TWI674540B - A system for immediate assessment and early warning of industrial and commercial disasters caused by wind and flood - Google Patents

Abstract

The invention relates to a system capable of real-time assessment and early warning of industrial and commercial disasters caused by wind and flood disasters. A management server in the system is provided with a plurality of evaluation models, and can communicate with a plurality of terminal devices and at least one weather data. The monitoring server is connected, wherein the management server can receive the meteorological data of the aforementioned area by the meteorological data monitoring server according to the geographic area corresponding to the user of the terminal device, and import the aforementioned meteorological data into each of the evaluations separately. The model can estimate the damage assessment value of each user's geographic area in real time, and when the damage assessment value exceeds a preset gate value, an early warning signal is generated for each user, so that the system can Remind users to prepare for disaster prevention in advance, so as to prevent disasters before they occur, so as to effectively and significantly reduce the possible industrial and commercial disasters.

Description

System capable of real-time assessment and early warning of industrial and commercial damage caused by wind and flood disasters

The invention relates to a system for instant assessment of disaster losses and early warning of disasters, and more particularly, to a system for instant assessment and early warning of industrial and commercial disasters caused by wind and flood disasters.

According to Taiwan, Taiwan is in a high-risk area with frequent natural disasters (such as typhoons and earthquakes, especially typhoon-derived "wind disasters" and "floods"). Disasters often cause serious loss of people's lives and property. Prevention and rescue technology helps the government to achieve effective reduction of disaster losses and civilian casualties. It is not only an important issue for the government ’s priority consideration, but also an important issue that has been concerned by the relevant academic and industrial circles for many years and is anxious to solve. The inventor is the "National Disaster Prevention and Rescue Technology Center". In order to effectively improve the level of disaster prevention and rescue technology, for many years, he has been committed to integrating various types of disaster information and prevention and technology resources across sectors and ministries, so as to bring relevant The results of the disaster prevention research and development are implemented in the government's governance of disaster management, and effectively achieve the ultimate goal of preventive measures, and can significantly reduce the loss of people's lives and property.

The “National Disaster Prevention Technology Center” (that is, the unit to which the applicant and the inventor belongs, hereinafter referred to as the “Disaster Prevention Technology Center”) is the “Ministry of Science and Technology of the Executive Yuan” of the Republic of China to promote the development of national basic and applied science and technology. Work, specially set up one "administrative legal person" to coordinate, plan and promote related issues of disaster prevention and rescue technology research and development, and assist and support disaster prevention technology research and development Development and implementation of applications. In view of this, the "Disaster Prevention Technology Center" organizes data samples of disaster events in recent years at home and abroad, review data of disaster trends, inventory data of disaster prevention technology, etc., and considers large environmental needs and "disaster prevention Based on the attributes of human resources in the Science and Technology Center, the focus of work on future disaster prevention technology that needs to be strengthened and actively developed is summarized as follows: 1. Constructing an early warning and response network platform; 2. Constructing an advisory capable of assessing disaster risks And service platform; and 3. Construct a consulting and service platform that can effectively improve disaster prevention capabilities.

Among the aforementioned work priorities, the “Disaster Prevention Technology Center” intends to strengthen the following important topics from 2017 to 2020: 1. New thinking on disaster response management should be added: In this regard, the “Disaster Prevention Technology Center” is particularly The contents of the 7 major disaster reduction goals and 4 priority promotion work items proposed in the "2015-2030 Disaster Reduction Program" adopted by the Third World Conference on Disaster Reduction in March 2015 in Sendai, Japan were provided to the Government, academic and industry-related units are used as a reference for disaster prevention research, and their corresponding architecture, content and target planning and R & D proposals will become China's due governance actions and recommendations in the future, so that China's disaster prevention governance and performance can Synchronize with advanced countries in the world; 2. A more complete and accurate vertical disaster analysis structure should be achieved: check, "Integrated Research on Disaster" carried out by the International Council for Science (ICSU) Risk (IRDR) analysis and research has pointed out that in the past, countries around the world have mainly dealt with disaster issues. Approaches derived from learning from disaster experience (the so-called Retrospective Longitudinal Analysis, or RLA for short), however, in recent years, Events such as huge changes and changes in many extreme environments. In the future, the understanding of natural disasters should obviously focus more on how to perform more accurate prediction and assessment of disaster situations (Projective Longitudinal Analysis, referred to as PLA). Ability to face the severe challenges posed by changing disaster events. In response, the Disaster Prevention Science and Technology Center has made great strides in this direction in recent years, especially in the planning of work. In addition to striving to carry out detailed analysis of disaster causes, disaster investigations and current conditions based on past natural disasters and other events, , According to the corresponding disaster event book, in order to continue to accumulate experience and response capabilities of the major natural disaster events, and for many extreme weather events in recent years, such as: Large-scale earth and rock collapse, large-scale earthquakes, and climate change will be studied separately in order to strengthen the speculation and estimation of possible scenarios in the future, so as to serve as the basis for accurate assessment of the impact of disasters; To be meticulous and precise: In terms of dealing with disaster issues, it can be clearly seen through review and analysis that, regardless of the resolution or time and space scale, more detailed massive data and processing technology are needed to achieve orders. The disaster prevention and early warning technology has a more detailed and precise goal. Based on this, the Disaster Prevention Technology Center has strengthened this The promotion of various aspects includes, for example, flood warning for metropolitan areas, raising the level of flood warning to more detailed grid early-warning analysis technology, which can be used to make risk warning maps of different protection levels; for extreme disaster events (Such as: heavy rain with short delay), improve the ability of early warning of very short-term disaster weather (short-time heavy rainfall, warning of very short-term disaster weather); improve the ability of real-time assessment of disaster losses ... and so on; and 2. We should make good use of newly developed technologies to achieve the goal of making the application of disaster prevention technology in the lives of the people: With the advancement of the times and the rapid development of information technology, more innovative technologies can be used. Value-added applications in disaster prevention work, such as: the popular aerial telemetry technology in recent years is a best example; in addition, through the popular smartphones and social networks in recent years, people ’s channels for receiving the latest information and The source is also very different from the past. Therefore, if you can make good use of these innovative technologies, such as: analysis and value-added of relevant huge amounts of information, resource pooling and integration of various types of information, and social networking ( Social Media) near-term or real-time information for disaster prevention and rescue; cloud computing, high-speed mobile networks (such as 5G) or the Internet of Things for disaster prevention and rescue; 3D and visualization technologies for disaster prevention and rescue Applications; and cross-device (mobile phones, tablets, desktops, etc.), cross-platform services in disaster prevention and rescue applications ... etc .; so that each of these innovative technologies can be successfully added to the disaster prevention work, it is bound to It can effectively improve the effectiveness and results of disaster prevention operations.

In recent years, major natural disaster events that have spread frequently at home and abroad have made all sectors of the society pay more attention to disaster prevention and relief work. Taking the island of Taiwan as an example, with the rapid development of socioeconomic and natural environment in many aspects, many have been added. New disaster prevention issues and needs, this phenomenon has also gradually increased the potential disaster threats facing the entire society, thereby forcing government departments and all sectors of society to work hard to face up, plan and adjust many phenomena and problems in development. Ability to respond to the tasks and issues of future disaster prevention and rescue work. Taking the “National Center for Disaster Prevention Technology” and the “National Climate Change Adaptation Policy Disaster Field Action Plan” jointly promoted by relevant government ministries as examples, especially after the passage of the “Greenhouse Gas Reduction and Management Law”, In the process of adaptation in related fields, all relevant competent units must specifically propose corresponding adaptation action plans. Therefore, the Disaster Prevention Technology Center has accumulated considerable research results that can be used for reference by relevant government departments. Alas, based on the increasing awareness of disaster prevention by the central and local governments, governments at all levels are naturally paying more and more attention to disaster prevention and relief efforts, and the need for disaster prevention technology (such as disaster prevention maps) Resources, emergency response information, disaster vulnerability analysis, etc.) are also increasingly eager. Because of the increasing demand for disaster prevention technology from all walks of life and the government, the "Disaster Prevention Technology Center" can smoothly combine resources at all levels, coordinate units at all levels, and jointly develop and promote related disaster prevention technology tasks.

In addition, in response to changes in global climate and the environment, the needs of disaster management at all stages (disaster mitigation, improvement, response and reconstruction) are bound to be more diverse and arduous. In the past, Maplecroft has conducted large-scale research and analysis on the risks of natural disasters exposed to economic activities in various countries around the world, and based on this, it has proposed a disaster risk analysis atlas (Natural Hazards Risk Atlas 2011). After considering natural disasters such as earthquakes, volcanoes, tsunamis, storms, floods, landslides, etc., the risk management consulting firm clearly stated in its analysis report that in 196 countries worldwide, if only the absolute value is calculated (that is, denominated in US dollars) Natural disaster risk of natural disasters), Taiwan ’s natural disaster risk for economic activities ranks fourth in the world. Therefore, the risk management consultant company has listed Taiwan as a “high-risk” region for natural disasters, and has seen frequent “complex disasters” in recent years, such as: The disasters caused by typhoons such as Kameiji, Xinlock, Morak, and Sudir were not only increasing in scale, but even those that far exceeded the technological protection limits or can be expected from past disaster experiences are all special and extreme cases. , The existing disaster prevention and rescue plan, protection standards and operating mechanisms, naturally, should be immediately improved and Strong's urgent needs.

In view of this, how to compile, analyze and calculate the rich sample of weather data, disaster data, and disaster distribution areas accumulated in Taiwan over the past years due to typhoon-induced wind and flood disasters, to construct Disaster assessment and early warning systems in various regions can effectively and significantly reduce the potential industrial and commercial disasters. That is to become an inventor and related academic and industrial circles are working hard to research, an important issue that needs to be broken, and an important subject that this invention intends to explore and uncover here, in order to effectively enhance the metropolitan industry and commerce industry and its key facilities. Anti Disaster capabilities, and the ability to analyze the time and space of major industrial and commercial disasters can better meet the urgent needs of the business community, so that the business community can immediately recognize and identify the coming of disasters, so that they can face disasters, respond quickly, and research Develop the best disaster prevention countermeasures to achieve the goal of reducing disaster losses.

In response to the aforementioned problems in Taiwan where industrial and commercial disaster losses occur almost every year due to typhoons or other abnormal weather-induced "wind disasters" and "floods", the inventors have studied the disaster prevention early warning mechanisms of countries around the world for many years and have extensive experience in research With the assistance of huge data samples on “wind disasters” and “flood disasters” accumulated and provided by relevant Taiwanese authorities, academia, and industry over the years, after long-term efforts in analysis, research, and experimentation, it was finally developed. A system for real-time assessment and early warning of industrial and commercial disasters caused by wind and flood disasters is designed according to the present invention, so as to effectively and greatly reduce the possible industrial and commercial disasters.

An object of the present invention is to provide a system capable of real-time assessment and early warning of industrial and commercial disasters caused by wind and flood disasters. The system includes a plurality of terminal devices, the Internet, a management server, and at least one weather data monitor. A server, wherein each of the terminal devices is a separate owner of an insurance account registered to the system, and can be connected to the management server through the Internet, and the management server is connected through the Internet The network is connected to each of the meteorological data monitoring servers. Among them, at least one industrial and commercial wind damage assessment model and one industrial and commercial flood damage assessment model are installed in the management server. Provide the relevant industrial and commercial disaster data samples of each area during the typhoon-affected period and the meteorological data samples of each area at that time. Determine the "industrial damage amount" as "dependent variables", and use the Geographical coordinates, corresponding meteorological data at that time, industry category, Use floors, total insurance amount, etc ... as "independent variables", and substitute each corresponding independent variable in each industrial and commercial disaster data sample into at least two multiple regression equations (multiple linear regression equation) ), Through the calculation of the equations, the industrial and commercial wind disaster damage assessment models and the industrial and commercial flood damage assessment models for each area are separately constructed, and at least one assessment model is installed in the management server; The meteorological data monitoring server obtains the current meteorological data in a certain indicator area, and when it is determined that the current meteorological data in the indicator area is abnormal, it will read from an insurance bank database corresponding to the industrial and commercial disaster assessment model The basic information of each insurer with independent variables, 嗣, according to the geographic coordinates of each insurer, read the current meteorological data corresponding to the geographic coordinates of each insurer from the meteorological data monitoring server, so that the management server can The basic information of the insurer and the current weather data corresponding to the geographic coordinates of each insurer are imported one by one to the worker. Disaster assessment model, so as to estimate the current industrial and commercial disaster assessment value of each insurer in real time. The management server judges that the industrial and commercial disaster assessment value and flood depth assessment value of the geographical location of each insurer are separately When it exceeds a preset gate value, the management server will mark each area that exceeds the preset gate value as an early warning area, and it will automatically register as an insurer and the geographic coordinates are distributed in the system. Each insurer in the warning area is marked as a warning insurer, so as to send an early warning signal to each terminal device to which each of the warning insurers belongs via the Internet, reminding each insurer to be warned Prepare for disasters to prevent disasters before they occur, so as to effectively and significantly reduce the potential damage to business and industry.

In order that the review committee can further understand and understand the implementation principles, structural features, and objectives of the present invention, the embodiments are described in detail with the drawings, as follows:

[Learning]

no

〔this invention〕

10‧‧‧Terminal device

20‧‧‧Internet

30‧‧‧ Management Server

40‧‧‧Meteorological data monitoring server

51‧‧‧Insurer Database

52‧‧‧Sample historical database of industrial and commercial disasters

Industrial and commercial disaster data corresponds to the current weather data

53‧‧‧ historical database

55‧‧‧Address and Geographic Coordinate Database

70‧‧‧ early warning area

80‧‧‧Pending insurance

90‧‧‧ communication system

100 ~ 106‧‧‧ steps

200 ~ 209‧‧‧ steps

Fig. 1 is a schematic diagram of the system architecture of the present invention; Fig. 2 is a process of constructing an industrial and commercial disaster assessment model in each region based on a historical database of historical and industrial disaster disaster data samples in a preferred embodiment of the present invention Schematic diagram; Figure 3 is an annual listing of the names of disaster events included in the business disaster historical data sample database; Figure 4 is a sample of claims data sampled in the business disaster historical data sample database. Schematic diagram of industrial classification; Figure 5 is a schematic diagram of the sample distribution number of claims in the industrial and commercial disaster historical data sample database in Taiwan; Figure 6 is a schematic of the industrial and commercial disaster historical data sample database. Annual statistical diagram of the number of disaster samples; Figure 7 is a statistical diagram of the types and proportions of the disaster samples covered by the industrial and commercial disaster historical data sample database; Figure 8 is a sample of the industrial and commercial disaster historical data The statistical schematic diagram of the disaster samples covered by the database in the type and proportion of the disaster; Figure 9 shows the disaster samples covered by the business disaster historical data sample database. Statistics and schematic diagrams of disaster damage items and proportions; Figure 10 is a statistical diagram of the amount and proportion of disaster damage samples covered by the business disaster historical data sample database; Figure 11 is the invention Schematic diagram of the processing flow of a system capable of real-time assessment and early warning of industrial and commercial damage caused by wind and flood disasters; FIG. 12 is a schematic diagram of the system of the present invention labeling each early warning area and the insurance insurers to be warned; and FIG. 13 is a schematic diagram of the system of the present invention sending an early warning message to each insurer to be warned.

In the past years, during the typhoon-affected period, the economic disaster losses (hereinafter referred to as "disaster damages") of the industrial and commercial and service industries in various regions of Taiwan were related to the business interests of the respective business units. Data samples are often classified as company secrets and are not disclosed to the public, which has made relevant academic and industrial circles interested in conducting disaster assessment and research for a long time due to the lack of basic information on industrial and commercial disasters, which has been difficult to construct. Specific assessment models and systems for business disasters. According to this, the inventor is thinking about cooperating with "Fubon Products Insurance Company" (hereinafter referred to as "Fubon Property Insurance"), which has a long history in Taiwan and has business items and product insurance for business services. Negotiations are a formal cooperation agreement reached in 2015 to jointly conduct research and development on real-time assessment and early warning systems for industrial and commercial disasters caused by wind and floods in various regions of Taiwan.

Please refer to FIG. 1. The present invention is a system capable of real-time assessment and early warning of industrial and commercial disasters caused by wind and flood disasters. The system includes a plurality of terminal devices (such as smart phones or computers). Network 20, a management server 30, and at least one meteorological data monitoring server 40 (such as the instant meteorological server of the Central Meteorological Bureau and the flood warning range server of the Water Conservancy Department), wherein each of the terminal devices 10 is The insurers (or "members") who log in to the system can be individually owned by individuals, companies, or their employees, and can connect to the management server 30 through the Internet 20 respectively. The management server 30 is installed with at least one industrial and commercial disaster assessment model (such as an industrial and commercial wind disaster assessment model or an industrial and commercial flood assessment model), and the management server 30 can communicate with each of the meteorological data monitoring servers separately. 40. An insurance account database 51. One worker The business disaster damage historical data sample database 52 and an address are connected to the geographic coordinate comparison database 55. The present invention is constructing the system so that the system can use it to evaluate the industrial and commercial disasters in various regions of Taiwan. The industrial and commercial disaster assessment model is mainly constructed according to the following processing steps shown in Figure 2 respectively.

(100) First, establish the industrial disaster historical data sample database 52: During the research and development cooperation between the inventor and Fubon Property & Casualty Insurance, Fubon Property & Casualty Insurance specialized in product insurance agencies over the years Accumulate a large amount of disaster damage data sample data for applying for insurance claims due to typhoon-induced floods and industrial and commercial disasters, and provide it to the inventor, and input each disaster data sample to the computer one by one according to the data format designated by the inventor. In order to construct the industrial and commercial disaster historical data sample database 52, the database 52 contains at least the basic information of the affected households (such as: industry type, floor use), the affected households' disaster experience (such as: disaster event, date of disaster) And the amount of loss (such as: the amount of building damage, the amount of decoration loss, the amount of loss of machinery and equipment, and the amount of downtime loss, etc.), please refer to Figure 3, the industrial and commercial disaster historical data sample database 52 is covered From 2001 to 2016, a total of 1,293 data samples of industrial and commercial disaster claims caused by 40 typhoons or heavy rains in 16 years, of which, please refer to Figure 4 According to the number of samples in the industry, there are 274 in the industry and 1,019 in the business, and the number of samples distributed in various regions is shown in Figure 5. Taipei City and New Taipei City had the highest numbers of about 250, Taoyuan City and Taichung City had about 180 and 100, and the counts in other counties were less than 100. Please refer to Figure 2. For the establishment of the industrial and commercial disaster historical data sample database 52, the inventor and Fubon Property & Casualty Insurance, in order to consider the confidentiality of the sample data, recorded the data of each case. In addition to only representing the affected households with unique numbers, the affected area is only based on the area (or road section) where the disaster occurred (for example, New Taipei City, New Taipei City) No. 1-300, Beixin Road, Dianqu), and do not record a clear address; (101) Each industrial and commercial disaster data sample in the industrial and commercial disaster historical data sample database 52 and the then Meteorological data is aggregated and merged: after the establishment of the industrial and commercial disaster historical data sample database 52 is completed, because the location of each industrial and commercial disaster data is recorded in the area of the occurrence, without a clear address, Therefore, when the current meteorological data provided by the Central Meteorological Bureau is further expanded and the business disaster historical data database 52 is expanded, it is only necessary to replace the area closest to the area at the time when typhoon disasters provided by the Central Meteorological Bureau in the past years. The meteorological data measured by the meteorological monitoring station (such as 24-hour accumulated rainfall and maximum wind speed) are imported into the meteorological data field corresponding to each disaster case in the area, which can be included in the historical data of the industrial and commercial disaster The sample database 52 expands the relationship between each disaster damage claim sample and the corresponding area range and its current meteorological data, thereby expanding the A historical sample database 53 corresponding to the meteorological data at that time was produced; (102) Descriptive statistics on the industrial and commercial disaster data in each area of the historical sample database 53 corresponding to the then-time meteorological data (descriptive statistics): According to the information in the historical sample database 53 of the corresponding meteorological data of the industrial and commercial disaster data at the time, descriptive statistics are made on the industrial and commercial disaster data in each region to understand the basics of the industrial and commercial disaster sample. Statistical data. A special mention here is that because there are many disaster-stricken samples in New Taipei City and Taipei City, the statistical results can represent the historical industrial and commercial disasters in the Shuangbei area. Therefore, this step uses the samples in the Shuangbei area as an example to describe statistics. Chen Ming first. In the aforementioned sample data of claims for industrial and commercial disasters, please refer to Figure 6. The typhoon-induced floods and disasters caused by the typhoon, and the typhoon-induced floods and wind disasters in 2007, 2012, and 2013, respectively. The number of disaster damage samples is large, accounting for nearly 70% of the total number of disaster damage samples, and the corresponding typhoon events are in sequence Typhoon Sha, Typhoon Sura, and Typhoon Suli, etc. In terms of industries, please refer to Figure 7; the number of samples in the commercial and service industries is large, accounting for nearly 90% of the total sample of disasters. The disaster type is shown in Figure 8. It is mainly caused by wind disasters, accounting for more than 80%; (103) 俟 After descriptive statistics are completed, the independent variables in the business disaster assessment model (independent) can be defined. variables (also called "previous variables") and dependent variables (or "consequence variables"): the inventor's historical sample database corresponding to the weather data at the time After performing correlation analysis, comparison and collation of each industrial and commercial disaster data sample in 53 regions, it was clearly found that the "claim amount" in each industrial and commercial disaster data sample should be set as "consequence variable", And set its "basic data" (such as: industry type, area, floor and total insurance amount) and the current weather data (such as 24-hour cumulative rainfall and maximum wind speed) as "pre-variables", so, From which One of the feasible business and industry damage assessment models, so that the business and industry damage assessment model can be used as an assessment tool for developing disaster early warning and disaster prevention strategies. In the present invention, in order to avoid that the "claim amount" does not conform to the normal distribution, The business disaster damage data corresponds to each disaster damage claim sample in the historical sample database 53 of the meteorological data at that time. When data analysis is performed, the relevant assumptions of multiple regression and correlation analysis are violated. Therefore, the subsequent multiple regression equation is executed in the present invention. In the process of calculation and analysis, the inventor specially calculated the “claim amount” of each disaster loss claim sample by taking log10 to make it more consistent with the normal distribution, thereby effectively increasing the statistical result of the multiple regression equation. In addition, the present invention mainly uses Pearson's correlation coefficient to perform correlation analysis to find the important factors (hereinafter referred to as "disaster factors") that cause disasters in business and services, and then starts with multivariate regression. (multiple regression) Analysis and operation of equations to construct a business and industry disaster assessment model (such as: a business and industry disaster assessment model or a business and industry disaster assessment model); The industrial and commercial disaster data corresponded to the historical sample database 53 of historical meteorological data at the time. The inventor believes that the industrial and commercial disaster data corresponds to the historical sample database 53 of the current weather data. The number of samples caused by ground flooding and basement flooding is small, so before considering the representativeness of the samples, in the present invention, the damage caused by wind damage is mainly used to construct the business damage. The data corresponds to one of the industrial and commercial wind damage assessment models corresponding to the historical sample database 53 of the meteorological data at that time. After collecting enough samples of the industrial and commercial damage caused by ground flooding and basement flooding in the future, the same can be followed. Procedures and models, a corresponding sample database of historical data on industrial and commercial disasters and its corresponding model of industrial and commercial flood damage assessment are established together.

Alas, after the inventor conducted statistics and analysis on each data sample in the historical sample database 53 corresponding to the meteorological data at that time, in view of the damage caused by strong winds to the business and service industries, please refer to As shown in Figure 9, it is mainly the auxiliary structures of the building (such as signboards, doors and windows and their glass, awnings or shelters covered on the top floor, etc.), followed by interior / office decoration and equipment; as for wind damage Regarding the proportion of the amount, please refer to Figure 10, the interior / office decoration and equipment accounted for a maximum of 59% of the amount of the Prime Minister's compensation. In addition, the industrial and commercial disaster data corresponds to the current weather data. After correlation research and analysis of each sample data in the historical sample database 53, it is also clearly shown that as shown in Table 1 below, the total insurance amount and claim amount of each claim sample and each hazard factor (such as Typhoon 24 There is obviously a positive correlation between hourly accumulated rainfall, typhoon maximum wind speed, and business building height. In other words, the greater the maximum wind speed, the higher the business building height, and the total insurance premium. The larger the amount, the greater the amount of claims received;

Although the aforementioned correlation coefficient shows the correlation strength between the two hazard factors (or "pre-variables", independent variables) and the amount of claims (or "consequential variables", dependent variables), it cannot display multiple disasters The relationship between factors and the amount of claims; (104) Performing analysis and calculation of multiple regression equations: Before performing analysis and calculation of multiple regression equations in the present invention, each of the disaster-causing factors shown in Table 1 is used as a multiple regression. The "preliminary variable" of the equation, and the "consumption amount" as the "consequence variable" of the multiple regression equation, and based on the business disaster data corresponding to each of the historical sample database 53 of the meteorological data at that time The "claim amount" corresponding to the claims sample and each of the "causative factors" are imported into the multiple regression equation one by one, and operations are performed, that is, after each coefficient in the multiple regression equation is obtained, the so-called "invention" can be constructed. One of the industrial and commercial wind damage assessment models; although, the hazard factors shown in Table 1, such as: 24-hour accumulated rainfall by typhoon, industry type and whether to use the basement? Related In the analysis, it did not reach a significant standard, but these hazard factors can still be used as “pre-variables” and imported into the business wind and damage assessment model referred to in the present invention, and it is clearly shown that The results of its multiple regression evaluation are shown in Table 2 below. The performance in the coefficient direction is still consistent with that shown in Table 1 above, that is, the "24-hour cumulative rainfall of typhoons", "industry category" and "whether used" Basement? "And other hazard factors, the" Typhoon maximum wind speed "," Business building height "and" Total insurance amount "can still be used to predict the amount of disaster damage claims, so the three hazard factors are still Important assessment factors for disaster assessment.

Because the coefficients in the multiple regression equation are related to the mutually confidential business secrets between the inventor and Fubon Property and Casualty Insurance, in the present invention, the inventor is unable to calculate the results of the multiple regression equation and the damage caused by the industrial and commercial storm. The coefficients in the evaluation model are made public, and only the characteristics of the coefficients can be listed with symbols and signs for the understanding of the amount of claims and the various hazard factors. Positive or negative correlation. In addition, since the present invention is the research result of the cooperation between the inventor and Fubon Property and Casualty Insurance, it is possible to borrow valuable data samples of Taiwan ’s industrial and commercial and service industry typhoon damage claims accumulated by Fubon Property and Casualty Insurance over the years. To construct the wind and rain damage assessment model required by the present invention, however, based on the disaster damage compensation samples, it is mainly attributed to wind damage losses, and from the results of the damage compensation claims, it is also found that strong winds most often cause damage to the business and service industries. The loss item is the destruction of the auxiliary structure of the building. The interior / office decoration has a greater impact on the amount of loss. Based on this, the data correlation and multiple regression analysis and calculation are used to more clearly show the maximum wind speed, building height, and total insurance. The amount is also an important factor that can be used to evaluate the damage caused by wind disasters. Based on the foregoing research results, it is clear that during the typhoon approach, the hazard factor that the industrial and commercial and service industries need to continue to pay attention to may be the maximum wind speed of the typhoon. Corresponding disaster preparedness and disaster prevention measures shall be carried out for the auxiliary structures of the buildings and the interior of the business place, so as to minimize the disaster damage.

According to the present invention, when performing the analysis and calculation of the multiple regression equation, the above-mentioned dependent variables and independent variables corresponding to each disaster sample in the business disaster history sample database 53 are sequentially and one by one imported. To one of the following multiple regression equations (1): In the following multiple regression analysis and calculation of the present invention, it is mainly based on "the expected value of the dependent variable Y is a function of the independent variables X _i , i = 1, 2, ..., k, and ε _i is an independent random error. "Variables" theoretical basis, so, according to the linear mode, it is assumed that there are k + l variables Y and X ₁ , X ₂ , ..., X _k ; where X _i is k independent variables; Y is The dependent variable is a variable that changes randomly according to the change of the independent variable X _l . Therefore, the above equation (1) can also be expressed in a matrix as shown in the following equation (2):

Use the Least Squares method to find the minimum error value of the random error variable ε _i , that is, ; 嗣, perform partial differential analytic operations on B ₀ , B ₁ , ..., B _k , respectively, to obtain the following equations (3) and (4):

嗣, calculate the parameters B ₀ , B ₁ , ..., B _k in the above equations (3) and (4), and set b ₁ , b ₂ ..., b _k as the least squares of the parameters, respectively. Multiply the estimated value, and let Q (B ₀ , B ₁ , ..., B _k ) be zero (minimum), and then obtain the following normal equations (5):

Thus, it is clear that in the normal equation (5) listed above, the coefficient matrix is a symmetric matrix, where A represents the coefficient matrix and B represents the right-side constant term matrix, as shown in the following equations (6) and (7), respectively. :

According to this, the matrix type of the normalization equation (5) becomes as follows: (X 'X) b = X' Y or Ab = B, where b = (b ₀ , b ₁ ..., b _k ) Is the unknown of the normalized equation (5). Under the full rank of the coefficient matrix A, the inverse matrix of (XX) exists, so the coefficient b can be obtained from the following equation (8): b = ( X'X ) ^-1 X'Y (8), and via the above equation (8), and one of the following multivariate linear models Y = b ₀ + b ₁ X ₁ + b ₂ X ₂ + b ₃ X ₃ + ... + b _K X _K (9 )

(105) Determine whether the damage assessment model can pass the statistical hypothesis test? If yes, continue to the following steps; otherwise, return to step (102), readjust the settings of each "independent variable", and continue to perform the multiple regression analysis and calculation of the present invention; and (106) obtain a disaster loss assessment model (ie, Equation (9)).

According to statistical research based on social sciences, in order to find out the relationship between disaster losses and hazard factors, multiple regression is often used to construct a prediction model so that in the future, it can be based on the current hazard factors. To assess the current disaster losses in real time, in which the meaning of multiple means that this prediction model contains multiple independent variables (or "preliminary variables", that is, disaster factors). The present invention The multiple regression method is mainly used to understand the quantitative relationship between the dependent variable (also called "consequence variable", that is, disaster loss) and multiple independent variables, and find out the dependent variable. These independent variables are significantly related, so that in addition to the mathematical In addition to the disaster damage prediction model, it is expected that in the future, by collecting different quantified independent variables (24-hour accumulated rainfall of typhoons), and then through the same construction procedures and models of the wind damage prediction model, a mathematically based Evaluate the forecast model of industrial and commercial flood damage to predict or estimate the possible impact of 24-hour cumulative rainfall on industrial and commercial losses.

Accordingly, the inventor hereby specifically declares that because of the multiple regression theory itself, the multiple regression equations used in the multiple regression equations, and their calculations and analysis, are all existing conventional techniques and theories before the present invention, rather than The key points of protection claimed by the present invention; on the contrary, the core of the technology claimed by the present invention lies in the data in each sample in the historical sample database 53 corresponding to the current weather data corresponding to the industrial and commercial disaster data. After summarizing, statistic and collating the interrelationships, it is determined that “industrial storm damage” and “industrial flood damage” are used as “dependent variables”, respectively, and the geographical coordinates of each area (such as latitude and longitude) and each area are determined. Corresponding meteorological data (such as 24-hour accumulated rainfall and maximum wind speed), "industry category", "used floor", and "total insurance amount", etc. are set as "independent variables", respectively. The corresponding data in each sample in the industrial and commercial disaster history sample database 53 is imported into at least two multiple regression equations, Through the calculation of the equations, a business and industry damage assessment model is constructed separately. Please refer to Figure 1 and install each of the wind or flood damage assessment model in the management server 30. One of the inventions is an instant disaster damage assessment and early warning system, which enables the management server 30 to perform the following steps, please refer to FIG. 11 to implement the assessment of the current industrial and commercial wind or flood damage, and execute the disaster accordingly. Early warning of damage and flooding: (200) The management server 30 will be connected to at least one weather data monitoring server 40 (such as the real-time weather data monitoring server of the Central Meteorological Bureau) through the Internet 20, And since The meteorological data monitoring server 40 reads real-time meteorological data (such as "24-hour cumulative rainfall data" and / or "maximum wind speed data") issued by the meteorological data monitoring server 40; (201) in the management server The device 30 judges whether the current meteorological data of a certain indicator area (such as the Xindian District of Taipei City) is abnormal? When it is judged that the current meteorological data in the indicator area is abnormal (such as: "24-hour accumulated rainfall" is greater than the historical average accumulated rainfall or "maximum wind speed" is greater than the historical average wind speed), step (202) is executed; otherwise, return Step (200), read the real-time weather data released by the meteorological data monitoring server 40 in the next period; (202) the management server 30 will be connected to the insurer database 51 to obtain information from the insurer. In the database 51, read the basic information (such as geographic coordinates, industry type, floor used and total insurance amount, etc.) of each insurer corresponding to the respective variables in the business disaster assessment model; (203) the management server 30 Based on the geographic coordinates of each insurer, the current meteorological data corresponding to the geographic coordinates of each insurer will be read from each of the meteorological data monitoring servers 40 (e.g., the current "24-hour accumulated rainfall data corresponding to each of these geographic coordinates" "And / or" maximum wind speed data "); (204) enables the management server 30 to collect the basic data of each insurer and the current weather data corresponding to the geographic coordinates of each insurer one by one To the industrial and commercial disaster assessment model; (205) to instantly estimate the current industrial and commercial disaster assessment value of each insurer; (206) the management server 30 will determine the current industrial and commercial disaster assessment of the geographical coordinates of each insurance household Has the value exceeded a preset gate value (such as the total insurance amount of the insurer or the historical average risk amount in the geographic coordinate area)? If yes, continue to step (207); otherwise, return to step (202); (207) When the management server 30 judges the industrial and commercial disaster damage assessment of the geographic coordinates of each insurer When the valuation has exceeded a preset gate value (such as the total insurance amount of the insurer or the historical average risk amount within the geographic coordinate area), please refer to Figure 12 and the management server 30 is The area of each geographic coordinate position exceeding the preset gate bar value will be marked as an early warning area 70, and the system will automatically log in to the system as an insurance and the geographic coordinate locations are distributed in each of the early warning areas 70. The insurer is marked as the insurer 80 to be warned; (208) The management server 30 will read the communication information (such as email address and mobile phone number) of each insurer 80 to be warned from the insurer database 51. Etc.) to send warning messages as shown in FIG. 13 to each terminal device 10 to which each of the to-be-insured insurance customers 80 belongs via the Internet 20 or a communication system 90 to remind the to-be-insured insurance The household 80 must perform disaster prevention preparations in advance to effectively prevent disasters before they occur, thereby greatly reducing the possibility of industrial and commercial disasters.

The above are only a few of the preferred embodiments of the present invention. However, when the present invention is actually implemented and applied, its features are not limited to this. In particular, it is not limited to being applied to the business service industry. Those skilled in the art, after considering the technical content of the present invention, can easily consider the equivalent changes without departing from the protection scope of the present invention.

Claims (6)

A system for real-time assessment and early warning of industrial and commercial disasters caused by wind and flood disasters. The system includes: a management server installed with an industrial and commercial disaster assessment model installed in the industrial and commercial disaster assessment model. The assessment model or an industrial and commercial flood damage assessment model can be based on the data samples of industrial and commercial disasters in each area during the typhoon impact period provided by each relevant unit and the samples of meteorological data at that time in each area. The data of the industrial and commercial disasters in the data sample are used as the dependent variables of one of the multiple regression equations corresponding to the industrial and commercial disaster assessment model, and the geographic coordinates in each disaster data sample, the corresponding meteorological data at that time, and the industry category The data of the floor and the total insurance amount used are the independent variables of the multiple regression equation corresponding to the evaluation model. Among them, the meteorological data is the 24 hour accumulated rainfall data and / or maximum wind speed corresponding to each geographic coordinate. Data, alas, each of the data is imported into the corresponding multiple regression equations, and the equations are solved by Calculation, that is, the industrial and commercial disaster assessment model for each area can be constructed separately; an insurance account database, which is connected to the management server, and stores the basic data of multiple insurance accounts registered in the system Each basic information includes at least the geographic coordinates, industry type, floor and total insurance amount of each insurer; the Internet; at least one weather data monitoring server, which can publish an Weather data, and can be connected to the management server through the Internet to transmit real-time weather data to the management server; and a plurality of terminal devices, which are insurance companies registered in the system Owners, and can be connected to the management server through the Internet respectively; 管理 The management server obtains the current status of a certain indicator area from each of the weather data monitoring servers. Weather data, and if it is determined that the current weather data in the indicator area is abnormal, it will read the basic data of each insurer corresponding to the independent variable in the business disaster assessment model one by one from the insurer database; 嗣, according to The geographic coordinates of each insurer read the current meteorological data corresponding to the geographic coordinates of each insurer from each of the meteorological data monitoring servers, so that the management server can combine the basic information of each insurer and the geography of each insurer. The current meteorological data corresponding to the coordinates are imported one by one into the industrial and commercial disaster assessment model, so that the current industrial and commercial disaster assessment value of each insurer can be estimated in real time, and the management server determines the geographic coordinates of each insurer. When the industrial and commercial disaster assessment value in the location area has exceeded a preset gate value, the management server will mark each area that exceeds the preset gate value as an early warning area, and automatically Until the system becomes an insurer and the geographic coordinates of the insurer are distributed in each of the early warning areas, it is marked as a to-be-insured insurer, through the Internet or a Fusion Systems, one by one to each of the each of the terminal devices to be relevant to the user warning safety early warning message. According to the system described in claim 1, the industrial and commercial disaster damage assessment model is the industrial and commercial wind damage assessment model, wherein the multiple regression equation corresponding to the industrial and commercial wind damage assessment model uses the industrial and commercial damage caused by the wind disaster as the multiple regression The dependent variables of the equation, and the industrial and commercial disasters include items caused by wind damage, building losses, decoration losses, mechanical equipment losses, and shutdown losses. The independent variables of the multiple regression equation are the geographic coordinates of the insurer, the geographic The weather data corresponding to the coordinates at that time, the industry type of the insurer, the floor used and the total insurance amount. According to the system described in claim 1, the industrial and commercial disaster assessment model is the industrial and commercial flood assessment model, wherein the multiple regression equation corresponding to the industrial and commercial flood assessment model uses the industrial and commercial disaster caused by the flood as the multiple regression The dependent variables of the equation, and the industrial and commercial disasters include items such as the amount of building damage caused by the flood, the amount of damage to the decoration, the amount of machinery and equipment, and the amount of downtime. The independent variables of the multiple regression equation are the geographic coordinates of the insurer, the geographic The weather data corresponding to the coordinates at that time, the industry type of the insurer, the floor used and the total insurance amount. The system according to any one of claims 2 to 3, further comprising a historical sample database of industrial and commercial disaster data, wherein the preset gate value is the geographic location of each insurer in the historical sample database of industrial and commercial disaster data. Historical average risk amount for a region. The system according to any one of claims 2 to 3, wherein the preset gate value is the total insurance amount of each insurer. The system according to any one of claims 2 to 3, wherein the preset gate value is the highest total insurance amount within the geographic coordinate area of the location.