CN118295045B - Meteorological monitoring alarm system and method based on big data - Google Patents

Abstract

The present invention relates to the field of meteorological monitoring technology, and discloses a meteorological monitoring and alarm system and method based on big data, comprising a meteorological data acquisition module for acquiring regional meteorological data of the area where each test point is located by the meteorological bureau; an identification module for identifying each test point with high-risk meteorology according to the regional meteorological data; a multi-dimensional partitioning module for forming multi-dimensional partitions of each test point according to each test point with high-risk meteorology; a multi-dimensional prediction module for calling local meteorological prediction models of each dimension according to the multi-dimensional partitions of each test point, and predicting the meteorological conditions of the test point in each dimension; a data fusion module for fusing meteorological prediction data of each dimension to form actual meteorological prediction data of the test point; and an early warning module for judging whether the test point is a dangerous area according to the actual meteorological prediction data of the test point, and if so, sending an alarm message to a client.

Description

A meteorological monitoring and alarm system and method based on big data

Technical Field

The present invention relates to the technical field of meteorological monitoring, and in particular to a meteorological monitoring and alarm system and method based on big data.

Background Art

In the field of meteorology, meteorological monitoring agencies have long been committed to providing the public with accurate and reliable weather forecast services by collecting and analyzing meteorological data. However, although existing meteorological forecasting technology has made great progress, it still faces many challenges and limitations when conducting meteorological forecasts for a single test point.

From the perspective of data sources, meteorological data usually comes from meteorological observation stations distributed in different locations. These stations provide the Meteorological Bureau with a basis for analysis and prediction by collecting data on meteorological elements such as air pressure, temperature, humidity, and wind speed. However, due to the uneven distribution of observation stations and the accuracy and resolution limitations of the observation equipment itself, there is a lack of meteorological data collection for a single test point, resulting in low accuracy of meteorological forecasts for a single test point.

From the perspective of prediction models, most existing meteorological prediction models are calculated based on regional data. These models simulate the physical and chemical processes of the atmosphere to predict weather changes in the future. However, due to the complexity of the models and the limitations of computing resources, these models often fail to fully consider the actual situation of a single test point when making predictions for the test point, resulting in large deviations in the meteorological prediction results for a single test point.

Summary of the invention

One of the purposes of the present invention is to provide a weather monitoring and alarm system and method based on big data, which can solve the problem of low accuracy of weather forecasting for a single test point in the prior art and improve the accuracy of weather forecasting.

In order to achieve the above purpose, a meteorological monitoring and alarm system based on big data is provided, including:

The meteorological data acquisition module is used to obtain regional meteorological data of the area where each test point is located from the meteorological bureau;

The identification module is used to identify the meteorological content and meteorological type corresponding to the meteorological data of each region according to the meteorological data of each region, and identify each test point corresponding to the high-risk meteorological condition based on the meteorological content and meteorological type;

The multi-dimensional partitioning module is used to partition each test point in multiple dimensions according to each test point corresponding to the identified high-risk weather conditions based on a preset multi-dimensional partitioning strategy, thereby forming multi-dimensional partitions corresponding to each test point;

The multi-dimensional prediction module is used to retrieve the local meteorological prediction model corresponding to each dimension according to the multi-dimensional partition corresponding to each test point, predict the meteorological conditions of the test point corresponding to each dimension, and form meteorological prediction data corresponding to each dimension;

A data fusion module is used to fuse the meteorological forecast data corresponding to each dimension according to the meteorological forecast data corresponding to each dimension of the point to be measured, so as to form the actual meteorological forecast data of the point to be measured;

The early warning module is used to determine whether the point to be tested is a dangerous area based on the actual weather forecast data of the point to be tested. If so, an alarm message is sent to the client;

The multi-dimensional partitioning strategy is:

Retrieve the actual historical meteorological data corresponding to each test point, the geographical information corresponding to the region, and the actual historical conditions of each test point in the region from the database;

Based on geographic information and actual historical meteorological data of each test point, each dimension in the dimension set pre-stored in the database is selected, and multiple dimensions that meet the actual situation of the area are selected from the dimension set;

According to the selected multiple dimensions, based on the historical actual conditions of each test point in the region and the actual historical meteorological data, each test point in the region is divided into multiple dimensions to form a multi-dimensional division plan for the region corresponding to each dimension.

The technical principle and effect of this scheme: In this scheme, the regional meteorological data corresponding to the area where each test point is located is first obtained from the Meteorological Bureau. By processing the regional meteorological data, the meteorological content and meteorological type corresponding to the area are identified, so as to realize the test points corresponding to each high-risk meteorology in the area, so as to realize the preliminary determination of the test points that need to be tested for meteorological forecasting, monitoring and alarm.

Afterwards, through the identified high-risk weather points corresponding to the test points, multi-dimensional partitioning strategies are used to realize multi-dimensional partitioning of each test point, thereby realizing multi-dimensional division of each test point in the area. For example, the test point i is divided together with other test points j, l, m, and n in dimension AA, and is divided together with test points j, l, b, and y in dimension BB.

Then, according to the multi-dimensional partitions corresponding to each test point, the local meteorological forecast model corresponding to each dimension is called to predict the multi-dimensional meteorological conditions, forming meteorological forecast data on multiple dimensions of the test point. Then, the meteorological forecast data on these dimensions are fused to obtain the actual meteorological forecast data of the test point. Based on the actual meteorological forecast data, it is possible to identify and monitor whether the test point is a dangerous area, and when the test point is a dangerous area, an alarm message is sent to the client to achieve accurate monitoring and timely alarm of the test point.

Compared with the prior art, the meteorological data of the Meteorological Bureau can only predict the region, but cannot accurately predict the weather at a single test point in the region. This makes the prediction of the meteorological data of a single test point in the region extremely inaccurate and unreliable, resulting in the inability to accurately monitor the meteorological data of a single test point in the region, and to issue a timely alarm when a danger occurs. In this application, the regional meteorological data corresponding to the region in the Meteorological Bureau is first used to make a preliminary judgment, so as to judge the test points that may have dangerous or high-risk weather. Then, in order to make accurate meteorological predictions for these test points, multi-dimensional division is performed, and the local meteorological models corresponding to each dimension after division are retrieved and predicted, so as to realize the prediction of multi-dimensional meteorological prediction data of the same test point. After that, through fusion, accurate and rapid prediction of the actual meteorological prediction data of the test point can be realized, which can solve the problem of low accuracy of meteorological prediction for a single test point in the prior art and improve the accuracy of meteorological prediction.

Further, the logic of selecting multiple dimensions that meet the actual situation of the region from the dimension set is: according to each dimension in the dimension set, each test point in the region is divided into blocks of a single dimension by clustering to form multiple sub-divisions corresponding to each dimension; according to the actual historical meteorological data, the first meteorological difference value between each test point in each sub-division and the second meteorological difference value between each sub-division are calculated; based on the first meteorological difference value and the second meteorological difference value, it is judged whether the first meteorological difference value is less than or equal to the first difference threshold value, and whether the second meteorological difference value is greater than or equal to the second difference threshold value; if both are satisfied, it is judged that the dimension is a dimension that meets the actual situation of the region, so as to realize the judgment of all dimensions in the dimension set, thereby selecting multiple dimensions that meet the actual situation of the region;

The calculation logic of the first meteorological difference value is:

According to the actual historical meteorological data corresponding to each test point in the same sub-area, the current corresponding season is determined; based on the current corresponding season, the historical temperature values of each test point corresponding to the current season are obtained from the actual historical meteorological data; the historical average temperature value corresponding to each test point in the current season is calculated; based on the first meteorological difference value calculation formula, the first meteorological difference value between each test point in the same sub-area is calculated;

The first meteorological difference value calculation formula is:

In the formula, is the first meteorological difference value, is the average temperature value of all test points in the same sub-division. is the historical average temperature value corresponding to the mth test point in the same sub-partition in the current season, is the total number of points to be tested in the corresponding sub-partition;

The calculation logic of the second meteorological difference value is:

Calculate the historical average temperature corresponding to each sub-area in the current season, and calculate the second meteorological difference value between different sub-areas based on the second meteorological difference value calculation formula;

The second meteorological difference value calculation formula is:

In the formula, is the second meteorological difference value, is the average temperature between different sub-intervals in the current season, is the historical average temperature value of the interval corresponding to subinterval n, is the total number of subintervals corresponding to the corresponding dimension.

Beneficial effect: By calculating and comparing the meteorological difference values between the test points within the sub-intervals and the meteorological difference values between the sub-intervals, the accuracy and reliability of selecting multiple dimensions that meet the actual conditions of the area are improved, providing reliable data support for the follow-up.

Furthermore, the multi-dimensional prediction module includes:

A partition overlap calculation module is used to calculate the partition overlap between sub-partitions corresponding to each dimension in the area according to the multi-dimensional partitions corresponding to each test point; each multi-dimensional partition includes multiple sub-partitions;

A judgment module is used to retrieve a preset overlap threshold from a database according to the overlap between each sub-partition, compare the overlap between the sub-partition and the preset overlap threshold, judge whether the overlap is greater than or equal to the preset overlap threshold, and calculate the proportion of the number of sub-partitions in a single dimension and the sub-partitions in other dimensions whose overlap is greater than or equal to the preset overlap threshold. If the proportion is greater than the preset threshold, it is judged that the meteorological correlation of each test point in the sub-partition is high; otherwise, it is judged that the meteorological correlation of each test point in the sub-partition is low;

The first prediction module is used to identify the corresponding test point in the sub-partition when the judgment result is that the meteorological correlation of each test point in the sub-partition is high, and randomly select the meteorological local prediction model corresponding to one of the dimensions, and based on the meteorological local prediction model, select the historical meteorological data corresponding to one of the test points at the previous moment and the regional meteorological data corresponding to the test point to predict the meteorological forecast data corresponding to the test point;

The second prediction module is used to identify the corresponding points to be tested in the sub-partition when the judgment result is that the meteorological correlation of the points to be tested in the sub-partition is low, and to call up the local meteorological prediction model corresponding to each dimension, and predict the meteorological prediction data on each dimension corresponding to each point to be tested, so as to form the meteorological prediction data on each dimension corresponding to each point to be tested.

Beneficial effect: In this scheme, it is fully considered that when dividing the intervals in various dimensions in the region, there may be duplications in a certain sub-interval between multiple dimensions. In this way, some test points can be in the same sub-interval regardless of the change of the dimension. In order to better identify this type of test points, the partition overlap degree is calculated by sub-intervals between different dimensions to judge the partition overlap degree between sub-intervals of different dimensions, and the proportion of the number of sub-intervals of a single dimension that overlap with sub-intervals of other dimensions is calculated. For example, when calculating the overlap degree of sub-interval i corresponding to dimension A with sub-intervals of other dimensions, the calculation If the proportion of the number of sub-intervals in other dimensions that overlap with the sub-interval i greater than or equal to the preset overlap threshold is greater than the threshold, it means that the various test points in the sub-interval i are mostly divided together when divided in different dimensions, which means that the meteorological correlation between the various test points in the sub-interval i is high. At this time, you can directly randomly select a local meteorological prediction model on a dimension to perform meteorological prediction to complete the meteorological prediction of all other test points in the sub-interval i, while ensuring the accuracy of the prediction of all test points in the sub-interval, while improving the prediction efficiency, solving the limited processing capacity of the system in the prior art, and improving the prediction efficiency is conducive to improving the processing performance of the system. If the meteorological correlation of the test point is low, a multi-dimensional local meteorological prediction model is used to predict each test point, and a multi-dimensional meteorological prediction of the test point with general meteorological correlation is achieved, which can improve the accuracy and reliability of the meteorological prediction of the test point.

Furthermore, the data fusion module includes:

A first allocation module is used to allocate the meteorological forecast data of a certain test point corresponding to the sub-division generated by the first prediction module to other test points corresponding to the sub-division at the same time, wherein the meteorological forecast data is the actual meteorological forecast data of all the test points in the sub-division;

A weight value calculation module is used to identify the position of the sub-interval of the test point in different dimensions according to the meteorological forecast data in each dimension corresponding to each test point formed by the second prediction module, calculate the distance value of the test point from the center position of the sub-interval in different dimensions based on the position of the test point and the center position of the sub-interval of the test point in different dimensions, and set the weight value corresponding to the meteorological forecast data corresponding to the test point in each dimension based on the distance value;

The first fusion module is used to fuse the meteorological forecast data corresponding to the test point in each dimension according to the weight value corresponding to the meteorological forecast data of the test point in each dimension, so as to form the actual meteorological forecast data of the test point.

Beneficial effect: In this scheme, when facing the fusion of meteorological forecast data in various dimensions corresponding to each test point formed by the second prediction module, full consideration is given to the different degrees of influence of meteorological forecast data in different dimensions on the test point. In order to accurately determine the actual meteorological data of the test point, the position of each test point in the sub-interval of each dimension is mainly determined to calculate the distance between the test point and the center position of the sub-interval in each dimension, so as to realize the determination of the weight value of the meteorological forecast data corresponding to the dimension, and then complete the fusion of meteorological forecast data in multiple dimensions, thereby improving the reliability of data fusion.

The present invention also provides a meteorological monitoring and alarm method based on big data, using the above-mentioned meteorological monitoring and alarm system based on big data.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a logic block diagram of a meteorological monitoring and alarm system based on big data according to an embodiment of the present invention.

DETAILED DESCRIPTION

The following is further described in detail through specific implementation methods:

Embodiment 1

A meteorological monitoring and alarm system based on big data is basically shown in Figure 1, including:

The meteorological data acquisition module is used to obtain regional meteorological data of the area where each test point is located from the meteorological bureau;

The identification module is used to identify the meteorological content and meteorological type corresponding to the meteorological data of each region according to the meteorological data of each region, and identify each test point corresponding to the high-risk meteorological condition based on the meteorological content and meteorological type;

In this embodiment, the meteorological content includes temperature, humidity, wind speed, wind direction change, and rainfall, and the meteorological types include sunny, cloudy, overcast, rainy, snowy, strong winds, and sandstorms. The identification of specific meteorological content and meteorological types is as follows: first, the meteorological data of each region is standardized, and the meteorological data of different units or formats are converted into a unified format and standard for subsequent processing. In this embodiment, the data corresponding to the meteorological content corresponding to the high-risk meteorology is greater than the corresponding thresholds, and the meteorological type is the corresponding rainy, snowy, strong winds or sandstorms.

The multi-dimensional partitioning module is used to partition each test point in multiple dimensions according to each test point corresponding to the identified high-risk weather conditions based on a preset multi-dimensional partitioning strategy, thereby forming multi-dimensional partitions corresponding to each test point;

The multi-dimensional partitioning strategy is:

The actual historical meteorological data corresponding to each test point, the geographical information corresponding to the region, and the historical actual situation of each test point in the region are retrieved from the database; in this embodiment, the geographical information includes the geographical location such as longitude and latitude, topography and landforms. The historical actual situation is the division situation corresponding to each test point in history.

Based on the geographic information and the actual historical meteorological data of each test point, each dimension in the dimension set pre-stored in the database is selected, and multiple dimensions that meet the actual conditions of the area are selected from the dimension set; in this embodiment, multiple dimensions are set in the dimension set, and each dimension in the dimension set includes an altitude dimension, a distance dimension, a surface vegetation dimension, a current meteorological similarity dimension, a historical meteorological similarity dimension, etc.

In this embodiment, the logic for selecting the actual situation in the area is: according to each dimension in the dimension set, each test point in the area is divided into blocks of a single dimension by clustering to form multiple sub-divisions corresponding to each dimension; according to the actual historical meteorological data, the first meteorological difference value between each test point in each sub-division and the second meteorological difference value between each sub-division are calculated; based on the first meteorological difference value and the second meteorological difference value, it is judged whether the first meteorological difference value is less than or equal to the first difference threshold value, and whether the second meteorological difference value is greater than or equal to the second difference threshold value; if both are satisfied, the dimension is judged to be a dimension that meets the actual situation in the area, so as to realize the judgment of all dimensions in the dimension set, and thus select multiple dimensions that meet the actual situation in the area.

The calculation logic of the first meteorological difference value is:

According to the actual historical meteorological data corresponding to each test point in the same sub-area, the current corresponding season is determined; based on the current corresponding season, the historical temperature values of each test point corresponding to the current season are obtained from the actual historical meteorological data; the historical average temperature value corresponding to each test point in the current season is calculated; based on the first meteorological difference value calculation formula, the first meteorological difference value between each test point in the same sub-area is calculated;

The first meteorological difference value calculation formula is:

In the formula, is the first meteorological difference value, is the average temperature value of all test points in the same sub-division. is the historical average temperature value corresponding to the mth test point in the same sub-partition in the current season, is the total number of points to be tested in the corresponding sub-partition;

The calculation logic of the second meteorological difference value is:

Calculate the historical average temperature corresponding to each sub-area in the current season, and calculate the second meteorological difference value between different sub-areas based on the second meteorological difference value calculation formula;

The second meteorological difference value calculation formula is:

In the formula, is the second meteorological difference value, is the average temperature between different sub-intervals in the current season, is the historical average temperature value of the interval corresponding to subinterval n, is the total number of subintervals corresponding to the corresponding dimension.

In this embodiment, many dimensions are set in the dimension set, but for different areas, the dimensions that conform to the actual situation of the area are different, so it is necessary to select dimensions that can represent the actual situation of the corresponding area from the dimension set, so as to achieve more accurate and scientific grouping of each test point in the area.

According to the selected multiple dimensions, based on the historical actual conditions of each test point in the region and the actual historical meteorological data, each test point in the region is divided into multiple dimensions to form a multi-dimensional division scheme for the region corresponding to each dimension. The multi-dimensional division scheme is a division scheme corresponding to each test point in the region in different dimensions. In this embodiment, the formation of the multi-dimensional division scheme is mainly based on the corresponding dimensions, clustering is performed, so as to complete the production of the corresponding dimensional division scheme. For example, the altitude dimension is based on the altitude data of each test point, and the altitude data within the first altitude threshold is divided into one category, the altitude data between the first altitude threshold and the second altitude threshold is divided into one category, and the altitude data greater than the second altitude threshold is divided into one category, so as to form a division based on the altitude dimension, and divide all the test points in the region into three sub-divisions. Then different dimensions are divided differently, so as to form a division scheme corresponding to each test point in the region in different dimensions. That is, when considering the altitude dimension, the test points with close altitude data are divided into one category, and when considering other dimensions, the test points with close data corresponding to the corresponding dimensions are divided into types.

In this embodiment, different dimensions are selected according to different regions. For example, when facing a plain, considering that the terrain corresponding to the plain is relatively flat, the corresponding meteorological changes within the preset range on the same plain will not be too large, so the test points on the plain may be ignored in the distance dimension. For example, those that do not exceed the preset range on the same plain can be classified into the same category, and those close to mountains or rivers on the plain may be classified into the same category.

The multi-dimensional prediction module is used to retrieve the local meteorological prediction model corresponding to each dimension according to the multi-dimensional partition corresponding to each test point, predict the meteorological conditions of the test point corresponding to each dimension, and form meteorological prediction data corresponding to each dimension;

The multi-dimensional prediction module comprises:

The partition overlap calculation module is used to calculate the partition overlap between the sub-partitions corresponding to each dimension in the area according to the multi-dimensional partitions corresponding to each point to be tested; each multi-dimensional partition includes multiple sub-partitions; for example, there are N corresponding points to be tested in area A, and they are numbered 1, 2, 3, ..., N respectively; the dimensions corresponding to this area are AA, BB, and CC respectively; when dividing each dimension, the sub-intervals AA1, AA2, and AA3 corresponding to the AA dimension are obtained; the sub-intervals BB1, BB2, and BB3 corresponding to the BB dimension; and the sub-intervals CC1, CC2, and CC3 corresponding to the CC dimension are obtained respectively; at this time, the partition overlap of each sub-interval is calculated in turn compared with the sub-intervals of other dimensions. In this embodiment, the partition overlap calculation logic is as follows: statistics are performed on the test points corresponding to the sub-partitions corresponding to each dimension to form a set of test points corresponding to the corresponding sub-partitions. When calculating the partition overlap corresponding to the sub-partitions of a certain dimension, based on the set of test points corresponding to each sub-partition corresponding to the dimension, the overlapping test points are calculated for the sub-partitions corresponding to each other dimension in turn, that is, the ratio of the number of overlaps between the test points corresponding to each sub-partition of the dimension and the test points corresponding to each sub-partition corresponding to other dimensions and the total number of different test points between the sub-partitions is the corresponding partition overlap.

The judgment module is used to retrieve a preset overlap threshold from a database according to the partition overlap between each sub-partition, compare the partition overlap with the preset overlap threshold, judge whether the partition overlap is greater than or equal to the preset overlap threshold, and calculate the proportion of sub-partitions in a single dimension and sub-partitions in other dimensions whose partition overlap is greater than or equal to the preset overlap threshold. If the proportion is greater than the preset threshold, it is judged that the meteorological correlation of each test point in the sub-partition is high, otherwise it is judged that the meteorological correlation of each test point in the sub-partition is low. In this embodiment, the partition overlap between the sub-interval of a single dimension and the sub-interval of other dimensions not only needs to be greater than the preset overlap threshold, but also the corresponding number of dimensions that meet the requirement must meet the requirement, which greatly improves the accuracy of determining the meteorological correlation of each test point in the sub-interval.

The first prediction module is used to identify the corresponding test point in the sub-partition when the judgment result is that the meteorological correlation of each test point in the sub-partition is high, and randomly select the meteorological local prediction model corresponding to one of the dimensions, and based on the meteorological local prediction model, select the historical meteorological data corresponding to one of the test points at the previous moment and the regional meteorological data corresponding to the test point to predict the meteorological forecast data corresponding to the test point;

The second prediction module is used to identify the corresponding points to be tested in the sub-partition when the judgment result is that the meteorological correlation of the points to be tested in the sub-partition is low, and to call up the local meteorological prediction model corresponding to each dimension, and predict the meteorological prediction data on each dimension corresponding to each point to be tested, so as to form the meteorological prediction data on each dimension corresponding to each point to be tested.

A data fusion module is used to fuse the meteorological forecast data corresponding to each dimension according to the meteorological forecast data corresponding to each dimension of the point to be measured, so as to form the actual meteorological forecast data of the point to be measured;

The data fusion module comprises:

The first allocation module is used to simultaneously allocate the meteorological forecast data of a certain test point corresponding to the sub-partition generated by the first prediction module to other test points corresponding to the sub-partition, and the meteorological forecast data is the actual meteorological forecast data of all the test points in the sub-partition; in this embodiment, for each test point in the sub-partition, considering their strong meteorological correlation, each prediction of the multi-dimensional local meteorological forecast model is not performed, but the local meteorological forecast model corresponding to one of the dimensions is randomly selected to predict the meteorological forecast data of all the test points in the sub-partition, which can not only ensure the accuracy of the prediction, but also improve the system prediction processing efficiency.

A weight value calculation module is used to identify the position of the sub-interval of the test point in different dimensions according to the meteorological forecast data in each dimension corresponding to each test point formed by the second prediction module, calculate the distance value of the test point from the center position of the sub-interval in different dimensions based on the position of the test point and the center position of the sub-interval of the test point in different dimensions, and set the weight value corresponding to the meteorological forecast data corresponding to the test point in each dimension based on the distance value;

The first fusion module is used to fuse the meteorological forecast data corresponding to the test point in each dimension according to the weight value corresponding to the meteorological forecast data of the test point in each dimension, so as to form the actual meteorological forecast data of the test point.

The early warning module is used to determine whether the point to be tested is a dangerous area based on the actual weather forecast data of the point to be tested. If so, an alarm message is sent to the client.

This embodiment also provides a meteorological monitoring and alarm method based on big data, using the above-mentioned meteorological monitoring and alarm system based on big data.

The above is only an embodiment of the present invention. The common sense such as the known specific structure and characteristics in the scheme is described too much here. The ordinary technicians in the relevant field know all the common technical knowledge in the technical field of the invention before the application date or priority date, can know all the existing technologies in the field, and have the ability to apply the conventional experimental means before that date. The ordinary technicians in the relevant field can improve and implement this scheme in combination with their own abilities under the enlightenment given by this application. Some typical known structures or known methods should not become obstacles for ordinary technicians in the relevant field to implement this application. It should be pointed out that for those skilled in the art, without departing from the structure of the present invention, several deformations and improvements can be made, which should also be regarded as the scope of protection of the present invention, which will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims (4)

1. A weather monitoring alarm system based on big data is characterized in that: comprising the following steps:

The meteorological data acquisition module is used for acquiring regional meteorological data of the region where each point to be detected is located from a meteorological bureau;

The identification module is used for identifying the weather content and the weather type corresponding to the weather data of each area according to the weather data of each area, and identifying each to-be-detected point corresponding to the high-risk weather based on the weather content and the weather type;

the multidimensional partitioning module is used for carrying out multidimensional partitioning on each point to be tested based on a preset multidimensional partitioning strategy according to each point to be tested corresponding to the identified high-risk weather, so as to form multidimensional partitions corresponding to each point to be tested;

the multidimensional prediction module is used for calling a local weather prediction model corresponding to each dimension according to the multidimensional partition corresponding to each to-be-measured point, and predicting the weather condition of the to-be-measured point corresponding to each dimension to form weather prediction data corresponding to each dimension;

the data fusion module is used for fusing the weather forecast data corresponding to each dimension according to the weather forecast data corresponding to each dimension of the point to be measured to form weather forecast actual data of the point to be measured;

The early warning module is used for judging whether the point to be detected is a dangerous area according to weather forecast actual data of the point to be detected, and if so, sending alarm information to the client;

the multi-dimensional partitioning strategy is:

The method comprises the steps of calling historical weather actual data corresponding to each to-be-measured point, geographic information corresponding to an area and historical actual conditions of each to-be-measured point in the area from a database;

Based on geographic information and historical weather actual data of each to-be-measured point, selecting each dimension in a dimension set stored in advance in a database, and selecting a plurality of dimensions conforming to the actual conditions of the area from the dimension set;

According to the selected multiple dimensions, based on the historical actual conditions and the historical weather actual data of each to-be-measured point in the area, carrying out multi-dimensional division on each to-be-measured point in the area to form a multi-dimensional division scheme of the area corresponding to each dimension;

The logic for selecting the plurality of dimensions which meet the actual conditions of the area from the dimension set is as follows: according to each dimension in the dimension set, partitioning the to-be-measured points in the area in a single dimension mode to form a plurality of sub-partitions corresponding to each dimension, calculating a first weather difference value between each to-be-measured point in each sub-partition and a second weather difference value between each sub-partition according to historical weather actual data, judging whether the first weather difference value is smaller than or equal to a first difference threshold value and judging whether the second weather difference value is larger than or equal to a second difference threshold value based on the first weather difference value and the second weather difference value, if so, judging that the dimension is the dimension conforming to the actual condition of the area, and judging all the dimensions in the dimension set, so that a plurality of dimensions conforming to the actual condition of the area are selected;

The calculation logic of the first weather difference value is as follows:

Determining the current corresponding season according to the historical weather actual data corresponding to each to-be-measured point of the same subarea, acquiring the historical temperature value of each to-be-measured point corresponding to the current season from the historical weather actual data based on the current corresponding season, calculating the historical average temperature value of each to-be-measured point corresponding to the current season, and calculating the first gas aberration difference value between each to-be-measured point in the same subarea based on a first gas aberration difference value calculation formula;

The first weather difference value calculation formula is as follows:

In the method, in the process of the invention, Is the difference value of the first gas phase difference,The average temperature value of all the to-be-measured points in the same sub-partition,The historical average temperature value corresponding to the mth point to be measured in the same sub-zone in the current season is used,The total number of the to-be-measured points in the corresponding sub-partition;

The calculation logic of the second weather difference value is as follows:

calculating the historical average temperature corresponding to each sub-subarea in the current season, and calculating a second weather difference value between different subareas based on a second weather difference value calculation formula;

The second weather difference value calculation formula is:

In the method, in the process of the invention, As the value of the second weather difference,For the average temperature of the different subintervals between the current seasons,The historical average temperature value of the interval corresponding to the subinterval n,The total number of subintervals corresponding to the corresponding dimensions.

2. The big data based weather monitoring and alarming system according to claim 1, wherein: the multi-dimensional prediction module includes:

the partition coincidence degree calculating module is used for calculating the partition coincidence degree between the sub-partitions corresponding to each dimension in the region according to the multi-dimension partitions corresponding to each to-be-measured point; each multi-dimensional partition includes a plurality of sub-partitions;

The judging module is used for calling a preset coincidence degree threshold value from the database according to the coincidence degree of the subareas, comparing the coincidence degree of the subareas with the preset coincidence degree threshold value, judging whether the coincidence degree of the subareas is larger than or equal to the preset coincidence degree threshold value, calculating the number ratio of the coincidence degree of the subareas in a single dimension to the coincidence degree of the subareas in other dimensions, which is larger than or equal to the preset coincidence degree threshold value, judging that the weather association of each measuring point in the subarea is high if the number ratio is larger than the preset threshold value, and otherwise judging that the weather association of each measuring point in the subarea is low;

The first prediction module is used for identifying the point to be detected corresponding to the sub-subarea when the judging result is that the weather association of each point to be detected in the sub-subarea is high, randomly selecting a weather local prediction model corresponding to one dimension, selecting the historical weather data at the last moment corresponding to one point to be detected and the regional weather data corresponding to the point to be detected based on the weather local prediction model, and predicting weather prediction data corresponding to the point to be detected;

And the second prediction module is used for identifying each to-be-measured point corresponding to the sub-partition when the judging result is that the weather association of each to-be-measured point in the sub-partition is low, calling a local weather prediction model corresponding to each dimension, and predicting weather prediction data in each dimension corresponding to each to-be-measured point to form weather prediction data in each dimension corresponding to each to-be-measured point.

3. The big data based weather monitoring and alarming system according to claim 2, wherein: the data fusion module comprises:

the first allocation module is used for simultaneously allocating the weather forecast data of one to-be-measured point corresponding to the sub-subarea generated by the first forecast module to other to-be-measured points corresponding to the sub-subarea, wherein the weather forecast data are weather forecast actual data of all to-be-measured points in the sub-subarea;

The weight value calculation module is used for identifying the positions of the subintervals of the points to be measured in different dimensions according to the weather forecast data of the points to be measured in each dimension formed by the second forecast module, calculating the distance value of the points to be measured from the central position of the subintervals of the points to be measured in different dimensions based on the positions of the points to be measured and the central positions of the subintervals of the points to be measured in different dimensions, and setting the weight value of the points to be measured corresponding to the weather forecast data of the points to be measured in each dimension based on the distance value;

the first fusion module is used for fusing the weather forecast data corresponding to the point to be tested in each dimension according to the weight value corresponding to the weather forecast data corresponding to the point to be tested in each dimension to form weather forecast actual data of the point to be tested.

4. A weather monitoring alarm method based on big data is characterized in that: use of a weather monitoring alarm system based on big data according to any of the preceding claims 1-3.