Earthquake disaster prediction method based on iron tower monitoring data

Abstract

The invention provides an earthquake disaster prediction method based on iron tower monitoring data, and relates to the technical field of disaster prediction, and the method comprises the steps: obtaining a monitoring data set, extracting an earthquake region, extracting earthquake time node data, constructing an earthquake iron tower reference set, carrying out the model training, obtaining an earthquake disaster prediction model, and obtaining a data prediction structure. And carrying out data standardization processing on the iron tower monitoring data, inputting a processing result into the earthquake disaster prediction model to obtain a prediction result, and finally carrying out disaster relief strategy guidance feedback. The method mainly solves the problems that the occurrence time, place and magnitude of an earthquake cannot be accurately predicted, the prediction time range is short, and the long-term prediction requirement cannot be met. Through earthquake monitoring on the iron tower, potential safety hazards can be found and solved in time, earthquake disasters can be dealt with better, and casualties and economic losses are reduced.

Description

Technical Field

[0001] The present application relates to the technical field of disaster prediction, and in particular to a method for predicting earthquake disasters based on tower monitoring data. Background Art

[0002] The earthquake disaster prediction method based on tower monitoring data is a method for predicting earthquake disasters by using tower monitoring data. A tower is a tall building, usually used for communication, power transmission and other functions. Due to the height and structural characteristics of the tower, it becomes an important monitoring point for earthquake activity. By monitoring the earthquake response data of the tower, relevant information about earthquake activity can be obtained, which can then be used to predict earthquake disasters. When an earthquake occurs, seismic waves will propagate outward in the form of waves, and the tower can serve as a monitoring point to collect and record information about the seismic waves.

[0003] The traditional method is to predict earthquakes by observing some abnormal phenomena that occur before an earthquake, such as surface deformation, changes in the geomagnetic field, changes in groundwater levels, etc. These changes may be signals before an earthquake occurs.

[0004] However, the inventors of this application found that the above technology has at least the following technical problems in the process of implementing the technical solution of the invention in the embodiment of this application:

[0005] There are limitations in prediction accuracy and time range. It is impossible to accurately predict the time, location and magnitude of an earthquake. The prediction time range is also relatively short and cannot meet long-term prediction needs. Summary of the invention

[0006] This application mainly solves the problem that it is impossible to accurately predict the time, location and magnitude of an earthquake, and the prediction time range is relatively short, which cannot meet long-term prediction needs.

[0007] In view of the above problems, the present application provides an earthquake disaster prediction method based on tower monitoring data. In the first aspect, the present application provides an earthquake disaster prediction method based on tower monitoring data, the method comprising: connecting to a tower big data system to obtain a monitoring data set; extracting earthquake area and earthquake time node data from the monitoring data set, and constructing an earthquake tower reference set in combination with earthquake disaster record data; using the earthquake tower reference set to perform model training to obtain an earthquake disaster prediction model; obtaining a data prediction structure; based on the data prediction structure, performing data standardization processing on the tower monitoring data at that time; inputting the standardized tower monitoring data at that time into the earthquake disaster prediction model to perform disaster prediction, obtain an earthquake disaster prediction result, and provide disaster relief strategy guidance feedback based on the earthquake disaster prediction result.

[0008] In the second aspect, the present application provides an earthquake disaster prediction system based on tower monitoring data, the system comprising: a monitoring data set acquisition module, the monitoring data set acquisition module is used to connect to the tower big data system to acquire a monitoring data set; a tower reference set construction module, the tower reference set construction module is used to extract earthquake area and earthquake time node data from the monitoring data set, and construct an earthquake tower reference set in combination with earthquake disaster record data; an earthquake disaster prediction model acquisition module, the earthquake disaster prediction model acquisition module is used to use the earthquake tower reference set to perform model training to obtain an earthquake disaster prediction model; a prediction structure acquisition module, the prediction structure acquisition module is used to acquire a data prediction structure; a standardization processing module, the standardization processing module is based on the data prediction structure, and performs data standardization processing on the tower monitoring data of the current time; a disaster relief feedback module, the disaster relief feedback module is used to input the standardized tower monitoring data of the current time into the earthquake disaster prediction model for disaster prediction, obtain an earthquake disaster prediction result, and provide disaster relief strategy guidance feedback based on the earthquake disaster prediction result.

[0009] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0010] The present application provides an earthquake disaster prediction method based on tower monitoring data, which relates to the technical field of disaster prediction. The method comprises: obtaining a monitoring data set, extracting earthquake area and earthquake time node data, constructing an earthquake tower reference set, performing model training to obtain an earthquake disaster prediction model, obtaining a data prediction structure, performing data standardization processing on the tower monitoring data, inputting the processing results into the earthquake disaster prediction model to obtain prediction results, and finally providing guidance and feedback on disaster relief strategies.

[0011] This application mainly solves the problem that it is impossible to accurately predict the time, location and magnitude of an earthquake, and the prediction time range is relatively short, which cannot meet the long-term prediction needs. Through earthquake monitoring of iron towers, potential safety hazards can be discovered and resolved in a timely manner, and earthquake disasters can be better responded to, reducing casualties and economic losses.

[0012] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0014] Figure 1 A schematic diagram of a method for predicting earthquake disasters based on tower monitoring data is provided for an embodiment of the present application;

[0015] Figure 2 A schematic flow chart of a method for obtaining an earthquake disaster prediction model in an earthquake disaster prediction method based on tower monitoring data is provided for an embodiment of the present application;

[0016] Figure 3 A schematic flow chart of a method for obtaining a training data set and a test data set in an earthquake disaster prediction method based on tower monitoring data is provided for an embodiment of the present application;

[0017] Figure 4 A structural schematic diagram of an earthquake disaster prediction system based on tower monitoring data is provided for an embodiment of the present application.

[0018] Explanation of reference numerals: monitoring data set acquisition module 10 , tower reference set construction module 20 , earthquake disaster prediction model acquisition module 30 , prediction structure acquisition module 40 , standardization processing module 50 , disaster relief feedback module 60 . DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0020] This application mainly solves the problem that it is impossible to accurately predict the time, location and magnitude of an earthquake, and the prediction time range is relatively short, which cannot meet the long-term prediction needs. Through earthquake monitoring of iron towers, potential safety hazards can be discovered and resolved in a timely manner, and earthquake disasters can be better responded to, reducing casualties and economic losses.

[0021] In order to better understand the above technical solution, the above solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods:

[0022] Embodiment 1

[0023] like Figure 1The earthquake disaster prediction method based on tower monitoring data is shown, and the method comprises:

[0024] Connect to the tower big data system to obtain monitoring data sets;

[0025] Specifically, connect to the tower big data system to obtain monitoring data sets and determine data requirements: First, it is necessary to clarify what types of monitoring data need to be obtained, such as seismic wave data, wind speed data, temperature data, etc. At the same time, it is also necessary to determine the time range, spatial range and other requirements of the data. Understand the data interface of the tower big data system: The tower big data system usually provides a data interface for external systems to obtain data. It is necessary to understand the specifications and standards of the data interface, including API interface, data format, authentication method, etc. Develop data acquisition program: According to the data interface specification of the tower big data system, develop a data acquisition program. The program usually needs to include modules such as data request, data parsing and data storage. Testing and debugging: After development is completed, testing and debugging are required to ensure that the program can correctly obtain and parse data. Deployment and operation: After testing and debugging, the program is deployed to the production environment and kept running to obtain monitoring data in real time. Analyze and manage data: For the acquired monitoring data, further analysis and management are required. For example, data cleaning, data conversion, data visualization and other operations are performed to better understand and use the data.

[0026] Extracting earthquake area and earthquake time node data from the monitoring data set, and building an earthquake tower reference set in combination with earthquake disaster record data;

[0027] Specifically, the acquired monitoring data set is used to extract earthquake region and earthquake time node data. Data cleaning: The acquired monitoring data set is cleaned to remove invalid, erroneous or duplicate data. Data screening: Extract earthquake-related data such as earthquake wave data and wind speed data as needed. Earthquake region extraction: Extract earthquake region information from the screened data, such as the longitude, latitude, and epicenter of the earthquake. Earthquake time node extraction: Extract earthquake time information from the screened data, such as the timestamp of the earthquake, focal depth, etc. After extracting the earthquake region and earthquake time node data, the earthquake tower reference set can be constructed in combination with the earthquake disaster record data. Collect earthquake disaster record data: Collect historical earthquake disaster record data, including information such as the region, time, and magnitude of the earthquake. Data matching: Match the extracted earthquake region and time node data with the earthquake disaster record data to find the earthquake disaster record corresponding to the extracted data. Construct earthquake tower reference set: Combine the matched earthquake disaster record data with the tower monitoring data to construct a set containing earthquake tower reference data. By constructing a seismic tower reference set, we can obtain the response data of the tower in the seismic active area, which can be used to analyze the performance and safety assessment of the tower in earthquake disasters. At the same time, this reference set can also provide data support for other related research, such as the development of earthquake early warning systems and the impact of earthquakes on social economy.

[0028] Using the earthquake tower reference set to perform model training to obtain an earthquake disaster prediction model;

[0029] Specifically, the constructed earthquake tower reference set is used for model training to obtain an earthquake disaster prediction model, which can be carried out using machine learning or deep learning methods. Data preprocessing: The earthquake tower reference set is preprocessed, including data cleaning, feature extraction, feature selection and other operations to prepare data for model training. Model selection: According to the nature of the problem and the characteristics of the data, a suitable machine learning or deep learning model is selected. For example, models such as support vector machine (SVM), random forest, neural network, etc. can be considered. Model training: Use the selected model to train with the earthquake tower reference set. During the training process, the parameters of the model need to be adjusted to optimize the performance of the model. Model evaluation: Use an independent test set to evaluate the trained model to determine the prediction ability and accuracy of the model. Different evaluation indicators can be used, such as accuracy, recall rate, F1 score, etc. Model optimization: According to the evaluation results, the model is optimized and adjusted to improve the prediction ability and accuracy of the model. The optimization measures that can be taken include adjusting model parameters, adding or reducing features, and improving model structure. Model application: Apply the trained model to actual earthquake monitoring data to achieve earthquake disaster prediction. The model can be integrated into the tower monitoring system, or a standalone application can be developed to use the model.

[0030] Get data prediction structure;

[0031] Specifically, obtain the data prediction structure and determine the prediction object and target: clarify the objects and targets to be predicted, such as predicting the time, location, and magnitude of earthquake disasters. Collect data: collect historical data and real-time data related to the prediction object and target. These data can come from different sources, such as tower monitoring data, earthquake record data, meteorological data, etc. Data preprocessing: clean, organize and convert the collected data to prepare for model training and prediction. Model selection and training: select a suitable prediction model, such as a time series prediction model, a causal relationship prediction model, a simulation model, etc., and use historical data to train the model. Prediction and result analysis: apply the trained model to real-time data, make predictions, and analyze the prediction results. According to the prediction results, corresponding measures can be taken, such as disaster prevention in advance and adjustment of resource allocation.

[0032] Based on the data prediction structure, data standardization processing is performed on the tower monitoring data at that time;

[0033] Specifically, based on the data prediction structure, the tower monitoring data can be standardized to better adapt to the model training and prediction process. Data standardization refers to converting data into a unified scale to avoid the impact of data of different scales on model training and prediction. Minimum-maximum normalization: Map the data to the range of [0,1], convert the maximum and minimum values ​​of the original data to 0 and 1 respectively, and scale other values ​​according to the ratio. Handling outliers: For the presence of outliers, appropriate processing is required, such as removing or replacing outliers, to avoid negative impact on model training and prediction. Handling missing values: If there are missing values ​​in the data, appropriate processing is required, such as filling missing values ​​or deleting data containing missing values, to avoid impact on model training and prediction. Handling duplicate values: If there are duplicate values ​​in the data, appropriate processing is required, such as removing or merging duplicate values, to avoid impact on model training and prediction. Data conversion: If the data does not meet the input requirements of the model, appropriate data conversion is required, such as unique hot encoding of categorical variables, to avoid impact on model training and prediction.

[0034] The standardized tower monitoring data is input into the earthquake disaster prediction model to perform disaster prediction, and an earthquake disaster prediction result is obtained. Based on the earthquake disaster prediction result, guidance and feedback of disaster relief strategies are performed.

[0035] Specifically, the standardized tower monitoring data is input into the earthquake disaster prediction model for disaster prediction, and the earthquake disaster prediction results are obtained. The disaster relief strategy guidance feedback is provided based on the earthquake disaster prediction results. Data standardization: The tower monitoring data is processed using the data standardization method, and the data is converted into a unified scale to avoid the impact of data of different scales on model training and prediction. Model input: The standardized tower monitoring data is input into the earthquake disaster prediction model. Disaster prediction: The input data is used for disaster prediction to obtain earthquake disaster prediction results. Disaster relief strategy guidance feedback: Based on the earthquake disaster prediction results, the disaster relief strategy guidance feedback is provided. The earthquake disaster prediction results are analyzed to assess the possible disaster impact scope, degree and affected population. According to the analysis results, the corresponding disaster relief strategy is formulated, including emergency response, resource allocation, personnel evacuation and other aspects. The disaster relief strategy is adjusted and optimized according to the actual situation to ensure the rationality and effectiveness of the strategy. The formulated disaster relief strategy is implemented, and the implementation effect is monitored and evaluated at the same time, so as to make timely adjustments and improvements. The entire disaster relief process is recorded and summarized, and the problems and shortcomings are analyzed to provide experience and reference for future disaster relief work. By inputting the standardized tower monitoring data into the earthquake disaster prediction model for disaster prediction, the earthquake disaster prediction results are obtained, and the disaster relief strategy guidance feedback is provided based on the results, which can better respond to earthquake disasters and reduce casualties and economic losses.

[0036] Furthermore, if Figure 2 As shown, the method of the present application uses the earthquake tower reference set to perform model training to obtain an earthquake disaster prediction model, including:

[0037] Preprocessing the seismic tower reference set to obtain a multi-dimensional target data set;

[0038] Performing data annotation based on the multi-dimensional target data set, dividing the annotated multi-dimensional target data set according to a preset ratio, and constructing a training data set and a test data set;

[0039] A neural network architecture is constructed, and the training data set and the test data set are used to train and test the neural network architecture to obtain the earthquake disaster prediction model.

[0040] Specifically, the earthquake tower reference set is preprocessed to obtain a multi-dimensional target data set: the earthquake tower reference set is subjected to data cleaning, data transformation, feature selection and other operations to prepare a multi-dimensional target data set for model training. The target data set should contain various dimensional data related to earthquake disasters, such as the time, location, and magnitude of the earthquake, as well as the geographical location, structural characteristics, and historical earthquake response data of the tower. Data annotation is performed based on the multi-dimensional target data set: data annotation refers to assigning a label or category to each sample in the data set, and this label or category is usually a text description that can be understood by humans. In this step, a label related to earthquake disasters is assigned to each sample in the multi-dimensional target data set, such as "earthquake occurred" or "no earthquake occurred". The annotated multi-dimensional target data set is split according to a preset ratio to construct a training data set and a test data set: Generally, splitting the annotated data set into a training data set and a test data set is a standard step in the process of machine learning and deep learning. The training data set is used to train the model, and the test data set is used to evaluate the performance of the model. In this step, the annotated multi-dimensional target data set is split into a training data set and a test data set according to a preset ratio. Build a neural network architecture: A neural network is a computing model that simulates the way neurons in the human brain are connected. It consists of nodes (neurons) and edges (weights) connecting these nodes. Design a neural network architecture that can receive multi-dimensional input data and learn and predict based on these input data. Use the training data set and test data set to train and test the neural network architecture: Training refers to using the training data set to adjust the weights in the neural network to minimize prediction errors. Testing refers to using the test data set to evaluate the performance of the model. In this step, the training data set is used to train the neural network architecture, and the test data set is used to evaluate the performance of the model. Obtain the earthquake disaster prediction model: Through the above steps, a trained and tested neural network model will be obtained, which can be used to predict earthquake disasters. The model can receive tower monitoring data as input and output earthquake disaster prediction results.

[0041] Furthermore, the method of the present application preprocesses the seismic tower reference set to obtain a multi-dimensional target data set, including:

[0042] Classifying the data of the earthquake tower reference set to obtain vibration monitoring data, signal monitoring data, and earthquake disaster record data;

[0043] Perform vibration direction and vibration amplitude analysis based on the vibration monitoring data to determine vibration dimension data;

[0044] Based on the signal monitoring data, regional coverage, weighted offline level, and number of tower offline analyses are performed to determine signal dimension data;

[0045] Perform disaster level assessment based on the earthquake disaster record data to determine disaster assessment dimension data;

[0046] The vibration dimension data, signal dimension data, and disaster assessment dimension data are aligned by region and time to construct the multi-dimensional target data set.

[0047] Specifically, the data in the earthquake tower reference set is classified to obtain vibration monitoring data, signal monitoring data, and earthquake disaster record data: according to the nature and needs of the data, the data in the earthquake tower reference set is classified to obtain vibration monitoring data, signal monitoring data, and earthquake disaster record data. Vibration monitoring data usually includes time series data of tower vibration caused by earthquakes, signal monitoring data may include signal coverage and signal quality of the tower, and earthquake disaster record data includes information such as the time, location, and magnitude of each earthquake. Based on the vibration monitoring data, vibration direction and vibration amplitude are analyzed to determine vibration dimension data: the vibration monitoring data is deeply analyzed to extract information related to vibration direction and vibration amplitude. The vibration direction can reflect the propagation direction of seismic waves, and the vibration amplitude can reflect the severity of the earthquake. This information is integrated into vibration dimension data for subsequent model training and analysis. Based on the signal monitoring data, regional coverage, weighted offline level, and number of tower offline are analyzed to determine signal dimension data: the signal monitoring data is processed and analyzed to extract information related to regional coverage, weighted offline level, and number of tower offline. The regional coverage rate can reflect the coverage range of the signal, the weighted offline level can reflect the quality of the signal, and the number of towers offline can reflect the working condition of the tower. This information is integrated into signal dimension data for subsequent model training and analysis. Disaster level assessment is performed based on the earthquake disaster record data to determine the disaster assessment dimension data: the earthquake disaster record data is used to perform disaster level assessment to obtain the disaster level of each earthquake event. The disaster level assessment can be comprehensively evaluated based on factors such as magnitude, affected population, and economic losses. The assessment results are integrated into disaster assessment dimension data for subsequent model training and analysis. The vibration dimension data, signal dimension data, and disaster assessment dimension data are aligned in region and time to construct the multi-dimensional target data set: in the multi-dimensional target data set, the vibration dimension data, signal dimension data, and disaster assessment dimension data need to be aligned in region and time. This can be achieved by constructing a common index containing time, location, and other related attributes. After the alignment is completed, the data of the three dimensions are merged into a multi-dimensional target data set for subsequent model training and testing. Through the above process, different types of dimension data can be obtained and processed from the earthquake tower reference set, thereby constructing a multi-dimensional target data set for earthquake disaster prediction. This multi-dimensional target dataset will provide comprehensive input information for the neural network model, helping to improve the accuracy and reliability of earthquake disaster prediction.

[0048] Furthermore, if Figure 3 As shown, the method of the present application performs data annotation based on the multi-dimensional target data set, divides the annotated multi-dimensional target data set according to a preset ratio, and constructs a training data set and a test data set, including:

[0049] Setting labels based on the disaster assessment dimensional data, and labeling the multi-dimensional target data set;

[0050] Constructing a sample data structure according to the mapping relationship of the labeled data in the multi-dimensional target data set, wherein the sample data structure includes identification information of vibration dimension data, signal dimension data and their corresponding disaster assessment dimension data;

[0051] The multi-dimensional target data set after the sample data structure processing is segmented according to a preset ratio of 8:2 to obtain the training data set and the test data set.

[0052] Specifically, the multidimensional target data set is labeled based on the disaster assessment dimension data, and a sample data structure is constructed, and then the data set is divided into a training data set and a test data set according to a preset ratio. The label is set based on the disaster assessment dimension data, and the multidimensional target data set is labeled: a label is set for each sample in the multidimensional target data set according to the disaster assessment dimension data. For example, corresponding labels can be set according to the disaster level assessment results, such as no disaster, low disaster, moderate disaster, severe disaster, etc. According to the mapping relationship of the labeled data in the multidimensional target data set, a sample data structure is constructed: the mapping relationship between the labeled data and the original data is determined, and the sample data structure is constructed based on this. The sample data structure should include vibration dimension data, signal dimension data, and identification information of the corresponding disaster assessment dimension data. In this way, each sample contains data in three dimensions of vibration, signal and disaster assessment, and each dimension of data has its corresponding identification information. The multidimensional target data set after the sample data structure processing is divided according to the preset ratio of 8:2 to obtain the training data set and the test data set: the multidimensional target data set after the sample data structure processing is divided into a training data set and a test data set according to the ratio of 8:2. Among them, 80% of the data is used to train the model, and 20% of the data is used to test the performance of the model. Through the above process, a labeled multi-dimensional target data set can be obtained, and a sample data structure containing three dimensions of vibration, signal, and disaster assessment can be constructed. Then, the data set is divided into a training data set and a test data set according to a preset ratio for subsequent neural network model training and testing.

[0053] Furthermore, the method of the present application performs data standardization processing on the tower monitoring data at that time based on the data prediction structure, including:

[0054] Obtain the current earthquake requirements to be analyzed, including earthquake area and earthquake time;

[0055] Based on the earthquake area and earthquake time, connect to the tower big data system to obtain the tower monitoring data of that time;

[0056] According to the data requirements of the vibration dimension data and the signal dimension data in the data prediction structure, the tower monitoring data at that time is classified and calculated to complete the standardized processing of the data prediction structure.

[0057] Specifically, the current earthquake requirements to be analyzed are obtained, and then the corresponding tower monitoring data are obtained from the tower big data system based on these requirements, and the data are classified, calculated and standardized according to the requirements of the data prediction structure. Obtain the current earthquake requirements to be analyzed: obtain the current earthquake requirements to be analyzed in some way, including the area and time of the earthquake. This information may come from the earthquake early warning system, user reports or other monitoring equipment. Connect the tower big data system to obtain the tower monitoring data at that time based on the earthquake area and earthquake time: connect to the tower big data system according to the obtained earthquake area and time, and obtain the tower monitoring data related to the earthquake event at that time. These data may include vibration data, signal data, etc. of the tower. Classify and calculate the tower monitoring data at that time according to the data requirements of the vibration dimension data and signal dimension data in the data prediction structure: classify and calculate the tower monitoring data at that time according to the defined data prediction structure. For example, for vibration dimension data, it may be necessary to calculate the direction and amplitude of vibration; for signal dimension data, it may be necessary to analyze the coverage and quality of the signal. Complete the standardized processing of the data prediction structure: standardize the classified and calculated data according to the requirements of the data prediction structure. This includes operations such as scaling, transforming or processing outliers to ensure that the data meets the requirements of model training and prediction. Through the above process, the tower monitoring data of the current earthquake to be analyzed can be obtained and processed, thereby providing input data for the subsequent earthquake disaster prediction model.

[0058] Furthermore, the present application method also includes:

[0059] Obtain the current earthquake region, use the current earthquake region and the earthquake tower reference set to perform regional matching, and extract the record data set in the same region;

[0060] Perform data feature analysis based on the tower monitoring data, and use the data features to perform traversal comparison with the earthquake tower reference set to obtain a monitoring data set of the same type;

[0061] Based on the same-region record data set and the same-type monitoring data set, respectively predict the tower monitoring data of the current time to obtain a first prediction result and a second prediction result;

[0062] Perform loss value settlement based on the first prediction result and the earthquake disaster prediction result to obtain a first loss value;

[0063] Perform loss value settlement based on the second prediction result and the earthquake disaster prediction result to obtain a second loss value;

[0064] The first loss value and the second loss value are fused and calculated according to preset weights to determine a deviation adjustment coefficient, and the earthquake disaster prediction result is adjusted using the deviation adjustment coefficient to generate an auxiliary earthquake disaster prediction result. The auxiliary earthquake disaster prediction result is merged with the earthquake disaster prediction result for guidance and feedback.

[0065] Specifically, a method for assisting in the prediction of earthquake disasters using tower monitoring data and other related data. Obtain the current earthquake region: obtain the regional information of the current earthquake by means of big data. Use the current earthquake region to match the earthquake tower reference set to extract the same region record data set: match the region in the earthquake tower reference set according to the obtained current earthquake region, and extract the historical record data set related to the region. Perform data feature analysis based on the current tower monitoring data, use the data features to traverse and compare with the earthquake tower reference set, and obtain the same type of monitoring data set: perform feature analysis on the current tower monitoring data, and extract key data features. Then, use these data features to traverse and compare in the earthquake tower reference set to find similar data sets. Based on the same region record data set and the same type of monitoring data set, respectively predict the current tower monitoring data to obtain the first prediction result and the second prediction result: use the extracted historical record data set and similar data sets to predict the current tower monitoring data to obtain two prediction results. Based on the first prediction result and the earthquake disaster prediction result, the loss value is settled to obtain the first loss value: the first prediction result and the earthquake disaster prediction result are compared to calculate and obtain the corresponding first loss value. Based on the second prediction result and the earthquake disaster prediction result, the loss value is settled to obtain the second loss value: the second prediction result and the earthquake disaster prediction result are compared to calculate and obtain the corresponding second loss value. The first loss value and the second loss value are used to perform fusion calculation according to the preset weights to determine the deviation adjustment coefficient, and the earthquake disaster prediction result is adjusted by the deviation adjustment coefficient to generate an auxiliary earthquake disaster prediction result: according to the first loss value and the second loss value, a fusion calculation is performed according to the preset weights to obtain a deviation adjustment coefficient. Then, the earthquake disaster prediction result is adjusted by the deviation adjustment coefficient to generate an auxiliary earthquake disaster prediction result. The auxiliary earthquake disaster prediction result is combined with the earthquake disaster prediction result for guidance feedback: the auxiliary earthquake disaster prediction result is combined with the earthquake disaster prediction result, and the final result obtained can be used as guidance feedback for the earthquake disaster prediction result. This guidance feedback can be used to further optimize the model or provide it to decision makers. Through the above process, tower monitoring data and other relevant data can be used to assist in predicting earthquake disasters, and the accuracy and reliability of the prediction can be improved by evaluating and adjusting the results.

[0066] Embodiment 2

[0067] Based on the same inventive concept as the earthquake disaster prediction method based on tower monitoring data in the aforementioned embodiment, Figure 4 As shown, the present application provides an earthquake disaster prediction system based on tower monitoring data, the system comprising:

[0068] A monitoring data set acquisition module 10, wherein the monitoring data set acquisition module 10 is used to connect to the tower big data system to acquire a monitoring data set;

[0069] The tower reference set construction module 20 is used to extract earthquake area and earthquake time node data from the monitoring data set, and construct an earthquake tower reference set in combination with earthquake disaster record data;

[0070] An earthquake disaster prediction model acquisition module 30, wherein the earthquake disaster prediction model acquisition module 30 is used to perform model training using the earthquake tower reference set to obtain an earthquake disaster prediction model;

[0071] A prediction structure acquisition module 40, wherein the prediction structure acquisition module 40 is used to acquire a data prediction structure;

[0072] A standardization processing module 50, wherein the standardization processing module 50 performs data standardization processing on the tower monitoring data at that time based on the data prediction structure;

[0073] The disaster relief feedback module 60 is used to input the standardized tower monitoring data into the earthquake disaster prediction model to perform disaster prediction, obtain earthquake disaster prediction results, and provide disaster relief strategy guidance feedback based on the earthquake disaster prediction results.

[0074] Furthermore, the system also includes:

[0075] The earthquake disaster prediction model acquisition module is used to preprocess the earthquake tower reference set to obtain a multi-dimensional target data set; perform data annotation based on the multi-dimensional target data set, divide the annotated multi-dimensional target data set according to a preset ratio, and construct a training data set and a test data set; build a neural network architecture, use the training data set and the test data set to train and test the neural network architecture, and obtain the earthquake disaster prediction model.

[0076] Furthermore, the system also includes:

[0077] A multi-dimensional target data set construction module is used to classify the data of the seismic tower reference set to obtain vibration monitoring data, signal monitoring data, and earthquake disaster record data; to analyze the vibration direction and vibration amplitude based on the vibration monitoring data to determine the vibration dimension data; to analyze the regional coverage, weighted offline level, and number of towers offline based on the signal monitoring data to determine the signal dimension data; to assess the disaster level based on the earthquake disaster record data to determine the disaster assessment dimension data; to align the vibration dimension data, signal dimension data, and disaster assessment dimension data by region and time to construct the multi-dimensional target data set.

[0078] Furthermore, the system also includes:

[0079] The training and test data set acquisition module sets labels based on the disaster assessment dimension data and annotates the multi-dimensional target data set; constructs a sample data structure according to the mapping relationship of the annotated data in the multi-dimensional target data set, wherein the sample data structure includes vibration dimension data, signal dimension data and identification information of their corresponding disaster assessment dimension data; and divides the multi-dimensional target data set after the sample data structure processing according to a preset ratio of 8:2 to obtain the training data set and the test data set.

[0080] Furthermore, the system also includes:

[0081] The standardized processing module is used to obtain the current earthquake requirements to be analyzed, including the earthquake area and earthquake time; based on the earthquake area and earthquake time, the tower big data system is connected to obtain the tower monitoring data of the current time; according to the data requirements of the vibration dimension data and the signal dimension data in the data prediction structure, the tower monitoring data of the current time is classified and calculated to complete the standardized processing of the data prediction structure.

[0082] Furthermore, the system also includes:

[0083] The auxiliary earthquake disaster prediction result acquisition module is used to obtain the current earthquake area, use the current earthquake area and the earthquake tower reference set for regional matching, and extract the record data set in the same area; perform data feature analysis based on the current tower monitoring data, use the data features to traverse and compare with the earthquake tower reference set to obtain the same type of monitoring data set; predict the current tower monitoring data based on the same area record data set and the same type of monitoring data set, respectively, to obtain a first prediction result and a second prediction result; settle the loss value based on the first prediction result and the earthquake disaster prediction result to obtain a first loss value; settle the loss value based on the second prediction result and the earthquake disaster prediction result to obtain a second loss value; use the first loss value and the second loss value to perform fusion calculation according to preset weights to determine the deviation adjustment coefficient, use the deviation adjustment coefficient to adjust the earthquake disaster prediction result, and generate an auxiliary earthquake disaster prediction result, and the auxiliary earthquake disaster prediction result is merged with the earthquake disaster prediction result for guidance feedback.

[0084] Through the detailed description of the earthquake disaster prediction method based on railway monitoring data mentioned above, technical personnel in this field can clearly understand the earthquake disaster prediction system based on railway monitoring data in this embodiment. For the system disclosed in the embodiment, since it corresponds to the device disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0085] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for predicting earthquake disasters based on tower monitoring data, characterized in that: include: Connect to the tower big data system to obtain monitoring data sets; Extracting earthquake area and earthquake time node data from the monitoring data set, and building an earthquake tower reference set in combination with earthquake disaster record data; Using the earthquake tower reference set to perform model training to obtain an earthquake disaster prediction model; Get data prediction structure; Based on the data prediction structure, data standardization processing is performed on the tower monitoring data at that time; The standardized tower monitoring data is input into the earthquake disaster prediction model to perform disaster prediction, and an earthquake disaster prediction result is obtained. Based on the earthquake disaster prediction result, guidance and feedback of disaster relief strategies are performed.

2. The method according to claim 1, characterized in that The earthquake tower reference set is used to perform model training to obtain an earthquake disaster prediction model, including: Preprocessing the seismic tower reference set to obtain a multi-dimensional target data set; Performing data annotation based on the multi-dimensional target data set, dividing the annotated multi-dimensional target data set according to a preset ratio, and constructing a training data set and a test data set; A neural network architecture is constructed, and the training data set and the test data set are used to train and test the neural network architecture to obtain the earthquake disaster prediction model.

3. The method according to claim 2, characterized in that Preprocessing the seismic tower reference set to obtain a multi-dimensional target data set includes: Classifying the data of the earthquake tower reference set to obtain vibration monitoring data, signal monitoring data, and earthquake disaster record data; Perform vibration direction and vibration amplitude analysis based on the vibration monitoring data to determine vibration dimension data; Based on the signal monitoring data, regional coverage, weighted offline level, and number of tower offline analyses are performed to determine signal dimension data; Perform disaster level assessment based on the earthquake disaster record data to determine disaster assessment dimension data; The vibration dimension data, signal dimension data, and disaster assessment dimension data are aligned by region and time to construct the multi-dimensional target data set.

4. The method according to claim 3, characterized in that Performing data annotation based on the multi-dimensional target data set, dividing the annotated multi-dimensional target data set according to a preset ratio, and constructing a training data set and a test data set, including: Setting labels based on the disaster assessment dimensional data, and labeling the multi-dimensional target data set; Constructing a sample data structure according to the mapping relationship of the labeled data in the multi-dimensional target data set, wherein the sample data structure includes identification information of vibration dimension data, signal dimension data and their corresponding disaster assessment dimension data; The multi-dimensional target data set after the sample data structure processing is segmented according to a preset ratio of 8:2 to obtain the training data set and the test data set.

5. The method according to claim 4, characterized in that Based on the data prediction structure, data standardization processing is performed on the tower monitoring data at that time, including: Obtain the current earthquake requirements to be analyzed, including earthquake area and earthquake time; Based on the earthquake area and earthquake time, connect to the tower big data system to obtain the tower monitoring data of that time; According to the data requirements of the vibration dimension data and the signal dimension data in the data prediction structure, the tower monitoring data at that time is classified and calculated to complete the standardized processing of the data prediction structure.

6. The method according to claim 1, characterized in that Also includes: Obtain the current earthquake region, use the current earthquake region and the earthquake tower reference set to perform regional matching, and extract the record data set in the same region; Perform data feature analysis based on the tower monitoring data, and use the data features to perform traversal comparison with the earthquake tower reference set to obtain a monitoring data set of the same type; Based on the same-region record data set and the same-type monitoring data set, respectively predict the current tower monitoring data to obtain a first prediction result and a second prediction result; Perform loss value settlement based on the first prediction result and the earthquake disaster prediction result to obtain a first loss value; Perform loss value settlement based on the second prediction result and the earthquake disaster prediction result to obtain a second loss value; The first loss value and the second loss value are fused and calculated according to preset weights to determine a deviation adjustment coefficient, and the earthquake disaster prediction result is adjusted using the deviation adjustment coefficient to generate an auxiliary earthquake disaster prediction result. The auxiliary earthquake disaster prediction result is merged with the earthquake disaster prediction result for guidance and feedback.

7. The earthquake disaster prediction system based on tower monitoring data is characterized by: include: A monitoring data set acquisition module, which is used to connect to the tower big data system to acquire a monitoring data set; A tower reference set construction module, which is used to extract earthquake area and earthquake time node data from the monitoring data set, and to construct an earthquake tower reference set in combination with earthquake disaster record data; An earthquake disaster prediction model acquisition module, the earthquake disaster prediction model acquisition module is used to perform model training using the earthquake tower reference set to obtain an earthquake disaster prediction model; A prediction structure acquisition module, wherein the prediction structure acquisition module is used to acquire a data prediction structure; A standardization processing module, wherein the standardization processing module performs data standardization processing on the tower monitoring data at that time based on the data prediction structure; The disaster relief feedback module is used to input the standardized tower monitoring data into the earthquake disaster prediction model to perform disaster prediction, obtain earthquake disaster prediction results, and provide disaster relief strategy guidance feedback based on the earthquake disaster prediction results.

Images