An accident analysis and early warning method and system based on disaster accident cases

Abstract

The present invention discloses an accident analysis and early warning method and system based on disaster accident cases, and relates to the field of accident analysis and early warning technology. The method comprises: constructing a disaster accident case library based on historical disaster accident information; analyzing the cause of the accident according to the disaster accident case library, extracting key features, classifying according to the key features, and obtaining natural disasters and non-natural disasters; building a disaster early warning channel in combination with the key features; presetting early warning indicators, which are embedded in the disaster early warning channel; obtaining target area information according to a geographic information system, inputting the target area information into the disaster early warning channel, and outputting target area early warning information; sending the target area early warning information to a user port for early warning. The present invention solves the technical problem that a single data source and analysis method cannot meet the early warning needs as the types of disaster accidents become more diverse and complicated in the prior art, and achieves the technical effect of improving the accuracy of disaster early warning.

Description

Technical Field

[0001] The present invention relates to the technical field of accident analysis and early warning, and in particular to an accident analysis and early warning method and system based on disaster accident cases. Background Art

[0002] With the global climate change and the intensification of human activities, the frequency and impact of disasters are gradually increasing, posing a serious threat to people's lives and property. Therefore, how to effectively prevent and respond to disasters and reduce the losses they cause has become an urgent problem to be solved.

[0003] Existing accident analysis and early warning methods rely on historical data and cases, but these data may not be complete or accurate, resulting in biased analysis results. As disaster accident types become more diverse and complex, a single data source and analysis method cannot meet the needs of early warning. Summary of the invention

[0004] The present application provides an accident analysis and early warning method and system based on disaster accident cases, which is used to solve the technical problem that a single data source and analysis method in the prior art cannot meet the early warning needs as the types of disaster accidents become more diverse and complicated.

[0005] In view of the above problems, the present application provides an accident analysis and early warning method and system based on disaster accident cases.

[0006] In a first aspect of the present application, there is provided an accident analysis and early warning method based on a disaster accident case, the method comprising:

[0007] A disaster accident case library is constructed based on historical disaster accident information, and the disaster accident case library includes accident causes, number of casualties and response plans; the cause of the accident is analyzed according to the disaster accident case library, key features are extracted, and classification is performed according to the key features to obtain natural disasters and non-natural disasters; a disaster warning channel is built in combination with the key features, and the disaster warning channel includes a natural disaster warning module and a non-natural disaster warning module; warning indicators are preset and embedded in the disaster warning channel, and the warning indicators include quantitative indicators and qualitative indicators; target area information is acquired according to the geographic information system, the target area information is input into the disaster warning channel, and target area warning information is output; based on the target area warning information, it is sent to the user port for warning.

[0008] The second aspect of the present application provides an accident analysis and early warning system based on disaster accident cases, the system comprising:

[0009] A disaster accident case library construction unit, wherein the disaster accident case library construction unit constructs a disaster accident case library based on historical disaster accident information, and the disaster accident case library includes accident causes, number of casualties and response plans; a key feature classification unit, wherein the key feature classification unit analyzes the cause of the accident according to the disaster accident case library, extracts key features, classifies according to the key features, and obtains natural disasters and non-natural disasters; a disaster warning channel construction unit, wherein the disaster warning channel construction unit combines the key features to build a disaster warning channel, and the disaster warning channel includes a natural disaster warning module and a non-natural disaster warning module; a warning indicator preset unit, wherein the warning indicator preset unit presets warning indicators, which are embedded in the disaster warning channel, and the warning indicators include quantitative indicators and qualitative indicators; a target area warning information acquisition unit, wherein the target area warning information acquisition unit acquires target area information according to a geographic information system, inputs the target area information into the disaster warning channel, and outputs target area warning information; a warning unit, wherein the warning unit sends the target area warning information to a user port for warning based on the target area warning information.

[0010] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0011] The present application constructs a disaster accident case library based on historical disaster accident information, and the disaster accident case library includes accident causes, casualties and response plans; analyzes the cause of the accident according to the disaster accident case library, extracts key features, classifies according to the key features, and obtains natural disasters and non-natural disasters; builds a disaster warning channel in combination with key features, and the disaster warning channel includes a natural disaster warning module and a non-natural disaster warning module; presets warning indicators, which are embedded in the disaster warning channel, and the warning indicators include quantitative indicators and qualitative indicators; obtains target area information according to a geographic information system, inputs the target area information into the disaster warning channel, and outputs target area warning information; sends the target area warning information to the user port for warning. The present invention solves the technical problem that a single data source and analysis method cannot meet the warning needs as the types of disaster accidents become more diverse and complicated in the prior art, and achieves the technical effect of improving the accuracy of disaster warnings through a series of steps such as building a disaster accident case library, analyzing and classifying the causes of accidents, building a disaster warning channel, presetting warning indicators, and obtaining and analyzing target area information in combination with a geographic information system. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0013] Figure 1 A schematic diagram of a process flow of an accident analysis and early warning method based on a disaster accident case provided in an embodiment of the present application;

[0014] Figure 2 A schematic diagram of the structure of an accident analysis and early warning system based on disaster accident cases provided in an embodiment of the present application.

[0015] Explanation of the accompanying drawings: disaster accident case library construction unit 11, key feature classification unit 12, disaster warning channel construction unit 13, warning indicator preset unit 14, target area warning information acquisition unit 15, warning unit 16. DETAILED DESCRIPTION

[0016] This application provides an accident analysis and early warning method and system based on disaster accident cases, aiming to solve the technical problem in the existing technology that a single data source and analysis method cannot meet the early warning needs as the types of disaster accidents become more diverse and complicated. The application achieves the technical effect of improving the accuracy of disaster early warning through a series of steps such as building a disaster accident case library, analyzing and classifying the causes of accidents, setting up a disaster early warning channel, presetting early warning indicators, and combining a geographic information system to obtain and analyze information in the target area.

[0017] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings 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.

[0018] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products or devices.

[0019] Embodiment 1

[0020] like Figure 1 As shown, the present application provides an accident analysis and early warning method based on disaster accident cases, the method comprising:

[0021] Step S100: constructing a disaster accident case database based on historical disaster accident information, wherein the disaster accident case database includes accident causes, casualties, and response plans;

[0022] In the embodiment of the present application, the type, time span, geographical area, etc. of the historical disaster accidents to be collected are determined. The historical disaster accident information is collected through various channels, such as news media, research institutions, professional books, etc. During the collection process, the accuracy and completeness of the information are ensured to ensure the reliability of the case library.

[0023] After collecting enough information, filter, organize and categorize it to remove duplicate, incomplete or irrelevant information.

[0024] During the sorting process, the causes of the disaster, the number of casualties and the response plan are determined. The cause of the accident is the key to understanding the mechanism of disaster occurrence and predicting future disaster risks; the number of casualties reflects the severity and scope of the disaster; the response plan is the effective measures to be taken when similar disasters occur.

[0025] The above steps complete the construction of the disaster accident case database.

[0026] Step S200: Analyze the causes of accidents according to the disaster accident case database, extract key features, classify according to the key features, and obtain natural disasters and non-natural disasters;

[0027] In an embodiment of the present application, detailed case information is collected from a disaster accident case library, including the time, location, environmental description, course of the accident, and casualties of the accident.

[0028] Analyze each case one by one to identify the various factors that led to the accident, including natural factors, human factors, technical factors, etc. Extract key features closely related to the cause of the accident from the analysis process. These features are specific numerical indicators or descriptive information.

[0029] According to the extracted key features, the disaster accidents are classified into natural disasters and non-natural disasters.

[0030] Step S300: Building a disaster warning channel in combination with the key features, wherein the disaster warning channel includes a natural disaster warning module and a non-natural disaster warning module;

[0031] In the embodiment of the present application, key features extracted from the disaster accident case library are analyzed. These key features are multiple factors, including the intensity, frequency, location of natural disasters, etc., as well as the cause, type, and scope of impact of the disasters. Through analysis, the occurrence mechanism and characteristics of different disaster types are determined.

[0032] Based on the analysis results of key features, the overall framework of the early warning module is designed, which includes key links such as data input, data processing, early warning analysis, and early warning release.

[0033] Based on the key characteristics of natural disasters, a natural disaster early warning module is built. This module receives and processes monitoring data related to natural disasters in real time, such as earthquake waves, rainfall, wind speed, etc. By using machine learning, such as convolutional neural networks, these data are analyzed and predicted, potential risks of natural disasters are discovered in a timely manner, and corresponding early warning information is generated.

[0034] For early warning of non-natural disasters, a corresponding early warning module is constructed based on key features. This module includes monitoring and analysis of multiple factors such as equipment status, human operation, and social environment. By building a suitable early warning model, risk factors that may lead to non-natural disasters can be identified and early warnings can be issued in a timely manner.

[0035] After completing the construction of the natural disaster warning module and the non-natural disaster warning module, they will be integrated to build a disaster warning channel.

[0036] Step S400: presetting early warning indicators, which are embedded in the disaster early warning channel, and the early warning indicators include quantitative indicators and qualitative indicators;

[0037] In the embodiment of the present application, the early warning indicators are the basis of the disaster early warning channel, which measures and judges the risk and possibility of disasters. These indicators are divided into two categories: quantitative indicators and qualitative indicators, which together constitute a complete system of disaster early warning.

[0038] Quantitative indicators are based on specific values ​​or data, obtained through statistics, and have clear value ranges and thresholds. Quantitative indicators include disaster intensity indicators, disaster frequency indicators, and disaster impact range indicators.

[0039] Qualitative indicators are set based on experience and include disaster precursor indicators, social environment indicators, etc.

[0040] Finally, the preset early warning indicators are embedded in the disaster early warning channel. The threshold of each indicator is set according to factors such as disaster type, regional characteristics, and historical experience.

[0041] Step S500: acquiring target area information according to the geographic information system, inputting the target area information into the disaster warning channel, and outputting target area warning information;

[0042] In the embodiment of the present application, a geographic information system is a system that can capture, store, manage, analyze and present geographic spatial data. When acquiring information about a target area, GIS technology is used to collect relevant geographic data of the target area. Relevant geographic data includes various types of data such as topography, climate characteristics, geological structure, vegetation coverage, population distribution, etc.

[0043] The collected data is cleaned, formatted and converted to ensure accuracy and consistency. Then, the analysis function of GIS is used to extract key information related to disaster warning in the target area, such as potential landslide areas and flood-prone areas.

[0044] The target area information is input into the disaster warning channel. After receiving the target area information, the disaster warning channel uses the preset warning model and algorithm to analyze and judge, and outputs the target area warning information.

[0045] Step S600: Sending warning information based on the target area to the user port for warning.

[0046] In the embodiment of the present application, the most critical and important warning information for the user is screened out according to factors such as the urgency of the warning and the scope of impact. The warning information is converted into a format suitable for reception and display by the user port. According to the user's device type, network environment and the nature of the warning information, the most appropriate communication channel is selected, such as SMS, email, mobile application push notification, social media, etc. Finally, the warning is issued through the selected communication channel.

[0047] Furthermore, step S200 in the method provided in the application embodiment also includes:

[0048] Selecting a first disaster case according to the disaster accident case database, and obtaining a first disaster key feature based on the first disaster case, wherein the first disaster key feature includes quantitative data and qualitative data of the accident;

[0049] Setting disaster attributes, including natural disasters and non-natural disasters;

[0050] Performing disaster attribute recognition on the first disaster key feature to obtain a first disaster attribute;

[0051] The disaster accident case library is traversed to obtain disaster attributes corresponding to the historical disaster accident information.

[0052] In the embodiment of the present application, a disaster accident case is randomly selected from the disaster accident case library as the first disaster case. The first disaster key features are extracted from the selected first disaster case, and the first disaster key features include quantitative data and qualitative data. Quantitative data includes specific values ​​such as the number of casualties, etc.; qualitative data includes a description of the cause of the accident, an assessment of the scope of impact, etc.

[0053] Set disaster attributes and divide them into two categories: natural disasters and non-natural disasters. Natural disasters include earthquakes, floods, typhoons, etc.; non-natural disasters include traffic accidents, fires, chemical leaks, etc.

[0054] The key feature of the first disaster is matched with the preset disaster attribute, and the disaster attribute to which the first disaster case belongs is determined based on the matching result.

[0055] By traversing the disaster accident case database, key information is extracted from each case, and the disaster attributes of each case are identified to determine the disaster attributes to which it belongs.

[0056] Furthermore, step S300 in the method provided in the application embodiment also includes:

[0057] The disaster warning channel includes natural disaster warning module and non-natural disaster warning module;

[0058] Inputting the target area image into the disaster warning channel, and inputting into the natural disaster warning module and the non-natural disaster warning module respectively;

[0059] Outputting natural recognition results and non-natural recognition results respectively based on the natural disaster warning module and the non-natural disaster warning module;

[0060] Making a judgment based on the natural recognition result and the non-natural recognition result to determine the recognition result of the target area image;

[0061] Based on the recognition results, preset warning indicators are analyzed to obtain accident warning information.

[0062] In the embodiment of the present application, the target area image is input into the natural disaster warning module and the non-natural disaster warning module at the same time. The natural disaster warning module identifies natural disaster features of the image, such as terrain changes, water body anomalies, etc. The non-natural disaster warning module identifies non-natural disaster features of the image, such as smoke, flames, and vehicle congestion.

[0063] The natural disaster warning module outputs the identification results of natural disasters such as earthquakes and floods. The non-natural disaster warning module outputs the identification results of non-natural disasters such as fires and traffic accidents. Combining the identification results of the two modules, it is determined whether there is a disaster risk in the target area. If any module identifies a disaster risk, it enters the warning analysis stage.

[0064] Based on the identification results, the preset warning indicators are analyzed. These indicators include the level of the disaster, the scope of impact, the expected time of occurrence, etc. The severity and urgency of the disaster risk are determined through analysis. Finally, based on the results of the warning indicator analysis, accident warning information is generated. Accident warning information includes detailed information such as the type of disaster, the location of occurrence, and the expected impact.

[0065] Furthermore, the disaster warning channel includes a natural disaster warning module and a non-natural disaster warning module, and the method further includes:

[0066] Based on the disaster attributes corresponding to the historical disaster accident information, obtaining historical natural disaster data;

[0067] Building an initial natural disaster warning module, and performing supervised training on the initial natural disaster warning module based on the natural disaster historical data;

[0068] The preset model accuracy is determined, and a natural disaster warning module is obtained based on the preset model accuracy.

[0069] In the embodiment of the present application, all historical disaster accident information related to natural disasters is screened out from the disaster accident case library according to the disaster attributes, and the screened data is cleaned to remove duplicate, erroneous or irrelevant information.

[0070] Design the overall architecture of the natural disaster warning module, including the data input layer, feature extraction layer, model calculation layer, and result output layer. According to the characteristics of natural disasters and warning requirements, select key feature indicators, such as geological data, meteorological data, historical disaster records, etc. Use feature extraction technology to extract effective feature information from the original data for subsequent model training. Select a machine learning or deep learning model suitable for natural disaster warning, such as a convolutional neural network. Based on the selected model, build an initial natural disaster warning model.

[0071] The historical data of natural disasters are divided into training set, validation set and test set. The training set is used to train the initial natural disaster warning module, and the parameters and structure of the model are optimized through iteration. During the training process, appropriate loss functions and optimization algorithms are used to improve the prediction accuracy and generalization ability of the model. The trained model is verified using the validation set. According to the verification results, the model is tuned, such as adjusting model parameters, changing model structure, etc., to further improve the warning accuracy.

[0072] According to the actual application requirements and scenarios, a reasonable model accuracy threshold is preset. This threshold balances the accuracy and false alarm rate of the warning to ensure the effectiveness and reliability of the warning system. The trained natural disaster warning module is tested using the test set, and its accuracy index on the test set is calculated. The accuracy index obtained from the test is compared with the preset model accuracy threshold.

[0073] If the accuracy index obtained from the test meets or exceeds the preset model accuracy threshold, the module will be used as a natural disaster warning module.

[0074] Furthermore, an initial natural disaster warning module is constructed, and supervised training is performed on the initial natural disaster warning module based on the natural disaster historical data. The method further includes:

[0075] Obtaining a preset data division ratio, and dividing the natural disaster historical data into a training data set and a verification data set according to the preset data division ratio;

[0076] Performing supervised training on the initial natural disaster warning module through the training data set, and when the model output result tends to converge, verifying the output result of the initial natural disaster warning module through the verification data set;

[0077] Preset model accuracy, when the initial natural disaster warning module meets the preset model accuracy, a natural disaster warning module is obtained.

[0078] In the embodiment of the present application, the preset data division ratio is determined based on experience. The division ratio includes 70% training data, 20% validation data, and 10% test data. When dividing the data, ensure the balance of data of different categories, so that the number of samples of each category in the training set and the validation set is relatively balanced, and avoid overfitting or underfitting of the model to a certain category.

[0079] Use random sampling or stratified sampling methods to divide the natural disaster historical data into training data sets and validation data sets according to a preset ratio. Load the training data set into the model and prepare for training. Use supervised learning methods to train the initial natural disaster warning module. By continuously adjusting the parameters and structure of the model, the model can better fit the relationship between the features and labels in the training data. During the training process, monitor the changes in the model's loss function value, accuracy and other indicators in real time. When the model output results tend to converge, that is, the loss function value gradually stabilizes or no longer decreases significantly, it is considered that the model has learned certain rules. Then input the validation data set into the trained initial natural disaster warning module to obtain the output results of the model. Compare the output results with the real labels in the validation data set, and calculate the model's accuracy, recall and other evaluation indicators. Evaluate the performance of the model based on the validation results.

[0080] The model accuracy obtained on the validation data set is compared with the preset accuracy threshold. If the model accuracy is higher than or equal to the threshold, the model is considered to meet the requirements; otherwise, the model needs to be further adjusted or optimized. When the model accuracy meets the requirements, the trained initial natural disaster warning module is used as the final natural disaster warning module.

[0081] Furthermore, a pre-set warning indicator is embedded in the disaster warning channel, and the warning indicator includes a quantitative indicator and a qualitative indicator. The method further includes:

[0082] Quantitative indicators include meteorological indicators, geological indicators, hydrological indicators and equipment status indicators;

[0083] Qualitative indicators include abnormal phenomenon indicators and abnormal behavior indicators;

[0084] Allocate weights to the quantitative indicators and the qualitative indicators, perform weighted calculations, and obtain early warning coefficients;

[0085] The warning coefficient and the warning index are judged and warning information is output.

[0086] In the embodiments of the present application, the quantitative indicators include meteorological indicators, geological indicators, hydrological indicators and equipment status indicators. Meteorological indicators include temperature, humidity, wind speed, wind direction, air pressure, precipitation, etc. Geological indicators include seismic activity, ground subsidence, soil moisture, rock stability, etc. Hydrological indicators include river flow, water level, water quality, reservoir water storage, etc. Equipment status indicators refer to the operating status of instruments and equipment used to monitor disasters, such as the sensitivity of sensors, the failure rate of equipment, etc.

[0087] Qualitative indicators include abnormal phenomenon indicators and abnormal behavior indicators. Abnormal phenomenon indicators are based on human observation and experience, such as abnormal animal behavior, ground cracks, and tree tilts. Abnormal behavior indicators refer to abnormal situations in human social activities, such as abnormal crowd gathering and abnormal traffic congestion.

[0088] Weights are assigned to quantitative and qualitative indicators. The weight assignment is determined based on historical data. For example, for earthquake warning, geological indicators have higher weights; and for flood warning, hydrological indicators have higher weights. After the weight assignment is completed, the weight is calculated according to the weight assignment of each indicator to obtain the warning coefficient. The calculation process is to multiply the value of each indicator by the weight and finally sum it up.

[0089] After the early warning coefficient is calculated, it is compared with the preset early warning index. When the early warning coefficient reaches or exceeds the early warning index, the early warning information is output.

[0090] Furthermore, step S600 in the method provided in the application embodiment also includes:

[0091] Establishing a user information management portal, wherein the user information management portal includes user contact information and location;

[0092] Construct an early warning information sending channel, transmit the target area early warning information through the early warning information sending channel, and send it to the user port for early warning.

[0093] In the embodiment of the present application, the user information management port is an important port for storing, managing and updating user information, including the user's contact information and location.

[0094] The warning information sending channel is the path for delivering warning information to users. When constructing the warning information sending channel, the warning information is sent to the user's mobile phone in the form of text messages through the operator's SMS gateway. At the same time, the target warning area is finely divided so that different warning information can be sent to users in different areas. Finally, efficient transmission protocols and algorithms are used to reduce the delay and packet loss rate of information transmission, ensuring the real-time and accuracy of warning information.

[0095] Through the above steps, the construction of the warning information sending channel is completed.

[0096] In the embodiments of the present application, in summary, the embodiments of the present application have at least the following technical effects:

[0097] The present application constructs a disaster accident case library based on historical disaster accident information, and the disaster accident case library includes accident causes, casualties and response plans; analyzes the cause of the accident according to the disaster accident case library, extracts key features, classifies according to the key features, and obtains natural disasters and non-natural disasters; builds a disaster warning channel in combination with key features, and the disaster warning channel includes a natural disaster warning module and a non-natural disaster warning module; presets warning indicators, which are embedded in the disaster warning channel, and the warning indicators include quantitative indicators and qualitative indicators; obtains target area information according to a geographic information system, inputs the target area information into the disaster warning channel, and outputs target area warning information; sends the target area warning information to the user port for warning. The present invention solves the technical problem that a single data source and analysis method cannot meet the warning needs as the types of disaster accidents become more diverse and complicated in the prior art, and achieves the technical effect of improving the accuracy of disaster warnings through a series of steps such as building a disaster accident case library, analyzing and classifying the causes of accidents, building a disaster warning channel, presetting warning indicators, and obtaining and analyzing target area information in combination with a geographic information system.

[0098] Embodiment 2

[0099] Based on the same inventive concept as the accident analysis and early warning method based on disaster accident cases in the aforementioned embodiment, Figure 2 As shown, the present application provides an accident analysis and early warning system based on disaster accident cases. The system and method embodiments in the present application are based on the same inventive concept. The system includes:

[0100] A disaster accident case library construction unit 11, wherein the disaster accident case library construction unit 11 constructs a disaster accident case library based on historical disaster accident information, wherein the disaster accident case library includes accident causes, casualties, and response plans;

[0101] A key feature classification unit 12, wherein the key feature classification unit 12 performs accident cause analysis based on the disaster accident case library, extracts key features, and classifies the disasters based on the key features to obtain natural disasters and non-natural disasters;

[0102] A disaster warning channel building unit 13, wherein the disaster warning channel building unit 13 builds a disaster warning channel in combination with the key features, wherein the disaster warning channel includes a natural disaster warning module and a non-natural disaster warning module;

[0103] An early warning indicator preset unit 14, wherein the early warning indicator preset unit 14 presets early warning indicators, which are embedded in the disaster early warning channel, and the early warning indicators include quantitative indicators and qualitative indicators;

[0104] A target area warning information acquisition unit 15, wherein the target area warning information acquisition unit 15 acquires target area information according to a geographic information system, inputs the target area information into the disaster warning channel, and outputs target area warning information;

[0105] The early warning unit 16 sends the target area early warning information to the user port for early warning.

[0106] Furthermore, the system also includes:

[0107] Selecting a first disaster case according to the disaster accident case database, and obtaining a first disaster key feature based on the first disaster case, wherein the first disaster key feature includes quantitative data and qualitative data of the accident;

[0108] Setting disaster attributes, including natural disasters and non-natural disasters;

[0109] Performing disaster attribute recognition on the first disaster key feature to obtain a first disaster attribute;

[0110] The disaster accident case library is traversed to obtain disaster attributes corresponding to the historical disaster accident information.

[0111] Furthermore, the system also includes:

[0112] The disaster warning channel includes natural disaster warning module and non-natural disaster warning module;

[0113] Inputting the target area image into the disaster warning channel, and inputting into the natural disaster warning module and the non-natural disaster warning module respectively;

[0114] Outputting natural recognition results and non-natural recognition results respectively based on the natural disaster warning module and the non-natural disaster warning module;

[0115] Making a judgment based on the natural recognition result and the non-natural recognition result to determine the recognition result of the target area image;

[0116] Based on the recognition results, preset warning indicators are analyzed to obtain accident warning information.

[0117] Furthermore, the system also includes:

[0118] Based on the disaster attributes corresponding to the historical disaster accident information, obtaining historical natural disaster data;

[0119] Building an initial natural disaster warning module, and performing supervised training on the initial natural disaster warning module based on the natural disaster historical data;

[0120] The preset model accuracy is determined, and a natural disaster warning module is obtained based on the preset model accuracy.

[0121] Furthermore, the system also includes:

[0122] Obtaining a preset data division ratio, and dividing the natural disaster historical data into a training data set and a verification data set according to the preset data division ratio;

[0123] Performing supervised training on the initial natural disaster warning module through the training data set, and when the model output result tends to converge, verifying the output result of the initial natural disaster warning module through the verification data set;

[0124] Preset model accuracy, when the initial natural disaster warning module meets the preset model accuracy, a natural disaster warning module is obtained.

[0125] Furthermore, the system also includes:

[0126] Quantitative indicators include meteorological indicators, geological indicators, hydrological indicators and equipment status indicators;

[0127] Qualitative indicators include abnormal phenomenon indicators and abnormal behavior indicators;

[0128] Allocate weights to the quantitative indicators and the qualitative indicators, perform weighted calculations, and obtain early warning coefficients;

[0129] The warning coefficient and the warning index are judged and warning information is output.

[0130] Furthermore, the system also includes:

[0131] Establishing a user information management portal, wherein the user information management portal includes user contact information and location;

[0132] Construct an early warning information sending channel, transmit the target area early warning information through the early warning information sending channel, and send it to the user port for early warning.

[0133] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. Other embodiments are within the scope of the attached claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0134] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

[0135] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. An accident analysis and early warning method based on disaster accident cases, characterized in that: The method comprises: Building a disaster accident case database based on historical disaster accident information, the disaster accident case database includes accident causes, casualties and response plans; Analyze the causes of accidents according to the disaster accident case database, extract key features, classify according to the key features, and obtain natural disasters and non-natural disasters; In combination with the key features, a disaster warning channel is built, which includes a natural disaster warning module and a non-natural disaster warning module; Preset early warning indicators, embedded in the disaster early warning channel, the early warning indicators include quantitative indicators and qualitative indicators; Acquire target area information according to the geographic information system, input the target area information into the disaster warning channel, and output target area warning information; Send the target area warning information to the user port for warning; Among them, the cause of the accident is analyzed according to the disaster accident case database, key features are extracted, and classification is performed according to the key features to obtain natural disasters and non-natural disasters, including: Selecting a first disaster case according to the disaster accident case database, and obtaining a first disaster key feature based on the first disaster case, wherein the first disaster key feature includes quantitative data and qualitative data of the accident; Setting disaster attributes, including natural disasters and non-natural disasters; Performing disaster attribute recognition on the first disaster key feature to obtain a first disaster attribute; Traversing the disaster accident case database to obtain disaster attributes corresponding to the historical disaster accident information; Among them, in combination with the key features, a disaster warning channel is built, and the disaster warning channel includes a natural disaster warning module and a non-natural disaster warning module, including: The disaster warning channel includes natural disaster warning module and non-natural disaster warning module; Inputting the target area image into the disaster warning channel, and inputting into the natural disaster warning module and the non-natural disaster warning module respectively; Outputting natural recognition results and non-natural recognition results respectively based on the natural disaster warning module and the non-natural disaster warning module; Making a judgment based on the natural recognition result and the non-natural recognition result to determine the recognition result of the target area image; Performing preset warning indicator analysis based on the recognition result to obtain accident warning information; Among them, the preset warning indicators are embedded in the disaster warning channel, and the warning indicators include quantitative indicators and qualitative indicators, including: Quantitative indicators include meteorological indicators, geological indicators, hydrological indicators and equipment status indicators; Qualitative indicators include abnormal phenomenon indicators and abnormal behavior indicators; Allocate weights to the quantitative indicators and the qualitative indicators, perform weighted calculations, and obtain early warning coefficients; The warning coefficient is compared with the preset index to output warning information.

2. The method according to claim 1, characterized in that The disaster warning channel includes a natural disaster warning module and a non-natural disaster warning module. The method includes: Based on the disaster attributes corresponding to the historical disaster accident information, obtaining historical natural disaster data; Building an initial natural disaster warning module, and performing supervised training on the initial natural disaster warning module based on the natural disaster historical data; The preset model accuracy is determined, and a natural disaster warning module is obtained based on the preset model accuracy.

3. The method according to claim 2, characterized in that An initial natural disaster warning module is constructed, and supervised training is performed on the initial natural disaster warning module based on the natural disaster historical data. The method includes: Obtaining a preset data division ratio, and dividing the natural disaster historical data into a training data set and a verification data set according to the preset data division ratio; Performing supervised training on the initial natural disaster warning module through the training data set, and when the model output result tends to converge, verifying the output result of the initial natural disaster warning module through the verification data set; Preset model accuracy, when the initial natural disaster warning module meets the preset model accuracy, a natural disaster warning module is obtained.

4. The method according to claim 1, wherein the warning information of the target area is sent to the user port for warning, the method comprising: Establishing a user information management portal, wherein the user information management portal includes user contact information and location; Construct an early warning information sending channel, transmit the target area early warning information through the early warning information sending channel, and send it to the user port for early warning.

5. An accident analysis and early warning system based on disaster accident cases, characterized in that: The system comprises: A disaster accident case library construction unit, wherein the disaster accident case library construction unit constructs a disaster accident case library based on historical disaster accident information, wherein the disaster accident case library includes accident causes, casualties, and response plans; A key feature classification unit, wherein the key feature classification unit performs accident cause analysis based on the disaster accident case library, extracts key features, and classifies the key features to obtain natural disasters and non-natural disasters; A disaster warning channel building unit, wherein the disaster warning channel building unit builds a disaster warning channel in combination with the key features, wherein the disaster warning channel includes a natural disaster warning module and a non-natural disaster warning module; An early warning indicator preset unit, wherein the early warning indicator preset unit presets early warning indicators, which are embedded in the disaster early warning channel, and the early warning indicators include quantitative indicators and qualitative indicators; a target area warning information acquisition unit, which acquires target area information according to a geographic information system, inputs the target area information into the disaster warning channel, and outputs target area warning information; An early warning unit, wherein the early warning unit sends the target area early warning information to the user port for early warning; Among them, the key characteristic classification units include: Selecting a first disaster case according to the disaster accident case database, and obtaining a first disaster key feature based on the first disaster case, wherein the first disaster key feature includes quantitative data and qualitative data of the accident; Setting disaster attributes, including natural disasters and non-natural disasters; Performing disaster attribute recognition on the first disaster key feature to obtain a first disaster attribute; Traversing the disaster accident case database to obtain disaster attributes corresponding to the historical disaster accident information; Among them, the key characteristic classification units include: The disaster warning channel includes natural disaster warning module and non-natural disaster warning module; Inputting the target area image into the disaster warning channel, and inputting into the natural disaster warning module and the non-natural disaster warning module respectively; Outputting natural recognition results and non-natural recognition results respectively based on the natural disaster warning module and the non-natural disaster warning module; Making a judgment based on the natural recognition result and the non-natural recognition result to determine the recognition result of the target area image; Performing preset warning indicator analysis based on the recognition result to obtain accident warning information; Among them, the early warning indicator preset unit includes: Quantitative indicators include meteorological indicators, geological indicators, hydrological indicators and equipment status indicators; Qualitative indicators include abnormal phenomenon indicators and abnormal behavior indicators; Allocate weights to the quantitative indicators and the qualitative indicators, perform weighted calculations, and obtain early warning coefficients; The warning coefficient is compared with the preset index to output warning information.

Images