Mountain torrent disaster comprehensive risk prediction method and system based on multi-source data

Abstract

The invention discloses a mountain torrent disaster comprehensive risk prediction method and system based on multi-source data, and relates to the technical field of mountain torrent disaster risk prediction.The method comprises the steps of multi-source data acquisition, multi-source data analysis and mountain torrent disaster prediction feedback, the multi-source data is accurately acquired, then the state corresponding to a region is evaluated, and the risk prediction accuracy is improved. And predicting whether the mountain torrent disaster occurs in the region and the disaster level, analyzing to obtain early warning maintenance of the mountain torrent disaster corresponding to the region, executing negative hidden danger cost pre-simulation of the adopted maintenance set under the mountain torrent disaster risk level, and evaluating to obtain the cost of the adopted early warning maintenance of the mountain torrent disaster corresponding to the region. And adaptive early warning maintenance under the corresponding mountain torrent disaster of the region is obtained through screening, so that accurate prediction of the corresponding mountain torrent disaster risk of the region is realized, effective measures are taken for prevention of a prediction result in time, and dangerous influence and risk loss caused by the mountain torrent disaster are reduced.

Description

Technical Field

[0001] This invention relates to the field of flash flood disaster risk prediction technology, specifically to a method and system for comprehensive flash flood disaster risk prediction based on multi-source data. Background Technology

[0002] With the increasing frequency of flash floods in contemporary society, a comprehensive risk prediction method for flash floods based on multi-source data is proposed to achieve efficient prediction of flash floods, thereby effectively avoiding being caught off guard when flash floods occur and reducing the harm and losses caused by flash floods. By achieving comprehensive data monitoring of the region, comprehensive multi-source data corresponding to the region is obtained. Based on the rich data support, intelligent and accurate prediction of flash floods in the region can be achieved.

[0003] Existing technologies, such as the invention patent application CN119090691A, disclose a method for assessing flash flood disaster risk based on multi-source data, which relates to the field of natural disaster risk assessment technology. The method includes: collecting multi-source indicator data of the object to be evaluated based on flash flood disaster risk assessment indicators to construct a feature value matrix; subsequently, using these data and standard intervals, determining the specific point value of each indicator through a membership degree calculation formula; next, calculating the relative membership degree between these point values ​​and feature values ​​to assess the relative risk level of each indicator; then, combining the relative risk membership degree and weight of each indicator, comprehensively calculating the overall risk membership degree of the object; based on this comprehensive risk membership degree, further calculating the risk level feature value; finally, determining the flash flood disaster risk level of the object to be evaluated based on the variable fuzzy evaluation criterion and the calculated risk level feature value. This method can achieve a quantitative characterization of the "fuzzy" and "clear" aspects of flash flood disaster risk, enabling accurate assessment of flash flood disaster risk.

[0004] The current technology has the following problems: 1. The above invention mainly collects multi-source indicator data of the object to be evaluated to construct a feature value matrix, and analyzes the risk level of flash flood disaster based on the matrix. However, the terrain of areas prone to flash floods is complex, and there are multiple influencing factors and emergencies, which will greatly affect the quality of multi-source data acquisition. There is a lack of monitoring and acquisition of multi-source data to avoid the influence of factors in the corresponding direction. The quality of multi-source data cannot be guaranteed, which has an erroneous influence on the final risk assessment of flash flood disaster, and the assessment results of flash flood disaster risk lack accuracy.

[0005] 2. At the same time, after confirming the risk level of flash flood disasters, there is a lack of analysis and feasibility assessment of disaster measures, and effective measures cannot be taken for known flash flood disaster risks. Timely prevention of corresponding flash flood disaster risks is not achieved, and it is difficult to reduce the dangerous impact and losses caused by flash flood disasters. Summary of the Invention

[0006] To address the aforementioned technical shortcomings, the present invention aims to provide a method and system for comprehensive risk prediction of flash flood disasters based on multi-source data.

[0007] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: The present invention provides a comprehensive risk prediction method for flash flood disasters based on multi-source data, including: Step 1, multi-source data acquisition: Based on the causal characteristics of historical flash flood disasters, the monitoring attention items corresponding to the current region are analyzed and obtained by collecting multi-source data, and the basic status of the current region is assessed.

[0008] Step 2: Multi-source data analysis: If the basic state of the current region is abnormal, based on the basic characteristics of the multi-source data of the observed region, predict whether a flash flood will occur in the current region and the level of the disaster, and analyze the early warning and maintenance of the flash flood disaster in the region.

[0009] Step 3: Flash Flood Disaster Prediction Feedback: When the current region is in an abnormal state or a flash flood disaster is expected, an early warning feedback is issued. At the same time, the pre-simulation technology corresponding to the model is used to simulate the early warning maintenance to be taken for the corresponding flash flood disaster in the region, evaluate the cost value of the early warning maintenance to be taken for the corresponding flash flood disaster in the region, and select the appropriate early warning maintenance for the corresponding flash flood disaster in the region.

[0010] In a second aspect, the present invention provides a comprehensive risk prediction system for flash flood disasters based on multi-source data, comprising: a multi-source data acquisition module, used to analyze and obtain monitoring attention items corresponding to the current region based on the causal characteristics of historical flash flood disasters, and to acquire multi-source data and assess the basic state of the current region.

[0011] The multi-source data analysis module is used to predict whether a flash flood will occur and its level if the basic state of the current region is abnormal, based on the basic characteristics of the multi-source data of the observed region, and to analyze and obtain the early warning and maintenance of the flash flood disaster in the region.

[0012] The flash flood disaster prediction and feedback module is used to provide early warning feedback when the current region is in an abnormal state or when a flash flood disaster is likely to occur in the region. At the same time, it uses the pre-simulation technology corresponding to the model to simulate the early warning maintenance to be taken for the corresponding flash flood disaster in the region, evaluates the cost value of the early warning maintenance to be taken for the corresponding flash flood disaster in the region, and selects the appropriate early warning maintenance for the corresponding flash flood disaster in the region.

[0013] The beneficial effects of this invention are as follows: 1. This invention provides a method and system for comprehensive risk prediction of flash floods based on multi-source data. By accurately acquiring multi-source data, it assesses the corresponding state of a region and predicts whether a flash flood will occur and the level of the disaster. It analyzes and obtains early warning maintenance for the corresponding flash flood in the region. At the same time, it performs a pre-simulation of the negative hidden danger cost of taking maintenance set under the risk level of flash flood, evaluates the cost of taking early warning maintenance for the corresponding flash flood in the region, and selects the appropriate early warning maintenance for the corresponding flash flood in the region. Thus, it achieves accurate prediction of the risk of flash floods in the region, and timely takes effective measures to prevent the prediction results, reducing the dangerous impact and risk loss caused by flash floods. At the same time, based on the efficient guarantee of the accuracy of multi-source data acquisition, it effectively avoids the shortcomings of prediction results caused by inaccurate data, and obtains correct and accurate prediction of flash flood risk results, so that the risk prediction of flash floods is comprehensively and effectively guaranteed.

[0014] 2. Based on the analysis of the monitoring considerations for the current region corresponding to the collection of multi-source data, multi-source data is obtained, and the basic status of the current region is assessed. This ensures the quality of multi-source data acquisition, provides an effective data reference benchmark for subsequent flash flood disaster risk prediction, and achieves efficient understanding of the corresponding status of the region.

[0015] 3. Predict whether flash floods will occur in the current region and the level of the disaster, and analyze the corresponding flash flood warning and maintenance in the region, so as to achieve accurate prediction of the corresponding flash flood risk in the region, timely understand the expected occurrence level of the corresponding flash flood, and take targeted risk response measures.

[0016] 4. By pre-simulating the cost of negative hidden dangers, the cost of early warning maintenance for corresponding mountain torrent disasters in the region is evaluated, and the appropriate early warning maintenance for corresponding mountain torrent disasters in the region is selected. This enables effective response to emergencies under the corresponding measures for mountain torrent disasters, avoids the inability to respond to emergencies and related hidden dangers, and ensures social safety and reduces the risk and loss caused by mountain torrent disasters. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of the implementation steps of the method of the present invention.

[0019] Figure 2This is a schematic diagram of the system structure connection of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Please see Figure 1 As shown, a comprehensive risk prediction method for flash flood disasters based on multi-source data includes: Step 1, multi-source data acquisition: Based on the causal characteristics of historical flash flood disasters, the monitoring attention items corresponding to the current region are analyzed and multi-source data are collected, and the basic status of the current region is assessed.

[0022] Based on the analysis of monitoring considerations for the current region corresponding to the collection of multi-source data, multi-source data is obtained and the basic status of the current region is assessed, thereby ensuring the quality of multi-source data acquisition, providing an effective data reference benchmark for subsequent flash flood disaster risk prediction, and achieving efficient understanding of the corresponding status of the region.

[0023] In a specific embodiment, the analysis obtains monitoring attention items corresponding to the current region's multi-source data collection. The specific analysis process is as follows: Extract the data anomaly types corresponding to historical failures in predicting flash floods from the database, and obtain the monitoring application equipment and monitoring placement direction corresponding to the data anomaly types. At the same time, obtain the data difference between the corresponding data for the data anomaly types. If the data difference is greater than the reference data difference stored in the database, it is determined that the monitoring placement direction of the equipment corresponding to the data anomaly type is abnormal. If the data corresponding to the data anomaly type contains all zero values, it is determined that the data equipment is abnormal. If both are true, there are two problems: equipment abnormality and incorrect placement direction. Based on the functional type corresponding to the abnormal equipment, the equipment is replaced, and the direction is adjusted. When the data difference between the monitoring data for a certain placement direction is less than the reference data difference, that direction is taken as the monitoring direction of the replaced equipment.

[0024] It should be noted that understanding the reasons for past flash flood prediction failures is crucial for mitigating these failures, thereby obtaining more accurate monitoring data, enabling more precise predictions of flash flood disasters, timely prevention, and reducing losses and dangers caused by flash floods. Sufficient preparation is essential to ensure adequate and effective monitoring. Furthermore, professional researchers have established reference function data sets for each newly added high-precision device. By systematically summarizing the functions of each device, a more intuitive presentation is possible when needed. Based on knowledge of past flash flood prediction failures, monitoring conditions for the monitoring equipment can be avoided, ensuring accurate multi-source data monitoring in the current region.

[0025] It should be noted that the reference data difference is set by professional researchers. The reference data difference is set based on the corresponding data from successful historical predictions of flash floods, providing an effective reference to ensure the accuracy of data acquisition. After identifying abnormal equipment, the equipment is replaced based on the monitored data and the corresponding function type. At the same time, based on the previous incorrect placement direction of the abnormal equipment, the incorrect direction is removed, and the monitoring data corresponding to the replacement equipment is acquired. The difference between the monitoring data in each direction is calculated. When the difference between the monitoring data in a certain placement direction is less than the reference data difference, it is taken as the correct placement monitoring direction for the replacement equipment. This serves as the monitoring equipment and placement direction for collecting multi-source data in the current area, and the monitoring precautions for collecting multi-source data in the current area are analyzed in this way.

[0026] In one specific embodiment, the multi-source data includes meteorological data, topographic data, and hydrological data.

[0027] It should be noted that meteorological data includes real-time rainfall and duration of rainfall, obtained from rain gauges or weather forecasts. Topographic data includes topographic relief, saturated hydraulic conductivity, and root density. Topographic relief is obtained using lidar sensors or satellite remote sensing sensors. Saturated hydraulic conductivity is obtained using dual-ring monitoring. Water is continuously supplied to the soil through two concentric rings, maintaining vertical infiltration in the inner ring and eliminating lateral seepage interference in the outer ring. Saturated hydraulic conductivity is calculated based on the stable infiltration rate, which is equal to the stable infiltration rate (cm / min) multiplied by 60 cm / h. Root density is obtained using ground-penetrating radar sensors or CCD imaging sensors. Hydrological data includes real-time water level and water flow velocity, obtained from water level stations or using flow sensors.

[0028] In a specific embodiment, the assessment obtains the basic state corresponding to the current region. The specific assessment process is as follows: the meteorological data, topographic data and hydrological data corresponding to the region are compared with the reference meteorological data threshold, reference topographic data threshold and reference hydrological data threshold stored in the database, respectively, to obtain the multi-source performance values ​​of the region under meteorological, topographic and hydrological conditions. If there is one or more multi-source performance values ​​of 1, the current region is determined to be in an abnormal state; otherwise, the current region is determined to be in a normal state.

[0029] It should be noted that reference meteorological data thresholds, reference topographic data thresholds, and reference hydrological data thresholds are set by professional researchers. These thresholds serve as reference values ​​for a more comprehensive and intuitive assessment of whether the basic state of the current region is normal, providing a more effective state reference for the prediction of flash floods. When there is one or more multi-source performance values ​​that are 0, the current region is determined to be in a normal state.

[0030] Step 2: Multi-source data analysis: If the basic state of the current region is abnormal, based on the basic characteristics of the multi-source data of the observed region, predict whether a flash flood will occur in the current region and the level of the disaster, and analyze the early warning and maintenance of the flash flood disaster in the region.

[0031] Predict whether flash floods will occur in the current region and the level of the disaster, and analyze the corresponding flash flood warning and maintenance in the region, so as to achieve accurate prediction of the risk of flash floods in the region, timely understand the expected occurrence level of flash floods in the region, and take targeted risk response measures.

[0032] In a specific embodiment, the prediction process for whether a flash flood will occur in the current region and the level of the disaster is as follows: Risk levels and data corresponding to historical flash floods are obtained from the database; line graphs corresponding to the risk levels of historical flash floods are constructed; based on the common features among the line graphs corresponding to the risk levels of historical flash floods, a line graph corresponding to the risk levels of historical flash floods is established; and a rule base corresponding to regional flash floods is established based on the line graphs corresponding to the risk levels of historical flash floods.

[0033] Meteorological, topographic, and hydrological data corresponding to a region are used to construct line graphs based on the chronological duration from the initial time period to the preset time period. With time as the horizontal axis and multi-source data as the vertical axis, a multi-source line graph corresponding to the region is formed. The multi-source line graph corresponding to the region is compared with the line graphs corresponding to the risk levels of historical actual flash floods in the region's flash flood disaster rule base. If the multi-source line graph corresponding to the region is the same as the line graph corresponding to a certain historical actual flash flood disaster risk level, it is determined that a flash flood disaster will occur in the region, and the risk level of the historical flash flood disaster is taken as the risk level of the flash flood disaster in the region.

[0034] It should be noted that each risk level of a flash flood disaster corresponds to a line graph, resulting in multiple line graphs. Based on the common data features among these line graphs, a single line graph corresponding to each risk level of a flash flood disaster is obtained, resulting in feature line graphs corresponding to each level of flash flood disasters. A rule base is constructed based on these line graphs. Line graphs are constructed from the multi-source data monitored in the current region, and the current data is compared with the historical flash flood disaster line graphs corresponding to each level stored in the rule base for similarity matching. Based on the trend trend of the data corresponding to the regional line graph and the trend trend of the corresponding historical line graphs in the rule base, the trend difference between the two is obtained. If the trend difference is less than a preset reference trend difference, the historical flash flood disaster risk level is taken as the risk level for a flash flood disaster in the region. Alternatively, if the regional line graph and the historical line graph are similar, the risk level is determined. Figure 1 In this case, historical flash flood disaster risk and disease are used as the corresponding disaster level for the current region; professional researchers set a reference trend difference based on their professional knowledge and historical experience, and researchers, based on their rich experience and knowledge, can predict the disaster risk level by analyzing the trend difference over a period of time.

[0035] In a specific embodiment, the analysis obtains the early warning maintenance for flash flood disasters in the corresponding region. The specific analysis process is as follows: when the disaster level of a flash flood disaster in the region is level four, blue is assigned as its early warning symbol; when the disaster level of a flash flood disaster in the region is level three, yellow is assigned as its early warning symbol; if the disaster level is level two, orange is assigned as its early warning symbol; if the disaster level is level one, red is assigned as its early warning symbol. After allocating the disaster-assigned early warning symbols for each level of flash flood disaster in the region, the hazard manifestations of historical flash flood disasters at each level are obtained from the database, and specific hazard characteristic data are extracted from the hazard manifestations. The specific hazard characteristic data is input into the flash flood disaster intelligent model. Based on the input data, the model will automatically generate the maintenance set for each level of flash flood disaster in the region.

[0036] It should be noted that the intelligent model of flash flood disaster was built by professional researchers using model building software. Based on the data corresponding to each historical flash flood disaster, the data was preprocessed and then input into the software to form the corresponding model. With the continuous input of data corresponding to each historical flash flood disaster, the model is continuously trained and improved to obtain the intelligent model of flash flood disaster. The model has a large amount of data storage and AI intelligent technology. Therefore, based on the intelligent AI computing technology corresponding to the model, it automatically generates maintenance sets of flash flood disasters at various levels based on the input specific hazard characteristic data.

[0037] Step 3: Flash Flood Disaster Prediction Feedback: When the current region is in an abnormal state or a flash flood disaster is expected, an early warning feedback is issued. At the same time, the pre-simulation technology corresponding to the model is used to simulate the early warning maintenance to be taken for the corresponding flash flood disaster in the region, evaluate the cost value of the early warning maintenance to be taken for the corresponding flash flood disaster in the region, and select the appropriate early warning maintenance for the corresponding flash flood disaster in the region.

[0038] By pre-simulating the costs of negative hazards, the cost of early warning and maintenance measures for corresponding flash floods in a region is assessed. Adaptive early warning and maintenance measures for corresponding flash floods in a region are then selected to enable effective response to emergencies under the corresponding measures for flash floods. This avoids the inability to respond to emergencies and related hazards when they occur, and ensures social safety and reduces the risk and loss caused by flash floods.

[0039] It should be noted that nature is full of unknowns. Therefore, for the maintenance sets corresponding to each level of flash flood disaster, the negative impacts of sudden events or natural factors under the execution of pre-set maintenance measures are extracted and simulated. The pre-set negative hazard characteristics are input into the model. Using the simulation operation and computing power of the model, the pre-simulation of sudden events under the established maintenance sets corresponding to each level of flash flood disaster is performed, and the cost calculation and evaluation of the established maintenance sets facing sudden negative events are carried out. This achieves a comprehensive evaluation of the measures taken for flash flood disasters, so as to ensure the effectiveness of the measures taken when disasters occur and minimize the losses and dangers caused by disasters. At the same time, the calculation method of the relevant models in this paper is the same as the above process, so it will not be elaborated further. Sudden situations include channel blockage, sudden precipitation, and high rise in groundwater level.

[0040] In a specific embodiment, the automatic generation of maintenance sets corresponding to various levels of flash flood disasters in a region is carried out as follows: After inputting specific hazard characteristic data corresponding to the occurrence of flash flood disasters into the model and inputting a preset number of maintenance requirements, maintenance measures for each direction corresponding to various levels of flash flood disasters in the region are generated. Based on the negative hazard characteristics extracted from the maintenance measures for each direction corresponding to various levels of flash flood disasters in the region, negative hazard characteristics are then input into the model to obtain cost data of negative hazard characteristics corresponding to each maintenance measure in each direction.

[0041] It should be noted that the cost data includes, but is not limited to, capital costs, equipment depreciation, personnel safety values, and time costs. Professional researchers preset the number of maintenance requirements based on all events that occurred during maintenance of historical flash flood disasters. Based on the characteristics of all events, each event is distinguished, and the number of maintenance requirements is preset based on the number of each event. This ensures that flash flood disasters receive more comprehensive maintenance and avoids insufficient disaster maintenance preparation.

[0042] In a specific embodiment, the assessment obtains the cost value of early warning maintenance for a region corresponding to a flash flood disaster. The specific assessment process is as follows: The cost data of each maintenance measure in each direction corresponding to each level of flash flood disaster in the region are input into the model to calculate the cost value of each maintenance measure in each direction corresponding to each level of flash flood disaster in the region. If the cost value of a maintenance measure in a certain direction corresponding to a certain level of flash flood disaster in the region is less than or equal to a preset cost value, it is determined to meet the maintenance standard. The maintenance measure in that direction is taken as the maintenance measure for the corresponding level of flash flood disaster in the region and recorded as the maintenance measure for the compliant direction. In this way, each maintenance measure for each compliant direction is obtained. If the cost value of a maintenance measure in a certain direction corresponding to a certain level of flash flood disaster in the region is greater than the preset cost value, it is determined to not meet the maintenance standard, and the maintenance measure for that direction corresponding to that level of flash flood disaster is removed.

[0043] At the same time, the specific loss cost value corresponding to the specific hazard characteristics of flash floods at each level is calculated using the model.

[0044] In a specific embodiment, the screening process for obtaining the appropriate early warning maintenance for the corresponding mountain torrent disaster in a given region is as follows: The cost value of each maintenance measure corresponding to each direction of each level of mountain torrent disaster is compared with the specific loss cost value corresponding to each level of mountain torrent disaster. If the cost value of a maintenance measure corresponding to a certain direction of a certain level of mountain torrent disaster is greater than the specific loss cost value corresponding to that level of mountain torrent disaster, then the measure is determined to be a match under that direction, and the measure is removed. Otherwise, the measure is determined to be a match and recorded as the target maintenance measure under that direction, thereby obtaining the maintenance measures for each direction.

[0045] The cost values ​​of each target maintenance measure in each direction corresponding to each level of flash flood disaster in the region are arranged in order from low to high. The one with the highest cost value is taken as the appropriate maintenance measure in each direction corresponding to each level of flash flood disaster in the region. In this way, the appropriate early warning maintenance for the corresponding flash flood disaster in the region is obtained.

[0046] Please see Figure 2 As shown, a comprehensive risk prediction system for flash floods based on multi-source data includes a multi-source data acquisition module, a multi-source data analysis module, a flash flood prediction feedback module, and a database.

[0047] The multi-source data acquisition module is connected to the multi-source data analysis module, the flash flood disaster prediction and feedback module, and the database, respectively. The multi-source data analysis module is connected to the flash flood disaster prediction and feedback module and the database, respectively.

[0048] The multi-source data acquisition module is used to analyze the causal characteristics of historical flash floods, obtain the monitoring attention items for the current region corresponding to the collection of multi-source data, acquire the multi-source data, and assess the basic status of the current region.

[0049] The multi-source data analysis module is used to predict whether a flash flood will occur and its level if the basic state of the current region is abnormal, based on the basic characteristics of the multi-source data of the observed region, and to analyze and obtain the early warning and maintenance of the flash flood disaster in the region.

[0050] The flash flood disaster prediction and feedback module is used to provide early warning feedback when the current region is in an abnormal state or when a flash flood disaster is likely to occur in the region. At the same time, it uses the pre-simulation technology corresponding to the model to simulate the early warning and maintenance measures to be taken for the corresponding flash flood disaster in the region, evaluates the cost of early warning and maintenance measures to be taken for the corresponding flash flood disaster in the region, and selects the appropriate early warning and maintenance measures for the corresponding flash flood disaster in the region.

[0051] The database is used to store data anomaly types, reference data differences, function types, multi-source data, reference meteorological data thresholds, reference topographic data thresholds, reference hydrological data thresholds, risk levels and data corresponding to each historical actual flash flood disaster, specific hazard characteristic data, negative hazard characteristics, and cost data.

[0052] This invention, through the accurate acquisition of multi-source data, assesses the status of a region and predicts whether flash floods will occur and their severity. It analyzes and obtains early warning and maintenance measures for flash floods in the region, while simultaneously performing a pre-simulation of the negative hazard costs of maintenance sets under different flash flood risk levels. This assesses the cost of early warning and maintenance measures for flash floods in the region and filters out suitable early warning and maintenance measures for the corresponding flash floods. This achieves accurate prediction of flash flood risks in the region, enabling timely and effective preventative measures to reduce the dangerous impacts and losses caused by flash floods. Furthermore, based on the efficient guarantee of the accuracy of multi-source data acquisition, it effectively avoids the shortcomings of inaccurate predictions caused by inaccurate data, resulting in correct and accurate predictions of flash flood risks, thus providing comprehensive and effective protection for flash flood risk prediction.

[0053] The examples described in this invention are not limited to the specific embodiments listed above. The examples are merely illustrative to facilitate understanding of the invention and do not constitute a limitation on the scope of protection of this invention. Any modifications, equivalent substitutions, etc., made within the spirit and principles of this invention should be included within the scope of protection.

[0054] The above description is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they should all fall within the protection scope of the present invention.

Claims

1. A comprehensive risk prediction method for flash floods based on multi-source data, characterized in that, Includes the following steps: Step 1: Multi-source data acquisition: Based on the causal characteristics of historical flash floods, analyze the monitoring points for collecting multi-source data in the current area, acquire the multi-source data, and assess the basic status of the current area. Step 2: Multi-source data analysis: If the basic state of the current region is abnormal, based on the basic characteristics of the multi-source data of the observed region, predict whether a flash flood will occur in the current region and the level of the disaster, and analyze the early warning and maintenance of the flash flood disaster in the region. Step 3: Flash Flood Disaster Prediction Feedback: When the current region is in an abnormal state or a flash flood disaster is expected, an early warning feedback is issued. At the same time, the pre-simulation technology corresponding to the model is used to simulate the early warning maintenance to be taken for the corresponding flash flood disaster in the region, evaluate the cost value of the early warning maintenance to be taken for the corresponding flash flood disaster in the region, and select the appropriate early warning maintenance for the corresponding flash flood disaster in the region.

2. The method for comprehensive risk prediction of flash flood disasters based on multi-source data according to claim 1, characterized in that, The analysis yields monitoring considerations for the current region based on multi-source data collection. The specific analysis process is as follows: Extract the data anomaly types corresponding to historical failures in predicting flash floods from the database, and obtain the corresponding monitoring application equipment and monitoring placement direction for each data anomaly type. Simultaneously, obtain the data difference between the corresponding data for each data anomaly type. If the data difference is greater than the reference data difference stored in the database, the monitoring placement direction of the equipment corresponding to the data anomaly type is determined to be abnormal. If all data corresponding to the data anomaly type are zero, the data equipment is determined to be abnormal. If both conditions are met, there are two problems: equipment abnormality and incorrect placement direction. Based on the functional type corresponding to the abnormal equipment, replace the equipment and perform direction adjustment. When the data difference between the monitoring data for a certain placement direction is less than the reference data difference, that direction is used as the monitoring direction of the replacement equipment.

3. The method for comprehensive risk prediction of flash flood disasters based on multi-source data according to claim 2, characterized in that, The multi-source data includes meteorological data, topographic data, and hydrological data.

4. The method for comprehensive risk prediction of flash flood disasters based on multi-source data according to claim 3, characterized in that, The assessment yields the basic state of the current region. The specific assessment process is as follows: The meteorological, topographic, and hydrological data corresponding to the region are compared with the reference meteorological, topographic, and hydrological data thresholds stored in the database to obtain the multi-source performance values ​​for the region under meteorological, topographic, and hydrological conditions. If there is one or more multi-source performance values ​​of 1, the current region is determined to be in an abnormal state; otherwise, the current region is determined to be in a normal state.

5. The method for comprehensive risk prediction of flash flood disasters based on multi-source data according to claim 4, characterized in that, The specific prediction process for determining whether a flash flood will occur in the current area and the level of the disaster is as follows: The risk levels and data corresponding to each historical flash flood disaster are obtained from the database. Line charts corresponding to the risk levels of each historical flash flood disaster are constructed. Based on the common features among the line charts corresponding to the risk levels of each historical flash flood disaster, a rule base corresponding to the risk levels of each historical flash flood disaster is established. Based on the line charts corresponding to the risk levels of each historical flash flood disaster, a rule base corresponding to the regional flash flood disaster is established. Meteorological, topographic, and hydrological data corresponding to a region are used to construct line graphs based on the chronological duration from the initial time period to the preset time period. With time as the horizontal axis and multi-source data as the vertical axis, a multi-source line graph corresponding to the region is formed. The multi-source line graph corresponding to the region is compared with the line graphs corresponding to the risk levels of historical actual flash floods in the region's flash flood disaster rule base. If the multi-source line graph corresponding to the region is the same as the line graph corresponding to a certain historical actual flash flood disaster risk level, it is determined that a flash flood disaster will occur in the region, and the risk level of the historical flash flood disaster is taken as the risk level of the flash flood disaster in the region.

6. The method for comprehensive risk prediction of flash flood disasters based on multi-source data according to claim 5, characterized in that, The analysis yielded early warning and maintenance information for flash floods in the corresponding regions. The specific analysis process is as follows: When a region experiences a flash flood disaster at level four, a blue warning symbol is assigned; when the disaster level is level three, a yellow warning symbol is assigned; when the disaster level is level two, an orange warning symbol is assigned; and when the disaster level is level one, a red warning symbol is assigned. After allocating warning symbols for each level of flash flood disaster in the region, the system retrieves the historical hazard characteristics of flash flood disasters at each level from the database, extracts specific hazard feature data from these hazard characteristics, and inputs this specific hazard feature data into the flash flood disaster intelligent model. Based on the input data, the model automatically generates a maintenance set for each level of flash flood disaster in the region.

7. The method for comprehensive risk prediction of flash flood disasters based on multi-source data according to claim 6, characterized in that, The automatic generation process yields the maintenance set corresponding to each level of flash flood disaster in the region. The specific generation process is as follows: After inputting specific hazard characteristic data corresponding to the occurrence of flash flood disasters into the model, and inputting the preset number of maintenance requirements, the model generates maintenance measures for each direction corresponding to each level of flash flood disaster in the region. Based on the negative hazard characteristics extracted from the maintenance measures for each direction corresponding to each level of flash flood disaster in the region, the negative hazard characteristics are then input into the model to obtain the cost data of the negative hazard characteristics corresponding to each maintenance measure in each direction.

8. The method for comprehensive risk prediction of flash flood disasters based on multi-source data according to claim 7, characterized in that, The assessment yielded the cost of implementing early warning and maintenance measures for flash flood disasters in the corresponding region. The specific assessment process is as follows: The cost data of each maintenance measure in each direction corresponding to each level of flash flood disaster in a region are input into the model to calculate the cost value of each maintenance measure in each direction corresponding to each level of flash flood disaster in the region. If the cost value of a maintenance measure in a certain direction corresponding to a certain level of flash flood disaster in a region is less than or equal to the preset cost value, it is determined to meet the maintenance standard. The maintenance measure in that direction is taken as the maintenance measure corresponding to the level of flash flood disaster in the region and recorded as the maintenance measure in the direction that meets the standard. This is how the maintenance measures in each direction that meet the standard are obtained. If the cost value of a maintenance measure in a certain direction corresponding to a certain level of flash flood disaster in a region is greater than the preset cost value, it is determined to not meet the maintenance standard. The maintenance measures in that direction corresponding to the level of flash flood disaster are removed. At the same time, the specific loss cost value corresponding to the specific hazard characteristics of flash floods at each level is calculated using the model.

9. The method for comprehensive risk prediction of flash flood disasters based on multi-source data according to claim 8, characterized in that, The screening process yields the appropriate early warning and maintenance system for the corresponding flash flood disaster in the region. The specific screening process is as follows: The cost value of each maintenance measure corresponding to each direction of each level of flash flood disaster is compared with the specific loss cost value corresponding to each level of flash flood disaster. If the cost value of a maintenance measure corresponding to a certain direction of a certain level of flash flood disaster is greater than the specific loss cost value corresponding to that level of flash flood disaster, the measure under that direction is determined to be matched and the measure is removed. Otherwise, the measure is determined to be matched and recorded as the target maintenance measure under that direction. In this way, the maintenance measures for each direction are obtained. The cost values ​​of each target maintenance measure in each direction corresponding to each level of flash flood disaster in the region are arranged in order from low to high. The one with the highest cost value is taken as the appropriate maintenance measure in each direction corresponding to each level of flash flood disaster in the region. In this way, the appropriate early warning maintenance for the corresponding flash flood disaster in the region is obtained.

10. A comprehensive flash flood disaster risk prediction system utilizing the multi-source data-based comprehensive flash flood disaster risk prediction method according to any one of claims 1-8, characterized in that, include: The multi-source data acquisition module is used to analyze the causal characteristics of historical flash floods, obtain the monitoring attention items for the current region corresponding to the collection of multi-source data, acquire the multi-source data, and assess the basic status of the current region. The multi-source data analysis module is used to predict whether a flash flood will occur and the level of the disaster if the basic state of the current region is abnormal, based on the basic characteristics of the multi-source data of the observed region, and to analyze and obtain the early warning and maintenance of the flash flood disaster in the region. The flash flood disaster prediction and feedback module is used to provide early warning feedback when the current region is in an abnormal state or when a flash flood disaster is likely to occur in the region. At the same time, it uses the pre-simulation technology corresponding to the model to simulate the early warning and maintenance measures to be taken for the corresponding flash flood disaster in the region, evaluates the cost of early warning and maintenance measures to be taken for the corresponding flash flood disaster in the region, and selects the appropriate early warning and maintenance measures for the corresponding flash flood disaster in the region.

Images