to more effectively identify potential geological disaster risks in the subsequent identification process, and then providing timely warnings and decision-making support.

[0021] In this embodiment, the method for establishing a mapping relationship between multiple geological disaster monitoring data and one geological disaster type includes the following steps: Respectively obtain the influence factors of each type of geological disaster monitoring data on the geological disaster type, wherein the influence factors include occurrence trend influence factors and action range influence factors, and the larger the value of the occurrence trend influence factor, the greater the influence degree on the occurrence trend, and the larger the value of the action range influence factor, the greater the influence degree on the action range; The mapping relationship is determined based on the influencing factors of each type of geological disaster monitoring data and the geological disaster type.

[0022] As a preferred solution of this embodiment, the impact factor is the product of the occurrence trend impact factor and the scope impact factor: ; in, Representative The geological disaster monitoring data mentioned above is The influencing factors of the above mentioned geological disaster types, Representative The geological disaster monitoring data mentioned above is The occurrence trend influencing factors of the above-mentioned geological disaster types, Representative The geological disaster monitoring data mentioned above is The influencing factors of the scope of action of the above-mentioned geological disaster types.

[0023] Specifically, the method for obtaining the occurrence trend influencing factor is: ; ; in, For the first The geological disaster monitoring data and The correlation coefficient of the occurrence of the above-mentioned geological disaster types is is the total number of types of geological disaster monitoring data, is the natural logarithm function, To find the function of the Pearson correlation coefficient of two parameters, For the first The time series data of the geological disaster monitoring data is as follows: For the first A time series data sequence of the geological disaster type; On the other hand, the method for obtaining the scope influence factor is: ; in, is a natural constant, is the total area of ​​the scenic area, For the The area where the geological disaster monitoring data mentioned above is effective within the scenic area.

[0024] On this basis, the method for determining the mapping relationship table according to the influencing factors of the geological disaster type for each type of geological disaster monitoring data is: Preset impact factor thresholds; Jordi The geological disaster monitoring data mentioned above is If the impact factor of the geological disaster type is greater than the impact factor threshold, the The geological disaster monitoring data is placed in the A mapping relationship table of the types of geological disasters.

[0025] When establishing a mapping relationship table, this embodiment can use data from a large number of regions to identify and establish data trend relationships. By analyzing the area characteristics of the scenic area, the relationship between local data and the degree of disaster impact is constructed, that is, a larger area of ​​action is given a higher correction factor. For example, the parameters can include rainfall, water level, etc. The rainfall action area can be set to the entire field, and the water level is for lakes, streams and other areas. By combining the two types of data, it can be kept simple and easy in calculation, and the relationship between a large amount of data can be effectively corrected to achieve a cost-effective monitoring effect. This method not only improves the efficiency of data processing, but also ensures the reliability of the results, so that the monitoring system can still provide accurate geological disaster assessment under limited resources. As a specific case, when warning against floods, the water level is generally an important indicator, but if the area of ​​the lakes in the scenic area is very small, the range impact factor is very small, so it can be screened out in the calculation of the impact factor and is more targeted to the scenic area, while in another scenic area dominated by lakes, it will have a very high range impact factor. It is particularly noted that when obtaining the occurrence trend impact factor, the first The time series data sequence of the geological disaster type can be a true value, that is, 1 if the corresponding geological disaster occurs and 0 if it does not occur, or it can be a specific parameter value, for example, for an earthquake, the earthquake level is recorded, and 0 if it does not occur.

[0026] In the following steps, the disaster prediction model includes an input layer, a data cleaning layer, a semantic matching layer, a feature extraction layer, and a recognition layer; The input layer is used to receive the type of geological disaster to be predicted, multiple geological disaster monitoring data for N consecutive days and the pre-stored mapping relationship table; The function of the data cleaning layer is: if any geological disaster occurs in N consecutive days, the multiple geological disaster monitoring data of the corresponding date will be cleared and the cleared data will be supplemented through the Kalman filter; The functions of the semantic matching layer are: The name of each of the input geological disaster monitoring data is converted into a word vector by using the Word2Vec function; Obtaining the word vector of the geological disaster monitoring data required for prediction according to the type of geological disaster to be predicted through the mapping relationship table; Performing semantic matching through word vectors, selecting the geological disaster monitoring data required for prediction from the input geological disaster monitoring data; The feature extraction layer is used to extract features based on the geological disaster monitoring data required for the prediction to obtain multiple feature values; The recognition layer is used to output the prediction result of the geological disaster based on the activation function after nonlinearizing the characteristic value.

[0027] The semantic matching layer set in this embodiment can utilize the semantic features of word vectors, which can better understand and match the relationship between geological disaster monitoring data and the type of geological disaster to be predicted, and thus has strong adaptability and versatility. When faced with measurement data from different monitoring equipment or personnel, the required data can be accurately matched even if the names have certain deviations.

[0028] Furthermore, the method for extracting features from the geological disaster monitoring data required for the prediction is: ; ; ; in, for and The eigenvalue array formed is, and The prediction The first of the geological disaster monitoring data required by the geological disaster type a first characteristic value and a second characteristic value of the geological disaster monitoring data; To find the standard deviation function, For N consecutive days The first of the geological disaster monitoring data required by the geological disaster type An array of geological disaster monitoring data, and They are Middle and elements.

[0029] When performing output judgment, the activation function is: ; ; in, For the The output of the activation function of the identification layer of the geological disaster type, Represents the input parameters of the activation function after nonlinearization, Represents the geological disaster monitoring data required for prediction The first The nonlinear output of the eigenvalues, , The total amount of geological disaster monitoring data required for prediction, is a natural constant, and Respectively represent and The geological disaster monitoring data mentioned above is The influencing factors of the types of geological disasters, , and They are the first training weight, the first training bias, and the second training weight respectively.

[0030] The method for determining the first training weight, the first training bias and the second training weight is preferably to minimize the first The loss function of the geological disaster type Sure: ; ; ; ; in, To determine the threshold, To find the function of the parameter that minimizes the function value, is the total number of training data sets, is the intermediate parameter, Representative The training data is calculated from The value of Representative The training data is calculated from The value of is the truth judgment function, Representative The first set of training data The true value corresponding to the prediction result of the geological disaster type is specifically expressed as a true value, that is, 1 is recorded when it occurs and 0 is recorded when it does not occur.

[0031] In the process of constructing the geological disaster prediction model, this embodiment assigns different weights to different types of geological disaster monitoring data based on the previously calculated influencing factors. This strategy makes monitoring more targeted, thereby enhancing the reliability of geological disaster monitoring. By reasonably allocating weights, the model can more accurately reflect the actual risks of various types of disasters, thereby improving the responsiveness of the overall monitoring system. In the model training process, an appropriate loss function is used to further optimize the specific coefficients of the weights. The loss function involves Partial optimization excludes the influence of weight changes and involves The part takes into account the impact of weight changes. The comprehensive establishment of loss function helps to improve the accuracy of prediction and ensure that the model can adapt to the changing environment and data characteristics.

[0032] Example 2 This embodiment provides a scenic area geological disaster monitoring system based on big data, and a scenic area geological disaster monitoring method based on big data applied to the above embodiment, including: An initialization module is used to set various geological disaster monitoring data and various types of geological disasters to be predicted; A mapping establishment module is used to establish a mapping relationship table corresponding to one geological disaster type based on the scope of the geological disaster monitoring data obtained based on the big data of the scenic area and the degree of influence on the occurrence trend of geological disaster types obtained based on the big data of multiple regions; The prediction module is used to predict each type of geological disaster to be predicted through a disaster prediction model.

[0033] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for monitoring geological disasters in scenic areas based on big data, characterized in that: The following steps are involved: Set up a variety of geological disaster monitoring data and a variety of geological disaster types to be predicted; According to the scope of the geological disaster monitoring data obtained based on the big data of the scenic area and the degree of influence on the occurrence trend of geological disaster types obtained based on the big data of multiple regions, the influencing factors are obtained, and a mapping relationship table corresponding to multiple geological disaster monitoring data and one geological disaster type is established; The disaster prediction model is used to make predictions for each type of geological disaster to be predicted.

2. A method for monitoring geological disasters in scenic areas based on big data according to claim 1, characterized in that: The method for establishing a mapping relationship between multiple geological disaster monitoring data and one geological disaster type comprises the following steps: Respectively obtain the influence factors of each type of geological disaster monitoring data on the geological disaster type, wherein the influence factors include occurrence trend influence factors and action range influence factors, and the larger the value of the occurrence trend influence factor, the greater the influence degree on the occurrence trend, and the larger the value of the action range influence factor, the greater the influence degree on the action range; The mapping relationship is determined based on the influencing factors of each type of geological disaster monitoring data and the geological disaster type.

3. A method for monitoring geological disasters in scenic areas based on big data according to claim 2, characterized in that: The impact factor is the product of the occurrence trend impact factor and the scope impact factor: ; in, Representative The geological disaster monitoring data mentioned above is The influencing factors of the above mentioned geological disaster types, Representative The geological disaster monitoring data mentioned above is The occurrence trend influencing factors of the above-mentioned geological disaster types, Representative The geological disaster monitoring data mentioned above is The influencing factors of the scope of action of the above-mentioned geological disaster types.

4. A method for monitoring geological disasters in scenic areas based on big data according to claim 3, characterized in that: The method for obtaining the occurrence trend influencing factor is: ; ; in, For the first The geological disaster monitoring data and The correlation coefficient of the occurrence of the above-mentioned geological disaster types is is the total number of types of geological disaster monitoring data, is the natural logarithm function, To find the function of the Pearson correlation coefficient of two parameters, For the first The time series data of the geological disaster monitoring data is as follows: For the first A time series data sequence of the geological disaster type; The method for obtaining the scope influence factor is: ； in, is a natural constant, is the total area of ​​the scenic area, For the The area where the geological disaster monitoring data mentioned above is effective within the scenic area.

5. A method for monitoring geological disasters in scenic areas based on big data according to claim 4, characterized in that: The method for determining the mapping relationship table based on the influencing factors of the geological disaster type according to each type of geological disaster monitoring data is: Preset impact factor thresholds; Jordi The geological disaster monitoring data mentioned above is If the impact factor of the geological disaster type is greater than the impact factor threshold, the The geological disaster monitoring data is placed in the A mapping relationship table of the types of geological disasters.

6. A method for monitoring geological disasters in scenic areas based on big data according to claim 1, characterized in that: The disaster prediction model includes an input layer, a data cleaning layer, a semantic matching layer, a feature extraction layer and a recognition layer; The input layer is used to receive the type of geological disaster to be predicted, multiple geological disaster monitoring data for N consecutive days and the pre-stored mapping relationship table; The function of the data cleaning layer is: if any geological disaster occurs in N consecutive days, the multiple geological disaster monitoring data of the corresponding date will be cleared and the cleared data will be supplemented through the Kalman filter; The functions of the semantic matching layer are: The name of each of the input geological disaster monitoring data is converted into a word vector by using the Word2Vec function; Obtaining the word vector of the geological disaster monitoring data required for prediction according to the type of geological disaster to be predicted through the mapping relationship table; Performing semantic matching through word vectors, selecting the geological disaster monitoring data required for prediction from the input geological disaster monitoring data; The feature extraction layer is used to extract features based on the geological disaster monitoring data required for the prediction to obtain multiple feature values; The recognition layer is used to output the prediction result of the geological disaster based on the activation function after nonlinearizing the characteristic value.

7. A method for monitoring geological disasters in scenic areas based on big data according to claim 6, characterized in that: The method for extracting features from the geological disaster monitoring data required for the prediction is: ; ； ； in, for and The eigenvalue array formed is, and The prediction The first of the geological disaster monitoring data required by the geological disaster type a first characteristic value and a second characteristic value of the geological disaster monitoring data; To find the standard deviation function, For N consecutive days The first of the geological disaster monitoring data required by the geological disaster type An array of geological disaster monitoring data, and They are Middle and elements.

8. A method for monitoring geological disasters in scenic areas based on big data according to claim 7, characterized in that: The activation function is: ； ; in, For the The output of the activation function of the identification layer of the geological disaster type, Represents the input parameters of the activation function after nonlinearization, Represents the geological disaster monitoring data required for prediction The first The nonlinear output of the eigenvalues, , The total amount of geological disaster monitoring data required for prediction, is a natural constant, and Respectively represent and The geological disaster monitoring data mentioned above is The influencing factors of the types of geological disasters, , and They are the first training weight, the first training bias, and the second training weight respectively.

9. A method for monitoring geological disasters in scenic areas based on big data according to claim 8, characterized in that: The first training weight, the first training bias and the second training weight are determined by minimizing The loss function of the geological disaster type Sure: ; ； ; ; in, To determine the threshold, To find the function of the parameter that minimizes the function value, is the total number of training data sets, is the intermediate parameter, Representative The training data is calculated from The value of Representative The training data is calculated from The value of is the truth judgment function, Representative The first set of training data The true value corresponding to the prediction result of the above-mentioned geological disaster type.

10. A scenic area geological disaster monitoring system based on big data, applied to a scenic area geological disaster monitoring method based on big data as claimed in any one of claims 1 to 9, characterized in that: include: An initialization module is used to set various geological disaster monitoring data and various types of geological disasters to be predicted; A mapping establishment module is used to establish a mapping relationship table corresponding to one geological disaster type based on the scope of the geological disaster monitoring data obtained based on the big data of the scenic area and the degree of influence on the occurrence trend of geological disaster types obtained based on the big data of multiple regions; The prediction module is used to predict each type of geological disaster to be predicted through a disaster prediction model.

Images