A smart weather information system

Abstract

This invention provides an intelligent weather information dissemination system, relating to the field of meteorological service technology. It includes: a data acquisition module for real-time acquisition of meteorological elements, environmental conditions, and behavioral trajectories at the user's location, constructing a spatiotemporally correlated multidimensional dynamic dataset; a unit partitioning module for dividing continuous meteorological response units based on the multidimensional dynamic dataset through spatiotemporal feature extraction, establishing a mapping relationship table between meteorological elements and geographical units; and an analysis module for determining the spatial distribution range of meteorological elements using a convex hull algorithm based on the mapping relationship table, calculating the local deviation coefficient of meteorological elements within each unit, and generating regional meteorological characteristic indicators. This invention can provide more targeted, timely, and effective meteorological services for different geographical units, improving the precision and response efficiency of meteorological services.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of meteorological service, in particular to an intelligent meteorological answering system. BACKGROUND

[0002] Most of the existing public meteorological service systems release meteorological information based on fixed geographical areas (such as administrative divisions). However, the meteorological environment has significant spatio-temporal dynamics, and the static division method of fixed geographical units cannot fully meet the fine service demand, which is specifically manifested in the following aspects:

[0003] The meteorological characteristics of different positions (such as city centers and suburbs) in the same administrative region may be different, but the existing system cannot dynamically quantify this local deviation; the early warning instructions are weakly bound to the specific geographical environment, so that different risk levels of scenarios in the same region may receive the same level of response measures. SUMMARY

[0004] The technical problem to be solved by the application is to provide an intelligent meteorological answering system to improve the efficiency and accuracy of responding to meteorological disasters.

[0005] To solve the above technical problems, the technical scheme of the application is as follows:

[0006] In a first aspect, an intelligent meteorological answering system comprises:

[0007] A data acquisition module is configured to acquire meteorological elements, environmental states and behavior trajectories of user positions in real time, and to construct a multi-dimensional dynamic data set associated with space and time;

[0008] A unit division module is configured to divide continuous meteorological response units by extracting time-space characteristics based on the multi-dimensional dynamic data set, and to establish a mapping relationship table of meteorological elements and geographical units;

[0009] An analysis module is configured to determine the spatial distribution range of meteorological elements by using a convex hull algorithm according to the mapping relationship table, to calculate the local deviation coefficient of meteorological elements in each unit, and to generate regional meteorological characteristic indexes;

[0010] An instruction generation module is configured to generate hierarchical early warning instructions and response measures bound to geographical units according to the regional meteorological characteristic indexes and a pre-set control strategy library;

[0011] A feedback module is configured to synchronize the hierarchical instructions to roadside devices and user terminals through a distributed architecture, to acquire instruction execution state data in real time and verify the response effect, and to form a closed-loop feedback.

[0012] In a second aspect, a computing device comprises:

[0013] One or more processors;

[0014] a storage device storing one or more programs, when executed by the one or more processors, cause the one or more processors to implement the system.

[0015] In a third aspect, a computer-readable storage medium stores a program, which, when executed by a processor, implements the system.

[0016] The above scheme of the present application at least has the following beneficial effects:

[0017] Real-time integration of user location associated weather, environmental and behavior data ensures the spatio-temporal correlation and timeliness of the data; based on the spatio-temporal feature extraction, continuous weather response units are divided and a mapping relationship is established, breaking the limitation of fixed area division, making the association of weather elements and geographical units more in line with the actual weather distribution law, and improving the scientificity and pertinence of unit division. The convex hull algorithm is used to determine the spatial distribution range and calculate the local deviation coefficient, the regional weather characteristic index is generated according to the historical data, and the spatial distribution and abnormal degree of weather elements are accurately quantified; according to the characteristic index and the preset strategy library, hierarchical instructions and measures bound to geographical units are generated, ensuring that the early warning and response measures can adapt to the differences in weather risks of different regions, improving the accuracy and effectiveness of the response. Through the distributed architecture, the instructions are synchronized and a closed loop feedback is formed, which can not only ensure that the instructions quickly reach the terminal, but also verify the execution state and optimize the strategy, continuously improve the response efficiency and service quality of the system, and realize the dynamic iteration and precise upgrading of the weather service. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 is a schematic diagram of an intelligent weather answering system provided by an embodiment of the present application.

[0019] Figure 2 is a flowchart of generating regional weather characteristic indexes by determining the spatial distribution range of weather elements using the convex hull algorithm and calculating the local deviation coefficient of weather elements in each unit according to the mapping relationship table. DETAILED DESCRIPTION

[0020] Exemplary embodiments of the present disclosure will be described below in greater detail with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be accurately conveyed to those skilled in the art.

[0021] As Figure 1 shown, an embodiment of the present application proposes an intelligent weather answering system, which comprises:

[0022] A data collection module is configured to collect meteorological elements, environmental states and behavior trajectories of a user in real time, and construct a multi-dimensional dynamic data set associated with time and space;

[0023] A unit division module is configured to divide continuous meteorological response units by time and space feature extraction based on the multi-dimensional dynamic data set, and establish a mapping relationship table of meteorological elements and geographical units;

[0024] An analysis module is configured to determine a spatial distribution range of meteorological elements by a convex hull algorithm according to the mapping relationship table, calculate a local deviation coefficient of meteorological elements in each unit, and generate a regional meteorological characteristic index;

[0025] An instruction generation module is configured to generate a hierarchical early warning instruction and a response measure bound to a geographical unit according to the regional meteorological characteristic index and a preset control strategy library;

[0026] A feedback module is configured to synchronize the hierarchical instruction to roadside equipment and a user terminal through a distributed architecture, collect instruction execution state data in real time and verify a response effect, and form a closed-loop feedback.

[0027] In the embodiment of the present application, meteorological, environmental and behavior data associated with the user's location are integrated in real time to ensure the time and space correlation and timeliness of the data. Continuous meteorological response units are divided based on time and space feature extraction and a mapping relationship is established, which breaks through the limitation of fixed regional division, makes the association of meteorological elements and geographical units more in line with the actual meteorological distribution law, and improves the scientificity and pertinence of unit division. The spatial distribution range is determined by a convex hull algorithm and the local deviation coefficient is calculated, the regional meteorological characteristic index is generated according to the historical data, the spatial distribution and abnormal degree of meteorological elements are accurately quantified, the hierarchical instruction and measure bound to the geographical unit are generated according to the characteristic index and the preset strategy library, which ensures that the early warning and response measures can adapt to the meteorological risk differences of different regions, and improves the accuracy and effectiveness of the response. The instruction is synchronized through the distributed architecture and a closed-loop feedback is formed, which can ensure that the instruction quickly reaches the terminal, and through the execution state verification and strategy optimization, the response efficiency and service quality of the system are continuously improved, and the dynamic iteration and precision upgrade of the meteorological service are realized.

[0028] In a preferred embodiment of the present application, meteorological elements, environmental states and behavior trajectories of a user in real time are collected to construct a multi-dimensional dynamic data set associated with time and space. Continuous meteorological response units are divided by time and space feature extraction, and a mapping relationship table of meteorological elements and geographical units can include:

[0029] The number of trajectory points per unit area is calculated according to the time and space distribution density of the behavior trajectory, and the central coordinates of a region with a density exceeding a density determination threshold are obtained as hotspot coordinates;

[0030] Based on hotspot coordinates, real-time data from topologically adjacent meteorological monitoring stations are obtained, and the absolute values ​​of the rate of change of temperature and visibility between topologically adjacent stations are calculated to generate a gradient field of meteorological elements with hotspot coordinates as the core.

[0031] Based on the contour line distribution of the meteorological element gradient field and environmental continuity parameters, a closed boundary for a meteorological response unit is generated in a continuous region where the rate of temperature change is less than or equal to the threshold for determining the rate of temperature change and the rate of visibility change is less than or equal to the threshold for determining the rate of visibility change. Specifically, this includes:

[0032] Based on topologically adjacent station data in the gradient field of meteorological elements, spatial interpolation is performed to generate continuous contour distribution data of temperature change rate and visibility change rate;

[0033] Extract the coordinate set of the envelope region where the rate of temperature change is less than or equal to the threshold for determining the rate of temperature change from continuous contour line distribution data;

[0034] Determine whether the wind speed direction standard deviation within the envelope area is less than or equal to the direction consistency threshold, and whether the terrain elevation range is less than or equal to the elevation difference threshold; merge the coordinate sets of adjacent envelope areas that have a temperature change rate less than or equal to the temperature change rate judgment threshold, a visibility change rate less than or equal to the visibility change rate judgment threshold, and meet the environmental continuity standard, to obtain the closed boundary data of the meteorological response unit.

[0035] Real-time meteorological element data are assigned to geographical coordinates and associated with the closed boundary of the meteorological response unit to obtain a unit-element mapping relationship table.

[0036] In this embodiment of the invention, meteorological monitoring sensors are deployed at 500-meter intervals along urban roads. Simultaneously, with user authorization, real-time location information and movement data are obtained via a mobile application. The sensors collect temperature, humidity, air pressure, and visibility data every 10 minutes, with accuracies controlled to ±0.5℃, ±3%, ±1hPa, and ±10 meters, respectively. Each meteorological element data point is accompanied by the acquisition time accurate to the second and the latitude and longitude coordinates of the sensor, for example, E116.3800°, N39.8900°. Environmental status data is obtained through an urban environmental monitoring network, including PM2.5 concentration and road surface humidity. PM2.5 concentration is updated hourly with an accuracy of ±5μg / m³, and road surface humidity is updated every 30 minutes, categorized into three levels: dry, wet, and waterlogged. This environmental status data is associated with the corresponding monitoring area boundary coordinates. User behavior trajectories are recorded every 30 seconds using the mobile phone's global positioning system; locations where the user stays for more than 5 minutes are marked as key stopping points. The trajectory data is grouped by user identifier, and consecutive coordinate points are strung together in chronological order to form a trajectory line. The movement speed of adjacent points is calculated by dividing the movement distance by 30 seconds. If the movement distance within 30 seconds is less than 50 meters, it is determined to be a stationary state. Using 1 square kilometer as the basic spatial unit and 5 minutes as the time interval, the meteorological elements and environmental status data within the same spatial unit and time interval are matched with the behavioral trajectory segments of all users in that area. For example, the temperature data and PM2.5 data of a certain area from 8:00 to 8:05 will be bound with the trajectory data of users who passed through that area during that time period, forming a dynamic dataset containing five dimensions of information: time, space, meteorology, environment, and behavior. Each data unit contains a unique spatiotemporal code to ensure that the data source can be accurately traced in subsequent calls.

[0037] When calculating hotspot coordinates, the city's administrative boundary is used as the overall area, divided into 100m x 100m grid units. The number of user behavior trajectory points within each grid over the past 24 hours is counted. Each user's coordinates every 30 seconds are counted as one trajectory point. Each grid has an area of ​​10,000 square meters. The number of trajectory points per unit area is obtained by dividing the total number of trajectory points in the grid by 10,000, expressed as points / square meter. For example, if a grid has 2,000 trajectory points, the density is 2,000 divided by 10,000, resulting in 0.2 points / square meter. Based on historical data, when the number of trajectory points per unit area exceeds 0.15 points / square meter, it is considered a densely populated area of ​​user activity. All grids meeting this condition are selected. Adjacent densely populated grids are merged into a continuous region, and the geometric center of this region is calculated. The formula is to take the average of the maximum and minimum values ​​of the latitude and longitude of the region boundary. For example, if the longitude of a region is 116.3° to 116.5° east and the latitude is 39.9° to 40.1° north, then the central longitude is the sum of 116.3° and 116.5° divided by 2, resulting in 116.4°. The central latitude is the sum of 39.9° and 40.1° divided by 2, resulting in 40.0°. That is, the central coordinates are 116.4° east longitude and 40.0° north latitude. These central coordinates are marked as hotspot coordinates. Each hotspot coordinate is accompanied by the total density value of trajectory points in that region. The total density value of trajectory points is calculated by adding the total number of trajectory points in all grids within the continuous region to get the total number of trajectory points in the region. Then, this total number is divided by the total area of ​​the continuous region (in square meters). The result is the total density value of trajectory points in that region (in points / square meter).

[0038] Using the hotspot coordinates as the center, meteorological monitoring stations are selected within a preset radius of 3 kilometers, with a maximum selection of 10 stations. The straight-line distance between stations is calculated using a geographic information system (GIS). Stations with a distance less than 2 kilometers (or a preset threshold) are considered topologically adjacent, and a station connection network is constructed. For each pair of topologically adjacent stations, temperature data at the same time is extracted, the temperature difference between the two stations is calculated, and this difference is divided by the straight-line distance between the two stations to obtain the temperature change rate, which is then taken as the absolute value. The visibility change rate is calculated similarly: visibility data at the same time is extracted, the difference is calculated, and then divided by the station distance, with the absolute value taken. The hotspot coordinates are then used as the reference point for the hotspot coordinates. A polar coordinate system is established with E116.4000° and N39.9000° as the origin, with the polar axis pointing due north. For each pair of topologically adjacent stations, the coordinates of the midpoint of the line connecting the two stations are calculated. The absolute values ​​of the temperature change rate and visibility change rate of the pair of stations are marked at the corresponding midpoint positions. For example, the midpoint coordinates of station A (E116.3900°, N39.9000°) and station B (E116.4100°, N39.9000°) are (E116.4000°, N39.9000°), and its absolute value of temperature change rate is marked as 0.8℃ / km; station C (E116... The midpoint coordinates of station D (E116.4000°, N39.8900°) and station D (E116.4000°, N39.9100°) are (E116.4000°, N39.9000°). The absolute value of the visibility change rate is marked as 120 m / km. Linear interpolation is used to fill the blank areas between stations. During interpolation, the weights of adjacent points are allocated according to distance, with closer points accounting for 60%-80% and farther points accounting for 20%-40%. Taking a 1km radius around a hotspot as an example, the temperature change rate gradually increases from 0.5℃ / km at the center towards the edge, reaching 2℃ / km at 1km. Furthermore, the rate of change increment is stable at 0.15℃ / km for every 0.1 km interval. That is, at 0.1 km, it is 0.5 plus 0.15, resulting in 0.65℃ / km; at 0.2 km, it is 0.5 plus 0.15 multiplied by 2, resulting in 0.8℃ / km, and so on. Taking the hotspot coordinates as the center, a specific range of interpolation area is defined, such as a 10 km × 10 km area from E116.3500°E to E116.4500°E and N39.8500°N to N39.9500°N. This interpolation area is divided into grids of preset sizes such as 100 m × 100 m, forming 10,000 grid points.

[0039] For the temperature change rate and visibility change rate of each grid point, an inverse distance weighted interpolation method is used for accurate calculation. Specifically, for each grid point in the interpolation area, the geographical coordinate distance is first calculated to select the five closest meteorological monitoring stations (this number is a preset value and can be adjusted according to actual needs). Based on the temperature change rate and visibility change rate data of these five stations, a calculation weight is assigned to each station according to the inverse distance weighting rule. The weight is inversely proportional to the square of the straight-line distance from the grid point to the corresponding station. That is, the closer the station is, the greater its influence on the calculation result of the grid point. To ensure the rationality and stability of the weight allocation, a weight range is set for each station. The weight value of the first closest station is between 0.3 and 0.5; the weight value of the second closest station is between 0.2 and 0.3; and the total weight of the remaining three stations (the third to fifth closest) is between 0.2 and 0.5. At the same time, the weight values ​​of the five stations are strictly summed to 1 to meet the normalization requirements of the interpolation calculation. When calculating the temperature change rate of the grid point, the temperature change rate of each station is multiplied by its corresponding weight, and the results of these five products are added together. The sum is the temperature change rate of the grid point. Similarly, when calculating the visibility change rate of grid points, the visibility change rate of each station is multiplied by its corresponding weight, and the results of the five multiplications are summed to obtain the visibility change rate of the grid point. When calculating the temperature change rate of grid points, the temperature change rate of each station is multiplied by its corresponding weight and summed. Similarly, when calculating the visibility change rate of grid points, the visibility change rate of each station is multiplied by its corresponding weight and summed. The obtained temperature change rates of each grid point are classified at preset intervals such as 0.2℃ / km, and the visibility change rates are classified at preset intervals such as 50m / km. A unique numerical code is assigned to each level to form a continuous numerical distribution dataset. The threshold for judging the temperature change rate is set to 0.5℃ / km. The temperature change rate data of all grid points are traversed, and grid points with values ​​≤0.5℃ / km are selected. The set of these grid points is the target area, and the boundary of the area is composed of grid points that satisfy the temperature change rate = 0.5℃ / km.

[0040] Coordinates are extracted from these boundary grid points. Along the boundary, in a clockwise direction, a point is selected every 10 meters, with the latitude and longitude accuracy of each point to 0.0001 degrees. For example, starting from the initial point E116.3800°, N39.8900°, the next point is E116.3801°, N39.8901°, and so on, for a total of 500 coordinate points, forming the boundary coordinate set of the envelope region. Each coordinate point has a unique serial number. Within the envelope region, 10 wind speed monitoring points are selected according to the principle of uniform distribution, with coordinates E116.3850°, N39.8950°, E116.3900°, N39.8950°, etc. Wind direction data is collected continuously at each monitoring point for 1 hour, with a collection frequency of once every 10 minutes. Therefore, each point collects wind direction data 6 times per hour. 10 monitoring points × 6 times / point = 60 We collect wind direction data and convert it into angle values ​​ranging from 0 to 360 degrees (e.g., 30°, 32°, 28°, 31°, 29°, 33°, etc.). The conversion rule is to use true north as the reference (0°) and rotate clockwise, increasing the angle each time. That is, true east is 90°, true south is 180°, true west is 270°, and a full rotation returns to true north at 360° (coinciding with 0°). For example, true north is converted to 0°, northeast-northeast to 30°, true east to 90°, southeast-southeast to 150°, true south to 180°, southwest-west to 240°, true west to 270°, and northwest-northwest to 330°, etc. Values ​​like 30°, 32°, 28°, 31°, 29°, and 33° are all converted angle values. To calculate the standard deviation of these 60 angle values, the first step is to calculate the sum of the 60 angle values ​​and then divide by 60. The first step is to obtain the average value. The second step is to subtract the average value from each angle value to obtain 60 deviation values, and then square each deviation value. The third step is to add up the 60 squared deviation values ​​and then divide by 60 (the total number of data points) to obtain the variance. The fourth step is to take the square root of the variance, and the result is the standard deviation. If the standard deviation is ≤15°, the wind speed direction consistency of the envelope area is determined to meet the standard. For example, after calculating the 60 angle values ​​of the above 10 monitoring points, the standard deviation is 12°, which meets the requirements.

[0041] Elevation data of all points within the envelope area are obtained through on-site measurement and aerial mapping, with a data accuracy of 1 meter (i.e., the elevation value of each point is accurate to the meter). The maximum and minimum values ​​are selected from these elevation data, and the result of subtracting the minimum value from the maximum value is the topographic elevation range. The elevation difference threshold is set at 50 meters. If the topographic elevation range is ≤50 meters, it indicates that the terrain undulation in the area is small and meets the requirements of environmental continuity. For example, in a certain envelope area, the highest measured elevation is 120 meters and the lowest elevation is 80 meters, with a topographic elevation range of 40 meters. 40 meters < 50 meters, which meets the threshold standard.

[0042] Adjacent eligible envelope regions are topologically merged using polygons. Overlapping boundaries are identified through coordinate comparison. For example, the boundary points E116.3900°, N39.8900° to E116.3905°, N39.8905° of region A overlap with the corresponding boundary points of region B, and these overlapping points are deleted. For regions with gaps, such as the 2-meter gap between the boundaries of regions C and D at E116.3910°, N39.8910° and E116.3912°, N39.8912°, these two points are connected by a straight line to form a continuous boundary. The final generated meteorological response unit closed boundary consists of 800 vertex coordinates, and each coordinate point is topologically checked to ensure no overlap or gaps.

[0043] Each meteorological response unit is assigned a unique unit identifier, such as U001, U002, etc., and its closed boundary coordinate range is stored, i.e., minimum east longitude, maximum east longitude, minimum north latitude, and maximum north latitude. For example, the range of U001 is E116.3700°–E116.3900°, N39.8800°–N39.9000°. For each real-time meteorological element data, such as the temperature of 25°C collected by the sensor at 8:30 on E116.3800°, N39.8900°, its coordinates are extracted and compared with the range of each unit. If it falls within the range of U001, it is assigned to U001. In the table, each unit identifier corresponds to one row of data, including the unit identifier, boundary coordinate range, real-time temperature, average humidity, maximum visibility, and P. Fields such as the median M2.5 concentration are included. The real-time temperature is the average of all sensor data within the unit, rounded to one decimal place. For example, if there are 5 sensors in U001 with temperature data of 24.3℃, 24.6℃, 24.5℃, 24.7℃, and 24.4℃, the average is calculated by adding these five values ​​together: 24.3 + 24.6 + 24.5 + 24.7 + 24.4, which equals 122.5℃. This sum is then divided by 5 to get 24.5℃. The values ​​are then entered into the corresponding fields in the table. The table is updated every 10 minutes, deleting historical data older than one hour. For example, if the update occurs at 9:00, data before 7:50 is deleted, new data from 8:50 to 9:00 is added, and all statistical values ​​are recalculated to ensure the timeliness and accuracy of the data.

[0044] By accurately locating hotspots of user activity, meteorological services become more aligned with actual user needs. The generated meteorological element gradient fields and continuous isopleths clearly reflect the spatial variation trends of meteorological elements, providing a basis for dividing reasonable meteorological response units. Furthermore, the closed boundaries of meteorological response units generated based on environmental continuity parameters ensure the consistency and stability of meteorological conditions within the units, improving the accuracy of the association between meteorological elements and geographical units. The resulting unit-element mapping table helps to achieve refined and targeted meteorological services, improving the quality and efficiency of meteorological services.

[0045] like Figure 2 As shown, in another preferred embodiment of the present invention, the spatial distribution range of meteorological elements is determined using a convex hull algorithm based on a mapping table, the local deviation coefficient of meteorological elements within each unit is calculated, and regional meteorological characteristic indicators are generated, which may include:

[0046] Based on the mapping relationship table, the coordinate set of meteorological monitoring stations is extracted, and the spatial distribution range boundary of meteorological elements is generated using the convex hull algorithm, specifically including:

[0047] Extract the geographic coordinates of all meteorological monitoring stations from the mapping table to form a spatial coordinate set;

[0048] Identify the coordinate point with the smallest latitude in the spatial coordinate set as the reference point, calculate the azimuth angle between the reference point and the remaining coordinate points in the spatial coordinate set, and generate a sorted coordinate sequence in ascending order of azimuth angle;

[0049] Traverse the sorted coordinate sequence, and determine whether the current coordinate point satisfies the vertex condition of the convex polygon. Add the coordinate points that meet the convexity condition to the vertex set of the convex hull.

[0050] Construct a closed polygon boundary by connecting the coordinate points in the convex hull vertex set in the order of connection, thus generating the spatial distribution range boundary of meteorological elements.

[0051] Based on historical meteorological data for the same period and the spatial distribution range boundary, the historical statistical average values ​​of temperature and visibility within the unit are calculated to obtain the unit-level spatiotemporal reference values.

[0052] Based on the real-time meteorological element data in the mapping table, the absolute temperature deviation and absolute visibility deviation are calculated by comparing them with the unit-level spatiotemporal reference values.

[0053] Extract the distribution density data of meteorological monitoring stations within the unit, and perform spatial density weighted calculations on the absolute temperature deviation and absolute visibility deviation to obtain the temperature density weighted deviation value and the visibility density weighted deviation value.

[0054] Anomaly level quantification is generated based on temperature density weighted deviation, visibility density weighted deviation, and environmental abrupt change characteristic data; regional meteorological characteristic indicators containing anomaly level and evolution trend are obtained based on the temporal evolution slope of meteorological elements.

[0055] In this embodiment of the invention, each record in the unit-element mapping table includes fields such as "Site ID", "Belonging Meteorological Response Unit", "East Longitude Coordinates (accurate to 0.01°)", and "North Latitude Coordinates (accurate to 0.01°)". For example, the mapping table of a certain unit contains 5 meteorological monitoring stations:

[0056] Station A: ID is QX001, longitude E116.30°, latitude N39.90°;

[0057] Station B: ID is QX002, longitude E116.40°, latitude N39.85°;

[0058] Station C: ID is QX003, longitude E116.25°, latitude N39.70°;

[0059] Station D: ID is QX004, longitude E116.35°, latitude N39.75°;

[0060] Station E: ID is QX005, longitude E116.45°, latitude N39.80°;

[0061] Extract the "Site ID-Longitude-Latitude" information for these 5 stations from the table, organize them according to the "ID (Latitude and Longitude)" format to form a spatial coordinate set. Compare the latitude values ​​of all stations in the spatial coordinate set: Station A is 39.90°N, Station B is 39.85°N, Station C is 39.70°N, Station D is 39.75°N, and Station E is 39.80°N. Among them, Station C's latitude of 39.70°N is the smallest (i.e., the southernmost) of all values. Therefore, Station C (E116.25°N, N39.70°N) is selected as the reference point. With the reference point C as the center, the direction of true north (i.e., the direction of increasing latitude) is 0°. The angle of the line connecting the reference point to the other stations is the azimuth angle. Station D (E116.35°N, N39.75°N) is northeast of the reference point. Based on the geographical location, the angle between the line connecting the reference point and true north is approximately 30° clockwise. Therefore, the azimuth angle is... The azimuth is 45°; Station E (E116.45°, N39.80°) is located northeast of the reference point, with the line connecting it to the due north direction making an angle of approximately 45°; Station B (E116.40°, N39.85°) is located northeast of the reference point, with an angle of approximately 60° and an azimuth of 60°; Station A (E116.30°, N39.90°) is located northeast of the reference point, with an angle of approximately 80° and an azimuth of 80°; If the azimuths are the same (e.g., assuming that the azimuths of Station F and Station D are both 30°), then the straight-line distance from the two points to the reference point is calculated, that is, the difference in latitude and longitude between Station D and C is smaller (the distance is closer), so Station D is ranked before Station F; Finally, the coordinates are sorted from smallest to largest according to the azimuth, resulting in the coordinate sequence: [Station D (30°), Station E (45°), Station B (60°), Station A (80°)].

[0062] Starting from the reference point C, add the reference point C as the initial vertex to the convex hull vertex set, which is now [C]. Select the first point D in the sorted coordinate sequence and add it to the convex hull vertex set, which is now updated to [C, D]. Since the set contains only two points, convexity is not checked yet. Continue to select the next point E in the sorted coordinate sequence and determine whether point E satisfies the convex polygon vertex condition: by observing the relative positions of the three points C, D, and E, point D is located 30 degrees northeast of point C. In the 45° direction, point E is located 45° northeast of point D. The line connecting the three points forms a "left turn" (i.e., point E is to the left of the line connecting points C and D), therefore point E satisfies the convexity condition. Add point E to the convex hull vertex set, which is now updated to [C, D, E]. Next, select the next point B in the sorted coordinate sequence and determine if point B satisfies the convex polygon vertex condition: point E is located 45° northeast of point D, point B is located 60° northeast of point E, and point B is to the left of the line connecting points D and E. The line connecting the three points forms a "left turn" (i.e., point E is to the left of the line connecting points D and E). "Turn left", therefore point B satisfies the convexity condition; add point B to the convex hull vertex set, now updated to [C, D, E, B]; select the next point A in the sorted coordinate sequence and determine if point A satisfies the convex polygon vertex condition: point B is located at 60° of point E, point A is located at 80° of point B, and point A is to the left of the line connecting point E and point B, forming a "turn left" line, therefore point A satisfies the convexity condition; add point A to the convex hull vertex set, finally obtaining the convex hull vertex set [C, D, E, B]. [A]; If the point to be judged in the sorted coordinate sequence (such as the newly added point G) does not meet the convex polygon vertex condition (i.e., the point is located to the right of the line connecting the last two points in the current vertex set, forming a "right turn"), then remove the last point in the vertex set (such as point B), and re-judge the direction of the line connecting the point to be judged with the remaining last two points in the vertex set (such as points E and A); repeat the above removal and judgment operations until the point to be judged meets the "left turn" condition (i.e., it is located to the left of the line connecting the corresponding two points), and then add it to the convex hull vertex set. After determining all points, the coordinates of the convex hull vertices are connected sequentially according to the filtering order: starting from the reference point C (116.25°E, 39.70°N), points D (116.35°E, 39.75°N), E (116.45°E, 39.80°N), B (116.40°E, 39.85°N), and A (116.30°E, 39.90°N) are connected in sequence, and the last point A is connected to the reference point C to form a closed polygon; the boundary of this closed polygon covers all meteorological monitoring stations and serves as the spatial distribution range of meteorological elements within the meteorological response unit (in the form of an irregular pentagon).

[0063] For the temperature parameter, using the current date (e.g., July 15th) as the baseline, historical temperature data for the same period (July 15th) over the past five years (2020-2024) were selected from five meteorological monitoring stations within the aforementioned spatial distribution range. Daily temperature data were divided into 24-hour segments (e.g., 8:00-9:00, 9:00-10:00, etc.), and the regional average temperature value was calculated for the five stations within each hourly segment (e.g., the average temperature value for the five stations during the 8:00-9:00 period on July 15, 2020). The temperatures were 25.1℃, 25.3℃, 25.5℃, 25.2℃, and 25.4℃ respectively, with the regional average for this period being 25.3℃. Then, the average of the regional averages for the same hourly period over the past 5 years was calculated as the spatiotemporal reference value for that hourly period (e.g., the regional averages for the period from 8:00 to 9:00 on July 15, 2020-2024 were 25.3℃, 24.9℃, 25.5℃, 25.1℃, and 25.2℃ respectively, so the temperature reference value for that period was 25.2℃).

[0064] For visibility parameters, daily visibility data is divided into preset time periods (6:00-8:00 AM, 12:00-2:00 PM, and 6:00-8:00 PM). For each time period, the regional average visibility data from five stations is calculated (e.g., on July 15, 2020, from 6:00 AM to 8:00 AM, the visibility at the five stations was 980 meters, 1020 meters, 1050 meters, 950 meters, and 1000 meters respectively, and the regional average for this time period is 1000 meters). Then, the average of the regional averages for the same time period over the past five years is calculated as the spatiotemporal reference value for visibility during that time period (e.g., the regional averages for the time periods from 6:00 AM to 8:00 AM on July 15, 2020-2024 are 1000 meters, 950 meters, 1050 meters, 980 meters, and 1020 meters respectively, so the reference value for visibility during that time period is 1000 meters).

[0065] The calculated temperature reference values ​​(divided by hourly segments) and visibility reference values ​​(divided by preset time periods) are associated and bound with the corresponding meteorological response units (such as U001) to form unit-level spatiotemporal reference values. For example, the temperature reference value for unit U001 from 8:00 to 9:00 is 25.2℃, and the temperature reference value from 9:00 to 10:00 is 25.5℃ (and so on for the other 22-hour segments); the visibility reference value is 1000 meters from 6:00 to 8:00 in the morning, 1500 meters from 12:00 to 14:00 in the afternoon, and 1200 meters from 18:00 to 20:00 in the evening.

[0066] From the unit-element mapping table, extract the real-time temperature and visibility data (real-time monitoring values ​​of all stations within the unit) for the current hour (or time period). For each station, subtract the corresponding hourly temperature baseline from the real-time temperature and take the absolute value to obtain the absolute temperature deviation for that station. Similarly, subtract the corresponding time period's visibility baseline from the real-time visibility and take the absolute value to obtain the absolute visibility deviation for that station. For example, if a station's real-time temperature is 26.5℃ and the baseline is 25.2℃, the absolute temperature deviation is 1.3℃. Calculate the average of the absolute temperature deviations for all stations within the unit as the absolute temperature deviation for that unit. Similarly, calculate the average of the absolute visibility deviations for all stations as the absolute visibility deviation for that unit. Statistical analysis of the atmospheric conditions within the unit is then performed. The distribution density (number of stations / km²) is obtained by dividing the number of monitoring stations by the unit area. For example, if there are 8 stations within 10 km², the density is 0.8 stations / km². The density is divided into three levels: high density (≥1 station / km²), medium density (0.5-1 station / km²), and low density (less than 0.5 stations / km²). Weights are assigned according to the density level (1.2 for high density, 1.0 for medium density, and 0.8 for low density). The absolute temperature deviation of the unit is multiplied by the corresponding weight to obtain the temperature-density weighted deviation value. Similarly, the absolute visibility deviation is multiplied by the weight to obtain the visibility-density weighted deviation value. For example, if the absolute temperature deviation of a medium-density unit is 1.3℃, the weighted deviation value is 1.3 × 1.0 = 1.3℃.

[0067] The system tracks abrupt environmental events occurring within the past 24 hours, specifically including sudden temperature changes and sudden drops in visibility. Sudden temperature changes are defined as the number of times the temperature changes by more than 5°C within one hour, while sudden drops in visibility are defined as the number of times visibility decreases by more than 500 meters within 30 minutes (e.g., one sudden temperature rise and two sudden drops in visibility). The anomaly level is calculated as follows: 1 point is awarded for every 1°C increase in the temperature density-weighted deviation; 1 point is awarded for every 100 meters increase in the visibility density-weighted deviation; 3 points are awarded for each sudden temperature change; and 3 points are awarded for each sudden drop in visibility. These scores are summed to obtain the anomaly level. For example, a temperature deviation of 1.3°C (1 point), a visibility deviation of 240 meters (2 points), one sudden temperature change (3 points), and two sudden drops in visibility (6 points) would result in a total of 12 points. Simultaneously, real-time meteorological data from the past 6 hours (1 hourly) is also used. (Data points), calculate the changes at adjacent time points and take the average to obtain the evolution slope, which reflects the average rate of change per hour. For example, if the temperature rose from 23.5℃ to 27.0℃ in the past 6 hours, with an average hourly temperature increase of 0.6℃, then the temperature evolution slope is 0.6. When calculating the visibility evolution slope, if the visibility gradually decreased from an initial 1000 meters to 652 meters in the past 6 hours, the total decrease is 1000 meters minus 652 meters, resulting in 348 meters. Dividing the total decrease of 348 meters by 6 hours gives an average hourly decrease of 58 meters, so its slope is -58. Finally, the anomaly level quantification value, along with the time evolution slopes of temperature and visibility, are used as regional meteorological characteristic indicators, for example, "Anomaly level 12, temperature shows an upward trend (0.6℃ per hour), visibility shows a downward trend (58 meters per hour)."

[0068] The spatial distribution boundary of meteorological elements is generated based on the coordinates of actual monitoring stations using the convex hull algorithm, avoiding the limitations of fixed area division and more realistically reflecting the natural distribution range of meteorological elements. The unit-level spatiotemporal reference value calculated based on historical data from the same period and the spatial boundary fully considers the historical patterns and spatial characteristics of regional meteorology, making the reference value more consistent with the actual situation of specific units and providing a reliable reference for judging whether real-time data is abnormal. The absolute deviation reflects the intuitive difference between real-time data and the reference value, while the spatial density-weighted deviation value further considers the differences in the density of monitoring station distribution, making the deviation assessment more consistent with the monitoring conditions of different regions. The anomaly level quantification value includes deviations in temperature and visibility, as well as environmental abrupt changes, comprehensively quantifying meteorological anomalies with multi-dimensional information, avoiding the one-sidedness of single-element assessment, and more comprehensively reflecting the abnormal situation of regional meteorology. The temporal evolution slope of meteorological elements not only reflects the current anomaly level but also clearly shows the future changing trend of meteorological elements (such as warming, cooling, and increased or decreased visibility), providing a dynamic basis for the formulation of early warning and response measures.

[0069] In a preferred embodiment of the present invention, generating tiered early warning instructions and response measures bound to geographic units based on regional meteorological characteristic indicators and a pre-set control strategy library may include:

[0070] Analyze regional meteorological characteristic indicators to obtain quantitative values ​​of anomaly levels and the temporal evolution slope of meteorological elements;

[0071] The abnormality level quantification value is matched with the hierarchical rules in the preset control strategy library. When the abnormality level quantification value is within the preset level range, the corresponding warning level is activated, specifically including:

[0072] Analyze the hierarchical rules in the pre-set control strategy library, and extract the minimum and maximum values ​​of the abnormal level range and the associated warning level identifier for each rule;

[0073] The anomaly level quantification value is compared with the numerical range of the grading rules. The comparison process includes:

[0074] When the abnormality level quantification value is greater than or equal to the minimum value of the numerical range and less than the maximum value of the numerical range, the primary warning level indicator is activated.

[0075] When the abnormality level quantification value is greater than or equal to the maximum value of the numerical range and less than the preset mutation threshold, the intermediate warning level indicator is activated.

[0076] When the anomaly level quantification value is greater than or equal to the preset mutation threshold, the advanced warning level indicator is activated.

[0077] The warning level is dynamically adjusted based on the temporal evolution slope of meteorological elements to generate a dynamic warning level.

[0078] Based on the dynamic early warning level, response measure templates are bound from the pre-set control strategy library. Based on the geographical attribute data and closed boundaries of the meteorological response unit, hierarchical early warning instructions and customized response measures bound to the geographical unit are generated.

[0079] In this embodiment of the invention, key data is extracted from regional meteorological characteristic indicators to clarify the specific numerical values ​​of the anomaly level quantification. For example, from "anomaly level 15, temperature is rising (0.8℃ per hour), visibility is falling (60 meters per hour)," the anomaly level quantification value is extracted as 15. Simultaneously, the temporal evolution slope of meteorological elements is separated, i.e., the temperature evolution slope is 0.8℃ per hour, and the visibility evolution slope is -60 meters per hour (the negative sign indicates a decrease), and the direction (rising or falling) and rate of slope change are recorded. The hierarchical rules in the pre-set control strategy library are analyzed, and the hierarchical rule document is retrieved. The document clearly defines the specific parameters of each rule; for example, the first rule defines the minimum anomaly level numerical range as 5 and the maximum as 10, associated with the primary warning level identifier "blue warning"; the second rule defines the minimum numerical range as 10 and the maximum as 20, associated with the intermediate warning level identifier "yellow warning"; the pre-set mutation threshold is set to 20. When the anomaly level quantification value reaches or exceeds 20, it is associated with the advanced warning level identifier "red warning." These rule parameters are organized into a table for easy reference. Subsequent comparisons were performed. The table included columns such as "Rule Number", "Minimum Value", "Maximum Value", "Warning Level", and "Mutation Threshold". The extracted anomaly level quantification value of 15 was compared with the numerical range in the grading rules one by one. First, it was checked whether the first rule (5≤15<10) was met. Obviously, 15 is not less than 10, so it does not meet the rule. Then, the second rule (10≤15<20) was checked. 15 met the condition, so the intermediate warning level "Yellow Warning" was initially activated. Since 15 is less than the preset mutation threshold of 20, there is no need to activate the advanced warning. The current preliminary warning level is determined to be intermediate.

[0080] A slope adjustment rule is set: when the temperature evolution slope is positive (rising) and exceeds 0.5℃ per hour, the warning level is raised by one level; when the visibility evolution slope is negative (falling) and exceeds 50 meters per hour, the warning level is raised by one level; if the slope change is gradual (temperature ≤ 0.3℃ per hour, visibility ≥ -30 meters per hour), the warning level is lowered by one level; otherwise, the original level is maintained. Taking the current data as an example, a temperature evolution slope of 0.8℃ / hour (> 0.5℃ / hour) triggers the upgrade condition; a visibility evolution slope of -60 meters / hour (less than -50 meters / hour) triggers the upgrade condition again. The initial warning level is medium. After two upgrades, the dynamic warning level is adjusted to high-level "red warning". Based on the dynamic warning level "red warning", the corresponding response template is retrieved from the pre-set control strategy library. The high-level warning template includes "closing outdoor construction sites", "suspending classes in primary and secondary schools", and "highway..." Measures such as "speed limit of 60 km / h on highways" are implemented. Based on the geographical attribute data of the meteorological response unit (e.g., the unit includes 3 schools, 2 construction sites, and 1 section of expressway) and the closed boundary (116.35°–116.45°E, 39.85°–39.95°N), the template is customized. Specifically, the names of the 3 schools are clearly marked (e.g., XX Primary School, XX Middle School), the locations of the 2 construction sites are specified (e.g., the construction site at the intersection of XX Road and XX Street), and the specific sections of the expressway are limited (e.g., XX Expressway K100–K120 section). The final generated graded warning instructions include the warning level "red", the boundary coordinates of the affected area, the list of specific locations involved, and customized response measures (e.g., "XX Primary School will be closed before 10:00 today" and "XX Expressway K100–K120 section will have a speed limit of 60 km / h from 9:00"). Each instruction is also given a unique number.

[0081] By analyzing regional meteorological characteristic indicators in detail, the degree and trend of meteorological anomalies can be accurately grasped, providing an accurate basis for early warning. The clear comparison process of the classification rules makes the determination of warning levels more standardized and transparent, avoiding subjective arbitrariness. The dynamic adjustment of the evolution slope of meteorological elements enables the warning level to respond to meteorological changes in real time, improving the timeliness and flexibility of early warning. Customized response measures based on geographical attributes make the instructions more in line with the actual situation of specific regions, improving the operability and effectiveness of response measures and minimizing the impact of meteorological disasters.

[0082] In a preferred embodiment of the present invention, a distributed architecture is used to synchronize hierarchical instructions to roadside equipment and user terminals, collect instruction execution status data in real time, and verify response effects to form a closed-loop feedback, which may include:

[0083] Through the routing nodes of the distributed architecture, the graded early warning instructions are synchronized to the corresponding roadside equipment and user terminals according to the closed boundary space of the meteorological response unit.

[0084] Real-time collection of execution log data from roadside equipment and operation response data from user terminals to generate instruction execution status dataset;

[0085] Compare the instruction execution status dataset with the preset response effect threshold, and count the qualified units that simultaneously meet the conditions; the qualified units that simultaneously meet the conditions are the units whose roadside equipment response delay is less than or equal to the delay threshold and whose user response ratio is greater than or equal to the ratio threshold.

[0086] Based on the proportion of compliant units, the hierarchical rules of the control strategy library are updated to complete the closed-loop feedback of meteorological response.

[0087] In this embodiment of the invention, the distributed architecture includes 10 routing nodes, each responsible for covering an area of ​​50 square kilometers. The nodes are connected via a fiber optic network, and the data transmission rate is maintained above 100 Mbps. First, the system matches the graded warning instructions with the closed boundary coordinates of the meteorological response units. For example, the meteorological response unit numbered U001 has a closed boundary of 116.3000°–116.3500° E and 39.9000°–39.9500° N. The system automatically identifies this area as being the responsibility of routing node 3. After receiving the instruction, routing node 3 first parses the area range in the instruction and filters out the list of roadside equipment within that range, including 20 traffic lights, 15 variable message signs, and 5 road monitoring cameras. Each device has a unique hardware code (such as TL-001, VB-002, etc.). At the same time, through user terminal location data, it filters out the user terminals currently in the area, totaling 1200 mobile APP users, each user terminal corresponding to a unique device ID.

[0088] The distribution process adopts a "broadcast + confirmation" mechanism. Router node 3 sends instruction data packets to roadside equipment, including instruction number (such as YJ20250728001), execution time (such as 9:00-12:00), and specific operation requirements (such as traffic lights flashing yellow and variable message signs displaying "Rainstorm Warning, Slow Down"). Pop-up messages are pushed to user terminals, including warning level, scope of impact, and risk avoidance suggestions. After receiving a command, each device and terminal must return a confirmation message within 10 seconds. Devices that do not receive confirmation will be marked as "pending retransmission," and the system will resend the command after 30 seconds, with a maximum of 3 retransmissions. Execution log data for roadside equipment is uploaded every 5 minutes. Traffic light logs include "command reception time," "start of flashing yellow light time," and "continuous status" (e.g., normal execution, fault interruption). For example, the TL-001 log shows "8:59:30 command received, 9:00:00 flashing started, continuous normal." Variable message sign logs include "content refresh time" and "display status." For example, the VB-002 log shows "8:59:45 command received, 9:00:10 refresh completed, display normal." User terminal operation responses... Data is collected through the APP backend, including "message reception time", "whether clicked to view", "whether shared with others", and "whether the recommended actions were taken" (such as staying home to avoid risks or adjusting travel routes). For example, the record of user ID 138xxxx5678 shows "received pop-up at 8:59:50, clicked to view at 8:59:55, and selected to adjust travel routes at 9:05:20". This data is classified and integrated into unit U001 according to the meteorological response unit to create a dedicated dataset, which includes the execution logs of 20 traffic lights, the execution logs of 15 variable message signs, and the operation records of 1200 user terminals. Each data entry is labeled with the corresponding device / user ID, timestamp, and specific behavior description, forming a complete instruction execution status dataset.

[0089] The preset response effect thresholds are set as follows: a roadside device response delay threshold of 5 seconds (the time difference from receiving the command to starting execution ≤ 5 seconds), and a user response ratio threshold of 60% (the proportion of users clicking to view the message to the number of receiving users ≥ 60%). The "command reception time" and "start execution time" of each device are compared one by one. For example, TL-001's reception time is 8:59:30, and its start execution time is 9:00:00, with a delay of 30 seconds, exceeding the 5-second threshold and thus deemed substandard; VB-002's reception time is 8:59:45, and its start execution time is 9:00:10, with a delay of 25 seconds, also failing to meet the standard. Statistical analysis revealed that 18 out of 20 traffic lights had delays exceeding 5 seconds. Of the 15 variable message signs, 12 had a delay exceeding 5 seconds, meaning the roadside equipment in this unit did not meet the overall delay requirement. Assuming unit U001 has 1200 receiving users, with 800 clicking to view messages, the response rate is 800÷1200≈66.7%, exceeding the 60% threshold and meeting the user response requirement. However, due to the substandard roadside equipment, this unit did not simultaneously meet both conditions and was judged as a non-compliant unit. The same calculation was performed on all 10 meteorological response units. For example, if the roadside equipment delay in unit U002 is ≤5 seconds (compliant) and the user response rate is 72% (compliant), then it is judged as a compliant unit. Finally, a total of 6 compliant units were identified, accounting for 60% of the total number of units.

[0090] The system establishes a correspondence between the proportion of compliant units and rule adjustments. Specifically, when the proportion of compliant units is less than 50%, the trigger threshold for the warning level is lowered (e.g., the primary warning is adjusted from 5-10 points to 4-9 points); when the proportion of compliant units is between 50% and 80%, the existing rules are maintained; and when the proportion is 80% or higher, the trigger threshold is increased (e.g., the primary warning is adjusted to 6-11 points). Currently, the proportion of compliant units is 60%, falling within the 50%-80% range. Therefore, the tiered rules for unit U001 will not be adjusted for the time being. However, regarding the issue of excessive latency for roadside equipment in this unit, the system automatically adds supplementary rules to the control policy library: "For roadside equipment with three consecutive instances of excessive response latency, bandwidth resources will be allocated preferentially during the next warning to ensure priority execution of instructions." The updated control policy library will be synchronized to the instruction generation module. The new rules will take effect during the next weather warning. For example, when the traffic lights in unit U001 receive instructions next time, they will be prioritized by the routing node, reducing response latency and forming a complete closed-loop feedback mechanism for weather response.

[0091] The distributed architecture ensures precise distribution of tiered early warning commands, enabling rapid coverage of roadside equipment and user terminals in target areas and preventing omissions or delays in information transmission. Real-time collected execution status data comprehensively reflects the implementation of commands, providing a solid basis for effectiveness verification. Clear threshold standards and compliance judgment rules ensure objective and fair evaluation of response effectiveness, accurately identifying areas requiring improvement. Updating the control strategy library based on compliance ratios allows the system to dynamically optimize rules according to actual execution, continuously improving the effectiveness and adaptability of early warning commands.

[0092] Embodiments of the present invention also provide a computing device, including: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, performs the system as described above. All implementations in the above system embodiments are applicable to this embodiment and can achieve the same technical effects.

[0093] Embodiments of the present invention also provide a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the system as described above. All implementations in the above system embodiments are applicable to this embodiment and can achieve the same technical effects.

[0094] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. An intelligent weather answering system, characterized by, The method comprises the following steps: A data acquisition module is used to collect meteorological elements, environmental conditions and behavior trajectories of user positions in real time, and to construct a multi-dimensional dynamic data set associated with time and space; A unit division module is used to divide continuous meteorological response units based on the multi-dimensional dynamic data set through time and space feature extraction, and to establish a mapping relationship table between meteorological elements and geographical units, including: calculating the number of trajectory points per unit area according to the time and space distribution density of the behavior trajectory, and obtaining the central coordinates of the area with a density exceeding a density determination threshold as hotspot coordinates; based on the hotspot coordinates, obtaining real-time data of spatially topologically adjacent meteorological monitoring stations, calculating the absolute value of the temperature and visibility change rate between topologically adjacent stations, and generating a meteorological element gradient field with the hotspot coordinates as the core; based on the topologically adjacent station data in the meteorological element gradient field, performing spatial interpolation calculation to generate continuous contour distribution data of temperature change rate and visibility change rate; extracting the envelope region coordinate set with a temperature change rate less than or equal to a temperature change rate determination threshold in the continuous contour distribution data; judging whether the envelope region meets the standard that the wind speed direction standard deviation is less than or equal to a direction consistency threshold, and the terrain elevation range is less than or equal to an elevation difference threshold; topologically merging the adjacent envelope region coordinate set with a temperature change rate less than or equal to a temperature change rate determination threshold, a visibility change rate less than or equal to a visibility change rate determination threshold and environmental continuity meeting the standard, to obtain meteorological response unit closed boundary data; attributing real-time meteorological element data to geographical coordinates and correlating it to the meteorological response unit closed boundary to obtain a unit-element mapping relationship table, specifically: assigning a unique identifier to each meteorological response unit, recording its boundary coordinate range, correlating real-time meteorological data to the corresponding unit according to the coordinate attribution, calculating the average value of the air temperature and humidity and the maximum value of the visibility in the unit, updating the table every ten minutes, and cleaning up expired data to form a real-time updated mapping relationship table; An analysis module is used to determine the spatial distribution range of meteorological elements by using a convex hull algorithm based on the mapping relationship table, calculate the local deviation coefficient of meteorological elements in each unit, and generate regional meteorological characteristic indexes; An instruction generation module is used to generate hierarchical early warning instructions and response measures bound to geographical units based on the regional meteorological characteristic indexes and a pre-set control strategy library; A feedback module is used to synchronize the hierarchical instructions to roadside devices and user terminals through a distributed architecture, collect instruction execution state data in real time and verify the response effect to form a closed-loop feedback.

2. The intelligent weather answering system according to claim 1, wherein, Based on the mapping relationship table, the spatial distribution range of meteorological elements is determined by using a convex hull algorithm, the local deviation coefficient of meteorological elements in each unit is calculated, and regional meteorological characteristic indexes are generated, including: Based on the mapping relationship table, the spatial distribution range of meteorological elements is determined by using a convex hull algorithm, the local deviation coefficient of meteorological elements in each unit is calculated, and regional meteorological characteristic indexes are generated, including: Based on the mapping relationship table, the spatial distribution range of meteorological elements is determined by using a convex hull algorithm, the local deviation coefficient of meteorological elements in each unit is calculated, and regional meteorological characteristic indexes are generated, including: Based on the mapping relationship table, the spatial distribution range of meteorological elements is determined by using a convex hull algorithm, the local deviation coefficient of meteorological elements in each unit is calculated, and regional meteorological characteristic indexes are generated, including: Extract the distribution density data of meteorological monitoring sites in the unit, and perform spatial density weighted calculation on the absolute deviation amount of temperature and the absolute deviation amount of visibility respectively to obtain the temperature density weighted deviation value and the visibility density weighted deviation value; According to the temperature density weighted deviation value, the visibility density weighted deviation value and the environmental mutation characteristic data, an abnormal grade quantization value is generated; according to the time evolution slope of the meteorological element, a regional meteorological characteristic index containing the abnormal grade and the evolution trend is obtained.

3. The intelligent weather answering system according to claim 2, wherein, Based on the mapping relationship table, the coordinate set of the meteorological monitoring site is extracted, and the convex hull algorithm is used to generate the spatial distribution range boundary of the meteorological element, including: Extract the geographic coordinates of all meteorological monitoring sites from the mapping relationship table to form a spatial coordinate set; Identify the coordinate point with the smallest latitude in the spatial coordinate set as the reference point, and calculate the azimuth angle of the reference point and the remaining coordinate points in the spatial coordinate set, and generate a sorted coordinate sequence in ascending order of azimuth angle; Traverse the sorted coordinate sequence, and judge whether the current coordinate point meets the convex polygon vertex condition in turn, and add the coordinate point meeting the convexity condition to the convex hull vertex set; The coordinate points in the convex hull vertex set are connected in sequence to construct a closed polygon boundary, and the spatial distribution range boundary of the meteorological element is generated.

4. The intelligent weather answering system according to claim 3, wherein, According to the regional meteorological characteristic index and the preset control strategy library, a hierarchical early warning instruction and response measure bound to the geographical unit are generated, including: Analyzing the regional meteorological characteristic index to obtain the abnormal grade quantization value and the time evolution slope of the meteorological element; Match the abnormal grade quantization value with the hierarchical rules in the preset control strategy library, and activate the corresponding warning level when the abnormal grade quantization value is in the preset grade interval; Based on the time evolution slope of the meteorological element, the warning level is dynamically adjusted to generate a dynamic warning level; According to the dynamic warning level, bind the response measure template from the preset control strategy library, and generate a hierarchical early warning instruction and customized response measure bound to the geographical unit according to the geographical attribute data and closed boundary of the meteorological response unit.

5. The intelligent weather answering system according to claim 4, wherein, Match the abnormal grade quantization value with the hierarchical rules in the preset control strategy library, and activate the corresponding warning level when the abnormal grade quantization value is in the preset grade interval, including: Analyzing the hierarchical rules in the preset control strategy library to extract the minimum value of the abnormal grade numerical range, the maximum value of the numerical range and the associated warning level identifier defined by each rule; Compare the abnormal grade quantization value with the numerical range of the hierarchical rule, the comparison process includes: When the abnormal grade quantization value is greater than or equal to the minimum value of the numerical range and less than the maximum value of the numerical range, activate the primary warning level identifier; When the abnormal grade quantization value is greater than or equal to the maximum value of the numerical range and less than the preset mutation threshold, activate the intermediate warning level identifier; When the abnormal grade quantization value is greater than or equal to the preset mutation threshold, activate the high-level warning level identifier.

6. The intelligent weather answering system according to claim 5, wherein, Synchronize the hierarchical instruction to roadside devices and user terminals through a distributed architecture, collect instruction execution state data in real time and verify the response effect to form a closed loop feedback, including: Through the routing node of the distributed architecture, synchronize the hierarchical early warning instruction to the corresponding roadside device and user terminal according to the closed boundary space range of the meteorological response unit; Real-time collection of the execution log data of the roadside device and the operation response data of the user terminal, to generate instruction execution state data set; Comparing the instruction execution state data set with the preset response effect threshold value, and counting the qualified units that meet the conditions at the same time; the qualified units that meet the conditions at the same time are the units whose roadside device response time delay is less than or equal to the time delay threshold value and the user response proportion is greater than or equal to the proportion threshold value; Based on the proportion of qualified units, updating the hierarchical rules of the management strategy library to complete the meteorological response closed-loop feedback.

7. A computing device, comprising: Comprise: One or more processors; A storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, so that the one or more processors implement the system as claimed in any one of claims 1 to 6.

8. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a program, which is executed by the processor to implement the system as claimed in any one of claims 1 to 6.

Images