Multi-disaster risk early warning method and system using weather forecast data

Abstract

The invention relates to a multi-disaster risk early warning method and system using weather forecast data, and the method comprises the steps: storing the weather forecast data of an electric energy meter to construct a database, the weather forecast data comprising a precision grade, a weather data parameter sequence and a weather data increment; dividing the database into sub-databases according to the precision levels, and clustering meteorological data parameter sequences in the sub-databases to obtain clusters of all the precision levels; acquiring real-time data detected by the meteorological data acquisition instrument; calculating the sub-error of the target precision grade based on the clustering cluster of the target precision grade and the real-time data, and integrating all the sub-errors of the target precision grade to obtain a real-time error, the target precision grade being equal to or greater than the precision grade of the meteorological data acquisition instrument; and comparing the real-time error with the precision grade of the meteorological data acquisition instrument to obtain an error verification result. According to the technical scheme, the error checking efficiency of the electric energy meter is improved, and the hysteresis problem of checking time is solved.

Description

Technical Field

[0001] The present application generally relates to the field of multi-disaster risk warning, and more particularly to a multi-disaster risk warning method and system using meteorological forecast data. Background Art

[0002] Meteorological disasters may trigger secondary disasters. For example, heavy rainfall may trigger secondary disasters such as flash floods, mud-rock flows, and landslides. Therefore, the data flow of meteorological data information is of great significance for the early warning of secondary disasters such as flash floods, mud-rock flows, and landslides.

[0003] The method and system for multi-hazard risk early warning using meteorological forecast data is a comprehensive solution that combines meteorology, computer science and data analysis technology. This system can predict the risks of various natural disasters (such as floods, typhoons, rainstorms, lightning, etc.) in advance and issue early warnings in time to reduce the impact of disasters on human life and property.

[0004] At present, on the one hand, the meteorological department will produce meteorological disaster forecast maps for the next N hours based on relevant meteorological data and corresponding models, and the meteorological department and relevant departments (such as natural resources department, water conservancy department, etc.) jointly produce geological disaster meteorological risk warning maps, mountain torrent disaster meteorological risk warning maps, flood disaster meteorological risk warning maps, etc. These maps are only pictures and difficult to carry out secondary development; on the other hand, due to data confidentiality, authority and other issues, it is difficult to obtain the real-time meteorological forecast data of the meteorological department, which makes it difficult for other departments to use real-time meteorological forecast data for disaster warning dynamic analysis and emergency command services.

[0005] Therefore, how to conduct secondary development and intelligent data analysis and processing based on public meteorological forecast data to serve dynamic disaster warning, emergency command, etc. has become a technical problem that needs to be solved urgently. Summary of the invention

[0006] In order to solve the above-mentioned technical problems in the prior art, the present application provides a multi-disaster risk warning method and system using meteorological forecast data, which can perform statistical analysis and warning based on previous meteorological forecast data and actual meteorological conditions as well as the frequency and severity of natural disasters, thereby solving the problem that it is difficult to obtain real-time meteorological forecast data from the meteorological department, which makes it difficult for other departments to use real-time meteorological forecast data for dynamic analysis of disaster warning and emergency command services. It realizes the secondary development of meteorological forecast data and intelligent analysis and processing of data, and provides necessary technical support for multi-disaster risk warning and emergency command.

[0007] The present invention provides a multi-disaster risk early warning method and system using meteorological forecast data, comprising: storing meteorological forecast data to construct a database, wherein the meteorological forecast data includes the accuracy level of the meteorological forecast data, a meteorological forecast data sequence within a preset time scale, and a disaster scale sequence, and the disaster scale sequence includes at least a disaster occurrence frequency sequence and a disaster occurrence type sequence; dividing the database into at least one sub-database according to the accuracy level, clustering the disaster occurrence frequency sequence in the sub-database to obtain a cluster cluster corresponding to each accuracy level; collecting meteorological forecast real-time data, wherein the real-time data includes the accuracy level of the meteorological forecast, a real-time meteorological forecast data sequence within a preset time scale corresponding to the current moment, and disaster scale data; calculating a sub-error of the target accuracy level based on the cluster cluster corresponding to the target accuracy level and the real-time data, and performing weighted summation on the sub-errors of all target accuracy levels to obtain a real-time error of the disaster scale sequence corresponding to the meteorological forecast data, wherein the target accuracy level is greater than or equal to the accuracy level of the meteorological forecast data; comparing the real-time error with the accuracy level of the corresponding predicted disaster scale sequence to obtain an error verification result.

[0008] In some embodiments, the disaster scale sequence also includes at least one of a temperature data sequence, a wind data sequence, a drought data sequence, and a precipitation data sequence when a disaster occurs.

[0009] In some embodiments, the meteorological data sequences in the sub-database are clustered to obtain clusters corresponding to each precision level, including: performing dimensionality reduction processing on all meteorological data sequences in any sub-database to obtain corresponding low-dimensional representations; clustering the low-dimensional representations of all meteorological data sequences according to a density clustering algorithm to obtain clusters of the precision level corresponding to the sub-database; wherein the density clustering algorithm is the HDBSCAN algorithm or the DBSCAN algorithm.

[0010] In some embodiments, the cluster cluster includes a cluster center and multiple meteorological data sequences, and the sub-error of the target accuracy level is calculated based on the cluster cluster corresponding to the target accuracy level and the real-time data. The calculation includes: for any cluster cluster corresponding to the target accuracy level, calculating the similarity between its cluster center and the real-time meteorological parameter sequence; based on the similarity, screening the cluster clusters corresponding to the target accuracy level to obtain the target cluster cluster; for any target cluster cluster, calculating the cluster error of the target cluster based on the real-time meteorological data information and the disaster data corresponding to multiple meteorological databases, wherein the multiple meteorological databases include the disaster data corresponding to each meteorological data sequence in the target cluster cluster; and calculating the sub-error of the target accuracy level based on the cluster errors and similarities of all target cluster clusters.

[0011] In some embodiments, the screening of the cluster clusters corresponding to the target accuracy level based on the similarity to obtain the target cluster cluster includes: in response to the similarity corresponding to any cluster cluster being greater than or equal to a similarity threshold, the cluster cluster is used as the target cluster cluster; in response to the similarity corresponding to any cluster cluster being less than the similarity threshold, the cluster cluster is deleted.

[0012] In some embodiments, for any target cluster, the cluster error of the target cluster is calculated based on the real-time meteorological data information and multiple meteorological databases, including: using an outlier detection algorithm to detect the multiple meteorological databases to obtain outliers of each meteorological data information; based on the outliers, the multiple meteorological databases are screened to obtain target disaster data information; based on the real-time meteorological data information, the target meteorological data information and the outliers of each target meteorological data information, the cluster error of the target cluster is calculated, and the cluster error satisfies the relationship:

[0013] in, is the number of target meteorological data information in the target cluster a, and are the mth target meteorological data and the outlier of the mth target meteorological data in the target cluster a, respectively. is the real-time meteorological data, is the sum of outliers of all target meteorological data in the target cluster a, Represents the cluster error of the target cluster a.

[0014] In some embodiments, the screening of the multiple meteorological databases based on the outliers to obtain target meteorological data increments includes: in response to the outlier corresponding to any meteorological data being less than an outlier threshold, taking the meteorological data as target meteorological data; in response to the outlier corresponding to any meteorological data being greater than or equal to the outlier threshold, deleting the meteorological data.

[0015] In some embodiments, the sub-error of the target accuracy level satisfies the relationship:

[0016] in, and are the similarity and cluster error of the target cluster a respectively, N is the number of target clusters corresponding to the target accuracy level, is the sum of the similarities of all target clusters corresponding to the target accuracy level, is the sub-error of the target accuracy level k.

[0017] In some embodiments, the weighted sum of the sub-errors of all target accuracy levels to obtain the real-time error of the disaster data corresponding to the real-time meteorological data collected by the meteorological data collector includes: calculating the reference degree of each target accuracy level, and the reference degree satisfies the relationship:

[0018] in, is the sum of all target accuracy levels, is the reference degree of the target accuracy level k; based on the reference degree, the sub-errors of all target accuracy levels are weighted and summed to obtain the real-time error of the meteorological data.

[0019] In some embodiments, the error check result includes normal and abnormal, and the method further includes: in response to the error check result being normal, storing the real-time data of the meteorological conditions to be measured in the database.

[0020] The above-mentioned multi-disaster risk warning method and system using meteorological forecast data provided in the embodiment of the present application stores meteorological forecast data of different accuracy levels to build a database, compares the meteorological forecast data corresponding to the target accuracy level with the collected real-time meteorological data, and calculates the sub-errors of each target accuracy level, wherein the target accuracy level is greater than or equal to the accuracy level of the measured meteorological data; the real-time error of the measured meteorological data at the current moment is obtained by combining the sub-errors of all target accuracy levels; the error verification result is automatically obtained based on the real-time error, thereby realizing the secondary development of meteorological forecast data and intelligent data analysis and processing, and providing necessary technical support for multi-disaster risk warning and emergency command.

[0021] Furthermore, when calculating the sub-errors of each target accuracy level, the meteorological forecast data are divided into multiple clusters according to the meteorological data sequence in the meteorological forecast data, and each cluster corresponds to a meteorological condition; according to the similarity between the cluster center in the cluster and the real-time meteorological parameter sequence, the target cluster with the same meteorological condition as the real-time meteorological parameter sequence is determined, the cluster error of each target cluster is calculated, and the reference value of the cluster error of each target cluster for the real-time error calculation is evaluated according to the similarity, so as to improve the accuracy of the real-time error.

[0022] Furthermore, in the process of calculating the error of each target cluster, considering that the meteorological forecast data corresponding to the high-precision meteorological data may also have errors, the outlier value of each meteorological data in the target cluster is obtained according to the outlier detection algorithm, and the reference value of each meteorological forecast data for the error calculation of the real-time meteorological data is evaluated according to the outlier value. The reference value of the meteorological data is , thereby improving the accuracy of real-time meteorological data errors.

[0023] Furthermore, when the error check result is normal, the real-time data of the measured meteorological data is stored in a database, and the database is continuously updated, thereby continuously improving the accuracy of the error check result of the electric energy meter.

[0024] Furthermore, error verification of natural disaster prediction results corresponding to real-time meteorological data is achieved through meteorological data parameter sequences, wherein the meteorological data parameter sequences include multiple time series data that affect meteorological data errors, including the accuracy level of meteorological forecast data, meteorological forecast data sequences within a preset time scale, and disaster scale sequences. Error verification is achieved by combining multiple meteorological data parameters, thereby improving the accuracy of error verification results of meteorological forecast data for disaster risk warnings from the data source aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] By reading the detailed description below with reference to the accompanying drawings, the above and other purposes, features and advantages of the exemplary embodiments of the present application will become easy to understand. In the accompanying drawings, several embodiments of the present application are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein: Figure 1 It is a flow chart of a multi-disaster risk warning method and system using meteorological forecast data according to an embodiment of the present application. DETAILED DESCRIPTION

[0026] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are 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 those skilled in the art without creative work are within the scope of protection of the present application.

[0027] It should be understood that when the terms "first", "second", etc. are used in the claims, specification and drawings of the present application, they are only used to distinguish different objects, rather than to describe a specific order. The terms "include" and "comprise" used in the specification and claims of the present application indicate the presence of the described features, wholes, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or their collections.

[0028] According to the first aspect of the present application, the present application provides a multi-disaster risk early warning method and system using meteorological forecast data. Figure 1 The figure is a flowchart of a multi-disaster risk early warning method and system using meteorological forecast data provided by a preferred embodiment of the present application. According to different requirements, the order of the steps in the flowchart can be changed, and some steps can be omitted.

[0029] S1, storing weather forecast data to construct a database, wherein the weather forecast data includes the accuracy level of the weather forecast data, a weather forecast data sequence within a preset time scale, and a disaster scale sequence, and the disaster scale sequence includes at least a disaster occurrence frequency sequence and a disaster occurrence type sequence.

[0030] In an optional embodiment, the preset time scale is denoted as T, and for any meteorological data collected at time t, a meteorological data parameter sequence between time t and time t+T is obtained, and the real-time meteorological data parameter sequence includes a wind duration sequence and an air humidity duration sequence, wherein the wind duration sequence and the air humidity time series include wind force level values ​​and air humidity values ​​for each unit time, and the lengths of the wind duration sequence and the air humidity time series are the same and are related to the setting of the unit time; further, the increase in the number of meteorological data representations between time t and time t+T is taken as an increment of meteorological data, and the increment of meteorological data, the meteorological data parameter sequence and the accuracy level of the meteorological database are stored in the database as a meteorological forecast data, wherein the meteorological data parameter sequence can reflect the meteorological data situation within the preset time scale.

[0031] The meteorological data can be any one of all time periods in the forecast area, or any one of all forecast areas within a certain time period. The accuracy level of meteorological data forecast refers to the accuracy level that the meteorological data collector can guarantee when measuring meteorological data, usually expressed as a number, such as 1.0 or 0.5, etc., where a meteorological data collector with an accuracy level of 1.0 can guarantee that its measurement error does not exceed 1.0% under the premise of normal operation.

[0032] The unit time may be set to 1 second or 3 minutes, which is not limited in the present application; preferably, the preset time scale may be any time scale between 10 minutes and 60 minutes.

[0033] In another optional embodiment, changes in altitude, seasonal changes, the ambient temperature of the meteorological data collector, and changes in air humidity in the area to be measured may affect the accuracy of the meteorological data collected by the meteorological data collector. In order to improve the accuracy of subsequent error checking of the meteorological data collector, the meteorological data parameter sequence in the meteorological forecast data may also include at least one of an altitude time series, a temperature time series, and a seasonal factor time series.

[0034] For example, when a manufacturer needs to perform error verification on all meteorological data acquisition instruments put into use, due to the large temperature differences in the areas where different meteorological data acquisition instruments are located, the altitude time series, temperature time series and seasonal factor time series can be used as meteorological data parameter sequences in the meteorological forecast data to avoid environmental differences affecting the accuracy of the error verification results of the meteorological data collected by the meteorological data acquisition instruments.

[0035] In one embodiment, weather forecast data from different meteorological data collectors at the same collection time, weather forecast data from the same meteorological data collector at different collection times, or weather forecast data from different meteorological data collectors at different collection times are collected; information output by all meteorological data collectors is stored in a database to complete the construction of the database.

[0036] In this way, the construction of the database is completed, which includes meteorological forecast data corresponding to meteorological data collectors of different accuracy levels, providing a data basis for realizing error verification of meteorological data collectors.

[0037] S2, dividing the database into at least one sub-database according to the accuracy level, clustering the meteorological data parameter sequences in the sub-database to obtain clusters corresponding to each accuracy level.

[0038] In one embodiment, the database includes weather forecast data corresponding to data collectors of different accuracy levels, and all weather forecast data of the same accuracy level are used as sub-databases corresponding to the accuracy level, wherein the sub-databases correspond to the accuracy levels one by one. For example, all weather forecast data of accuracy level 1.0 in the database are used as sub-databases of accuracy level 1.0. If the accuracy levels include 0.5, 1.0, 1.5, and 2.0, a total of 4 types, 4 sub-databases corresponding to the accuracy levels can be obtained.

[0039] In one embodiment, each sub-database includes multiple meteorological forecast data of the same accuracy level, and the meteorological data parameter sequence in the meteorological forecast data can reflect the occurrence of disasters corresponding to the meteorological data within a preset time scale. Then, clustering the meteorological data parameter sequence in the sub-database can obtain clusters for each meteorological data situation, which is specifically described as follows. Clustering the meteorological data parameter sequence in the sub-database to obtain clusters corresponding to each accuracy level includes: performing dimensionality reduction processing on each meteorological data parameter sequence in any sub-database to obtain a corresponding low-dimensional representation; clustering the low-dimensional representations of all meteorological data parameter sequences according to a density clustering algorithm to obtain clusters of the sub-database corresponding to the accuracy level; wherein the density clustering algorithm is an HDBSCAN algorithm or a DBSCAN algorithm.

[0040] Among them, the dimensionality reduction processing can be implemented using the PCA algorithm or the autoencoder network. Through the dimensionality reduction processing, a low-dimensional representation corresponding to each meteorological data parameter sequence of the sub-database can be obtained. The amount of data of the low-dimensional representation is smaller than the meteorological data parameter sequence, thereby reducing the computational complexity of the density clustering algorithm and improving the clustering efficiency.

[0041] Among them, both the HDBSCAN algorithm and the DBSCAN algorithm do not need to set the number of clusters, and the number of clusters can be adaptively determined according to the distribution of the low-dimensional representation of all meteorological data parameter sequences. The number of clusters is greater than or equal to 1.

[0042] Each cluster includes a cluster center and at least one meteorological data parameter sequence, wherein the cluster center is the average value of the at least one meteorological data parameter sequence; the meteorological data parameter sequences in the same cluster belong to the same meteorological data situation.

[0043] In this way, by clustering the meteorological data parameter sequences in each sub-database, a cluster corresponding to each accuracy level can be obtained, and each cluster corresponds to a meteorological data situation.

[0044] S3, collecting real-time data detected by the meteorological data collector, wherein the real-time data includes the accuracy level of the meteorological data collector, the real-time meteorological data parameter sequence within a preset time scale corresponding to the current moment, and the increment of the real-time meteorological data.

[0045] In an optional embodiment, the meteorological data collector is any one of all meteorological data collectors in a preset area or all meteorological data collectors sold by a certain manufacturer that needs to be error checked. Among them, the preset time scale corresponding to the current moment is the preset time scale corresponding to the end moment with the current moment as the preset time scale, that is, the preset time scale corresponding to the current moment is ,in, For the current moment.

[0046] The real-time meteorological data parameter sequence and the meteorological data parameter sequence in the meteorological forecast data include the same meteorological data parameters. For example, when the meteorological data parameter sequence in the meteorological forecast data includes an altitude time series, a temperature time series, and a seasonal factor time series, the real-time meteorological data parameter sequence also includes an altitude time series, a temperature time series, and a seasonal factor time series.

[0047] S4, calculating the sub-error of the target accuracy level based on the cluster cluster corresponding to the target accuracy level and the real-time data, and performing weighted summation on the sub-errors of all target accuracy levels to obtain the real-time error of the meteorological data acquisition instrument to be tested, wherein the target accuracy level is greater than or equal to the accuracy level of the meteorological data acquisition instrument to be tested.

[0048] In an optional embodiment, a meteorological data acquisition instrument with an accuracy level greater than the accuracy level of the meteorological data acquisition instrument to be tested can obtain a more accurate meteorological data increment, which can be used to calculate the real-time error of the meteorological data detector to be tested. In addition, a meteorological data detector with an accuracy level equal to the accuracy level of the meteorological data detector to be tested can also provide a certain reference value for calculating the real-time error of the meteorological data to be tested. The present application uses all accuracy levels equal to or greater than the meteorological data detector to be tested as target accuracy levels for calculating the real-time error of the meteorological data detector to be tested; that is, the number of the target accuracy levels is at least one, which improves the accuracy of the real-time detection error of the meteorological data detector to be tested while making full use of the existing meteorological forecast data.

[0049] In one embodiment, the number of clusters corresponding to the target accuracy level is at least one, and each cluster includes a cluster center and multiple meteorological data parameter sequences. The sub-error of the target accuracy level is calculated based on the clusters corresponding to the target accuracy level and the real-time data. The method includes: for any cluster corresponding to the target accuracy level, calculating the similarity between its cluster center and the real-time meteorological data parameter sequence; screening the clusters corresponding to the target accuracy level based on the similarity to obtain the target cluster; for any target cluster, calculating the cluster error of the target cluster based on the real-time meteorological data increment and multiple meteorological database increments, wherein the multiple meteorological database increments include meteorological data increments corresponding to each meteorological data parameter sequence in the target cluster; and calculating the sub-error of the target accuracy level based on the cluster errors and similarities of all target clusters.

[0050] The similarity between the cluster center and the real-time meteorological data parameter sequence can be obtained by using existing similarity algorithms such as cosine similarity or Euclidean distance.

[0051] It should be noted that the multiple meteorological data parameter sequences in the clustering cluster can be the original meteorological data parameter sequences or the low-dimensional representations corresponding to the meteorological data parameter sequences; among them, when the multiple meteorological data parameter sequences in the clustering cluster are the low-dimensional representations corresponding to the meteorological data parameter sequences, the cluster center of the clustering cluster is the average value of all low-dimensional representations.

[0052] In one embodiment, the screening of the clusters corresponding to the target accuracy level based on the similarity to obtain the target cluster includes: in response to the similarity corresponding to any cluster being greater than or equal to a similarity threshold, taking the cluster as the target cluster; in response to the similarity corresponding to any cluster being less than the similarity threshold, deleting the cluster. The similarity threshold is 0.6.

[0053] It can be understood that when the similarity corresponding to any clustering cluster is greater than or equal to the similarity threshold, it means that the meteorological data situation corresponding to the clustering cluster is the same as the meteorological data situation of the real-time meteorological data parameter sequence, and the clustering cluster is used as the target clustering cluster for calculating the real-time error of the meteorological data acquisition instrument to be tested; when the similarity corresponding to any clustering cluster is less than the similarity threshold, it means that the meteorological data situation corresponding to the clustering cluster is significantly different from the meteorological data situation of the real-time meteorological data parameter sequence, that is, the clustering cluster has no reference value for the calculation of the real-time error of the meteorological data acquisition instrument to be tested, and it is discarded.

[0054] In one embodiment, for any target cluster, the cluster error of the target cluster is calculated based on the real-time meteorological data information and multiple meteorological databases, including: using an outlier detection algorithm to detect the multiple meteorological databases to obtain outliers of each meteorological data information; based on the outliers, the multiple meteorological databases are screened to obtain target disaster data information; based on the real-time meteorological data information, the target meteorological data information and the outliers of each target meteorological data information, the cluster error of the target cluster is calculated, and the cluster error satisfies the relationship

[0055] in, is the number of target meteorological data information in the target cluster a, and are the mth target meteorological data and the outlier of the mth target meteorological data in the target cluster a, respectively. is the real-time meteorological data, is the sum of outliers of all target meteorological data in the target cluster a, Represents the cluster error of the target cluster a.

[0056] Among them, the outlier detection algorithm is the LOF algorithm or the isolation forest algorithm, which is used to calculate the outlier value of each meteorological data increment in the target cluster. The outlier value can reflect the error degree of the meteorological data increment, and the outlier value is involved in the calculation of the cluster error to avoid the influence of the error of the high-precision meteorological data acquisition instrument on the cluster error.

[0057] It can be understood that all meteorological data increments in the same target cluster are collected by meteorological data collectors with the same accuracy level under the same meteorological data conditions. When there are no errors in all meteorological data collectors, all meteorological data increments in the same target cluster should converge at one point.

[0058] In one example, when the outlier value of the meteorological data increment is large, the distance between the meteorological data increment and other meteorological data increments in the same target cluster is far, that is, the error of the meteorological data increment is large, and it has no reference value for the calculation of the real-time error of the meteorological data collector, and it should be discarded; the specific description is as follows. The screening of the multiple meteorological data increments based on the outlier value to obtain the target meteorological data increment includes: in response to the outlier value corresponding to any meteorological data increment being less than the outlier threshold, the meteorological data increment is used as the target meteorological data increment; in response to the outlier value corresponding to any meteorological data increment being greater than or equal to the outlier threshold, the meteorological data increment is deleted. Among them, the value of the outlier threshold is 0.4.

[0059] In one embodiment, after calculating the cluster error of each target cluster, the sub-error of the target accuracy level is calculated based on the cluster errors and similarities of all target clusters; wherein, the greater the similarity corresponding to the target cluster, the greater the reference value of the cluster error of the target cluster to the error verification result of the meteorological data collector at the current moment, and the sub-error of the target accuracy level satisfies the relationship:

[0060] in, and are the similarity and cluster error of the target cluster a respectively, N is the number of target clusters corresponding to the target accuracy level, is the sum of the similarities of all target clusters corresponding to the target accuracy level, is the sub-error of the target accuracy level k; It can reflect the cluster error of the a-th target cluster Reference value for real-time error calculation.

[0061] In one embodiment, after obtaining the sub-errors of all target accuracies, the real-time error of the meteorological data corresponding to the disaster data corresponding to the real-time meteorological data collected by the meteorological data collector at the current moment can be obtained by integrating the sub-errors of all target accuracies, which is specifically described as follows. The weighted summation of the sub-errors of all target accuracy levels to obtain the real-time error of the meteorological data collector includes: calculating the reference degree of each target accuracy level, and the reference degree satisfies the relationship:

[0062] in, is the sum of all target accuracy levels, is the reference degree of the target accuracy level k; based on the reference degree, the sub-errors of all target accuracy levels are weighted and summed to obtain the real-time error of the meteorological data collector, and the real-time error of the meteorological data collector satisfies the relationship:

[0063] in, and are the reference degree and sub-error of target accuracy level k respectively, Q is the set of all target accuracy levels, is the real-time error of the meteorological data logger.

[0064] In this way, when calculating the sub-errors of each target accuracy level, the meteorological forecast data are divided into multiple clusters according to the meteorological data parameter sequence in the meteorological forecast data, and each cluster corresponds to a meteorological data situation; according to the similarity between the cluster center in the cluster and the real-time meteorological data parameter sequence, the target cluster with the same meteorological data situation as the real-time meteorological data parameter sequence is determined, the cluster error of each target cluster is calculated, and the reference value of the cluster error of each target cluster for the real-time error calculation is evaluated based on the similarity, thereby improving the accuracy of the real-time error; further, in the process of calculating the cluster error of each target cluster, the situation where there are errors in the meteorological forecast data is taken into account, and the reference value of each meteorological data increment in the target cluster for the real-time error calculation is evaluated based on the outlier detection algorithm, thereby further improving the accuracy of the real-time error.

[0065] S5, comparing the real-time error with the accuracy level of the meteorological data collector to obtain an error verification result.

[0066] In an optional embodiment, in response to the real-time error being greater than the accuracy level of the meteorological data collector, it indicates that the real-time error exceeds the maximum allowable error of the meteorological data collector, that is, the error check result of the meteorological data collector is abnormal; in response to the real-time error being not greater than the accuracy level of the meteorological data collector, it indicates that the real-time error does not exceed the maximum allowable error of the meteorological data collector, that is, the error check result of the meteorological data collector is normal. The accuracy level of the meteorological data collector corresponds to the maximum allowable error.

[0067] In response to the error checking result being abnormal, the staff is reminded to perform maintenance.

[0068] In an optional embodiment, in response to the error check result being normal, the real-time data of the meteorological data acquisition instrument is stored in the database to update the database; the updated database is used for the next error check of the meteorological data acquisition instrument to continuously improve the accuracy of the error check result of the electric energy meter.

[0069] In this way, the error verification result at the current moment is automatically obtained by comparing the real-time error and the accuracy level of the meteorological data acquisition instrument, which solves the problem of lag in verification time and does not require staff to conduct on-site inspection of the meteorological data acquisition instrument, thereby improving the efficiency of error verification of the meteorological data acquisition instrument; at the same time, when the error verification result is normal, the real-time data of the meteorological data acquisition instrument is stored in the database, and the database is continuously updated, thereby continuously improving the accuracy of the error verification result of the meteorological data acquisition instrument.

[0070] The multi-disaster risk warning method and system using meteorological forecast data provided in the embodiment of the present application stores meteorological forecast data of different accuracy levels to build a database, compares the meteorological forecast data corresponding to the target accuracy level with the real-time data of the measured electric energy meter, and calculates the sub-errors of each target accuracy level, wherein the target accuracy level is greater than or equal to the accuracy level of the measured electric energy meter; the real-time error of the measured electric energy meter at the current moment is automatically obtained by combining the sub-errors of all target accuracy levels, without the need for staff to conduct on-site inspection of the meteorological data acquisition instrument, thereby improving the efficiency of error verification of the meteorological data acquisition instrument and solving the problem of lag in verification time.

[0071] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0072] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be construed as limiting the scope of the patent application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent application shall be subject to the attached claims.

Claims

1. A multi-disaster risk early warning method and system using meteorological forecast data, characterized in that: Storing weather forecast data to construct a database, wherein the weather forecast data includes the accuracy level of the weather forecast data, a weather forecast data sequence within a preset time scale, and a disaster scale sequence, wherein the disaster scale sequence includes at least a disaster occurrence frequency sequence and a disaster occurrence type sequence; Dividing the database into at least one sub-database according to the accuracy level, clustering the disaster frequency sequences in the sub-database to obtain clusters corresponding to each accuracy level; Collecting real-time data detected by a meteorological data collector, wherein the real-time data includes the accuracy level of the meteorological forecast, the real-time meteorological forecast data sequence within a preset time scale corresponding to the current moment, and disaster scale data; Calculating the sub-error of the target accuracy level based on the clusters corresponding to the target accuracy level and the real-time data, and performing weighted summation on the sub-errors of all target accuracy levels to obtain the real-time error of the disaster scale sequence corresponding to the meteorological forecast data, wherein the target accuracy level is greater than or equal to the accuracy level of the meteorological forecast data; The error verification result is obtained by comparing the real-time error with the corresponding accuracy level of the predicted disaster scale sequence.

2. A multi-disaster risk early warning method and system using meteorological forecast data as claimed in claim 1, characterized in that: The disaster scale sequence also includes at least one of a temperature data sequence, a wind data sequence, a drought data sequence and a precipitation data sequence when a disaster occurs.

3. A multi-disaster risk early warning method and system using meteorological forecast data as claimed in claim 1, characterized in that: The meteorological data sequences in the sub-database are clustered to obtain clusters corresponding to each accuracy level, including: Perform dimensionality reduction on all meteorological data sequences in any sub-database to obtain the corresponding low-dimensional representation; Clustering the low-dimensional representations of all meteorological data sequences according to a density clustering algorithm to obtain clusters of the sub-database corresponding to the accuracy level; Wherein, the density clustering algorithm is the HDBSCAN algorithm or the DBSCAN algorithm.

4. A multi-disaster risk early warning method and system using meteorological forecast data as claimed in claim 1, characterized in that: The cluster includes a cluster center and multiple meteorological data sequences, and the sub-error of calculating the target accuracy level based on the cluster corresponding to the target accuracy level and the real-time data includes: For any cluster corresponding to the target accuracy level, calculate the similarity between its cluster center and the real-time meteorological parameter sequence; Based on the similarity, the clusters corresponding to the target accuracy level are screened to obtain a target cluster; For any target cluster, a cluster error of the target cluster is calculated based on the real-time meteorological data information and disaster data corresponding to a plurality of meteorological databases, wherein the plurality of meteorological databases include disaster data corresponding to each meteorological data sequence in the target cluster; The sub-errors of the target accuracy level are calculated based on the cluster errors and similarities of all target clustering clusters.

5. A multi-disaster risk early warning method and system using meteorological forecast data as claimed in claim 4, characterized in that: The step of screening the clusters corresponding to the target accuracy level based on the similarity to obtain the target clusters includes: In response to the similarity corresponding to any cluster being greater than or equal to a similarity threshold, the cluster is taken as a target cluster; In response to the similarity corresponding to any cluster being less than the similarity threshold, the cluster is deleted.

6. A multi-disaster risk early warning method and system using meteorological forecast data as claimed in claim 4, characterized in that: For any target cluster, the cluster error of the target cluster is calculated based on the real-time meteorological data information and multiple meteorological databases, including: Using an outlier detection algorithm to detect the multiple meteorological databases to obtain an outlier value of each meteorological data information; Screening the plurality of meteorological databases based on the outliers to obtain target disaster data information; The cluster error of the target cluster is calculated based on the real-time meteorological data information, the target meteorological data information and the outlier value of each target meteorological data information, and the cluster error satisfies the relationship: Among them, Num(a) is the number of target meteorological data information in the target cluster a, W m and L m are the mth target meteorological data and the outlier of the mth target meteorological data in the target cluster a, respectively. * is the real-time meteorological data, ∑(1-L) is the sum of the outliers of all target meteorological data in the target cluster a, and ε(a) represents the cluster error of the target cluster a.

7. A multi-disaster risk early warning method and system using meteorological forecast data as claimed in claim 6, characterized in that: The screening of the plurality of meteorological databases based on the outliers to obtain target meteorological data increments comprises: In response to an outlier value corresponding to any meteorological data being less than an outlier threshold, taking the meteorological data as target meteorological data; In response to an outlier value corresponding to any meteorological data being greater than or equal to the outlier threshold, the meteorological data is deleted.

8. A multi-disaster risk early warning method and system using meteorological forecast data as claimed in claim 4, characterized in that: The sub-errors of the target accuracy level satisfy the relationship: Among them, Sim(a) and ε(a) are the similarity and cluster error of target cluster a respectively, N is the number of target clusters corresponding to the target accuracy level, ∑Sim is the sum of the similarities of all target clusters corresponding to the target accuracy level, Δ k is the sub-error of the target accuracy level k.

9. A multi-disaster risk early warning method and system using meteorological forecast data as claimed in claim 1, characterized in that: The real-time error of the disaster data corresponding to the real-time meteorological data obtained by weighted summing of the sub-errors of all target accuracy levels includes: The reference degree for each target accuracy level is calculated, and the reference degree satisfies the relationship: Among them, ∑K is the sum of all target accuracy levels, w k is the reference degree of the target accuracy level k; The real-time error of the meteorological data is obtained by weighted summing up the sub-errors of all target accuracy levels based on the reference degree.

10. A multi-disaster risk early warning method and system using meteorological forecast data as claimed in claim 1, characterized in that: The error checking result includes normal and abnormal, and the method further includes: In response to the error checking result being normal, the real-time data of the meteorological conditions to be measured is stored in the database.

Images