Intelligent rainfall monitoring and early warning system based on phased array radar

Abstract

The invention relates to an intelligent rainfall monitoring and early warning system based on a phased array radar, and belongs to the technical field of phased array radar rainfall detection data analysis and processing. According to the system, firstly, a moving track of a rainfall unit is determined to complete moving direction prediction and moving distance prediction of each monitoring point after the current moment; therefore, according to the deviation between the direction angle between the to-be-inspected monitoring point and other monitoring points and the predicted motion direction of other monitoring points, the distance deviation between the to-be-inspected monitoring point and the predicted position of other monitoring points and the current monitoring values of other monitoring points, the current monitoring values of other monitoring points can be obtained. The influence of other monitoring points on the monitoring point to be inspected at the next monitoring moment is determined, and the influence is used for representing the influence of other monitoring points on the future weather of the monitoring point to be inspected, which cannot be represented by a traditional prediction method under the rapid change of the weather, so that the monitoring value prediction of the monitoring point to be inspected can be more accurately completed; and the rainfall monitoring and early warning accuracy of the phased array radar in a weather rapid change scene is improved.

Description

Technical Field

[0001] This invention relates to the field of phased array radar rainfall detection data analysis and processing technology, and in particular to a smart rainfall monitoring and early warning system based on phased array radar. Background Technology

[0002] Phased array radar is a radar system composed of multiple transmitting / receiving units. By adjusting the phase of the transmitted signals of each unit, it can quickly complete scanning and positioning. Compared with traditional radar, phased array radar has a faster scanning speed and higher accuracy. In rainfall monitoring and early warning systems, it has higher real-time performance, accuracy, and flexibility, making it an increasingly important component of rainfall early warning systems.

[0003] However, in scenarios of sudden, short-duration heavy rainfall, rain cells generally exhibit rapid movement and rapid generation, dissipation, merging, and splitting. Weather conditions change rapidly and differ significantly from historical weather patterns. Although phased array radar is more flexible and precise, capable of forming higher-resolution monitoring points during meteorological monitoring and obtaining more accurate monitoring values ​​at each point, thus acquiring more detailed and accurate historical data, it still suffers from insufficient accuracy in rainfall monitoring and early warning in such scenarios. This is because long-term historical data lacks reference value for predicting current rapidly changing weather scenarios, while recent historical data is insufficient in terms of data volume and interpretation of current rapidly changing weather scenarios.

[0004] Therefore, how to further improve the accuracy of phased array radar in monitoring and warning of rainfall in rapidly changing weather scenarios has become an urgent technical problem to be solved. Summary of the Invention

[0005] In view of this, the present invention provides a smart rainfall monitoring and early warning system based on phased array radar to solve the technical problem of insufficient accuracy of rainfall monitoring and early warning by phased array radar in scenarios of rapid weather changes.

[0006] The present invention provides a smart rainfall monitoring and early warning system based on phased array radar, comprising:

[0007] The historical data acquisition module is used to acquire the monitoring values ​​of each monitoring point in the monitoring area by the phased array radar within a set time period, and to determine the rainfall cells at each monitoring time.

[0008] The rainfall cell trajectory determination module is used to take any rainfall cell at the current monitoring time as the target cell, determine the proximity of the target cell to each rainfall cell at the previous monitoring time, and determine the movement trajectory of the target cell at the set time period based on the proximity of the positions.

[0009] The module for determining the degree of influence of the predicted value is used to determine the predicted value of the movement direction and the predicted value of the movement distance of each monitoring point in the target unit at the next monitoring time based on the movement trajectory, thereby determining the predicted value of the movement direction and the predicted value of the movement distance of the monitoring points in each rainfall unit at the current monitoring time, and further determining the degree of influence of any monitoring point on other monitoring points at the current monitoring time.

[0010] The prediction correction and early warning module is used to determine the corrected prediction value of any monitoring point at the current monitoring time at the next monitoring time based on the predicted value of any monitoring point at the current monitoring time at the next monitoring time and the degree of impact, and to complete the rainfall early warning.

[0011] Furthermore, determining the rainfall cells at each monitoring time includes:

[0012] The monitoring points in the monitoring area whose monitored values ​​are greater than a preset monitoring threshold are clustered, and the clusters with an area greater than a preset area threshold are taken as the rainfall units.

[0013] Furthermore, determining the proximity of the target cell to the location of each rainfall cell at the previous monitoring time includes:

[0014] A predetermined number of edge monitoring points that are closest to the target rainfall cell at the previous monitoring time are continuously selected from the edge monitoring points of the target cell to be judged to form a target edge segment. The corresponding edge segment on the target rainfall cell at the previous monitoring time is determined. The proximity of each rainfall cell to the target cell at the previous monitoring time is determined based on the average distance between the target edge segment and the corresponding edge monitoring points in the corresponding edge segment.

[0015] Furthermore, determining the corresponding edge segment on the rainfall cell to be judged at the previous monitoring time that corresponds to the target edge segment includes:

[0016] A predetermined number of consecutive edge monitoring points are randomly selected from the edge monitoring points of the rainfall cell to be judged at the previous monitoring time to form the edge segment to be judged. The distance value between the target edge segment and the edge segment to be judged is determined according to the average distance between the target edge segment and the corresponding edge monitoring point in the edge segment to be judged. The edge segment to be judged corresponding to the maximum distance value is taken as the corresponding edge segment in the rainfall cell to be judged at the previous monitoring time.

[0017] Furthermore, determining the movement trajectory of the target unit within the set time period based on the proximity of the positions includes:

[0018] The target cell is taken as the cell to be analyzed. At the previous monitoring time, a set number of rainfall cells that are closest to the target cell in location are selected as the undetermined rainfall cells. The undetermined rainfall cells at the previous time are each selected as new cells to be analyzed. At the monitoring time before the previous monitoring time, a set number of rainfall cells that are closest to the new cells to be analyzed are selected as new undetermined rainfall cells. The process of determining the undetermined rainfall cells corresponding to the target cell is repeated until all monitoring times within the set time period are traversed.

[0019] Each combination of the target cell and the corresponding undetermined rainfall cells at each monitoring time is determined. The angular change and position change of the rainfall cells in any two adjacent monitoring times in the current cell combination are determined. The cell combination with the smallest angular change and position change is taken as the movement trajectory.

[0020] Furthermore, the step of determining the predicted movement direction and predicted movement distance of each monitoring point in the target unit at the next monitoring time based on the movement trajectory includes:

[0021] Based on the angle change of the target unit at any two adjacent monitoring times in the movement trajectory, predict the predicted angle change of the target unit at the next monitoring time relative to the current monitoring time. Determine the predicted value of the movement direction based on the movement direction of the target unit at the current monitoring time compared to the previous monitoring time and the predicted angle change. Determine the predicted value of the movement distance based on the average movement distance of the target unit at any two adjacent monitoring times in the movement trajectory.

[0022] Furthermore, the step of predicting the predicted angle change of the target unit relative to the current monitoring time at the next monitoring time based on the angle change of the target unit at any two adjacent monitoring times in the movement trajectory includes:

[0023] The angle change of the target unit at any two adjacent monitoring times in the movement trajectory is arranged into an angle change sequence according to time, and curve fitting is performed to obtain an angle change fitting curve. The value of the angle change fitting curve at the next monitoring time is used as the predicted angle change.

[0024] Furthermore, further determining the degree to which any monitoring point is affected by other monitoring points at the current monitoring time includes:

[0025] Taking any monitoring point at the current monitoring time as the target monitoring point, determine the direction angle from any monitoring point in any rainfall cell at the current monitoring time to the target monitoring point, and calculate the direction deviation between the direction angle and the predicted value of the movement direction of any monitoring point in any rainfall cell at the current monitoring time at the next monitoring time;

[0026] Based on the distance between the target monitoring point and any monitoring point in any rainfall cell at the current monitoring time, and the predicted movement distance of any monitoring point in any rainfall cell at the current monitoring time at the next monitoring time, the distance deviation between the target monitoring point and any monitoring point in any rainfall cell at the current monitoring time at the next monitoring time is determined.

[0027] The degree of influence is determined based on the directional deviation, the distance deviation, and the monitoring value of any monitoring point in a rainfall cell at the current monitoring time.

[0028] Furthermore, the degree of influence is as follows:

[0029]

[0030] Among them, H i N represents the degree to which the i-th monitoring point is affected by other monitoring points at the current monitoring time and at the next monitoring time. i This represents the total number of remaining monitoring points within the monitoring area, excluding the rainfall cell containing the i-th monitoring point. This represents the direction angle from the e-th monitoring point to the i-th monitoring point at the current monitoring time. This represents the predicted movement direction of the e-th monitoring point at the current monitoring time at the next monitoring time. express and The directional deviation between them, d i,e D represents the distance between the i-th monitoring point and the e-th monitoring point at the current monitoring time. e d represents the predicted distance traveled by the e-th monitoring point at the current monitoring time at the next monitoring time. i,e -D e |d represents the distance deviation between the i-th monitoring point and the e-th monitoring point at the next monitoring time after the current monitoring time. i,e -D e | indicates the calculation of d. i,e -D e The absolute value of Z, ε is used to ensure that the fraction is meaningful. e This represents the monitoring value of the e-th monitoring point at the current monitoring time, and norm() represents the normalization function.

[0031] Furthermore, determining the predicted value of any monitoring point at the current monitoring time at the next monitoring time includes:

[0032] Based on the monitoring values ​​of any monitoring point at the current monitoring time at each monitoring time within the set time period, the ARIMA algorithm is used to obtain the predicted value of any monitoring point at the current monitoring time at the next monitoring time.

[0033] The advantages of this invention compared to the prior art are:

[0034] This invention's system first determines the movement trajectory of individual rainfall cells to predict the movement direction and distance of each monitoring point at the next predicted time. By combining the deviations in azimuth angle between the monitoring point under investigation and other monitoring points, the predicted movement directions of other monitoring points, the distance deviations between the predicted positions of the monitoring point under investigation and other monitoring points, and the monitoring values ​​of other monitoring points at the current time, the system determines the degree of influence of other monitoring points on the monitoring point under investigation at the next monitoring time. This degree of influence characterizes the impact of other monitoring points on the future meteorological changes of the monitoring point under investigation, which is something that traditional prediction methods cannot characterize under rapidly changing weather conditions. Therefore, by incorporating this influence, the system can more accurately predict the monitoring values ​​of the monitoring point under investigation, improving the accuracy of rainfall monitoring and early warning by phased array radar in rapidly changing weather scenarios. Attached Figure Description

[0035] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0036] Figure 1 This is a structural block diagram of a rainfall intelligent monitoring and early warning system based on phased array radar provided in Embodiment 1 of the present invention. Detailed Implementation

[0037] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of the invention include a particular feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. Furthermore, a particular feature, structure, or characteristic in one or more embodiments may be combined in any suitable form, and the terms "comprising," "including," "having," and variations thereof mean "including, but not limited to," unless otherwise specifically emphasized.

[0038] It should be understood that the sequence number of each step in the following embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0039] To further illustrate the technical solution of the present invention, specific embodiments are described below.

[0040] System Implementation Example:

[0041] See Figure 1 This is a structural block diagram of a rainfall intelligent monitoring and early warning system based on phased array radar provided in Embodiment 1 of the present invention, as shown below. Figure 1 As shown, the system includes: a historical data acquisition module 11, a rainfall cell trajectory determination module 12, a forecast impact determination module 13, and a forecast correction and early warning module 14. The following is a description of each module:

[0042] The historical data acquisition module 11 is used to acquire the monitoring values ​​of each monitoring point in the monitoring area by the phased array radar within a set time period, and to determine the rainfall cells at each monitoring time.

[0043] By installing a phased array radar at the center of the warning area, the radar can perform three-dimensional spatial scanning at set intervals (e.g., 1 minute) at multiple set elevation and azimuth levels. Each scan obtains echo intensity data for the area, in the form of a three-dimensional data volume. Each data point, or monitoring point, includes its coordinates and echo intensity (reflectivity), which is the monitored value. The elevation angle controls the scanning height, and the azimuth angle controls the scanning depth. A 360° scan is performed at each elevation level to obtain echo intensity data at different heights. The number of elevation levels is typically set between 12 and 20, and the elevation angle can be adjusted according to actual weather conditions to allow for higher-frequency scanning of local areas. Echo intensity data from a set time period (e.g., 15 minutes) prior to the predicted time is used as historical data for prediction.

[0044] Since rainfall formation is often related to rain cells (clouds), and clouds can change rapidly under the influence of factors such as wind and temperature, leading to significant variations in rainfall distribution within the monitoring area, it's understandable that rainfall is more likely to increase along the direction of rain cell movement, while the amount and probability of rainfall gradually decrease in directions away from the rain cells. Therefore, the echo intensity at monitoring points along the direction of rain cell movement is more easily affected by the echo intensity at other monitoring points. Thus, to analyze the influence relationship between echo intensity at different monitoring points, it is first necessary to obtain the rainfall movement trend.

[0045] The movement trend of rainfall is the movement trend of individual rainfall cells, so it is first necessary to identify the rainfall cells in the current area at each monitoring time. The echo intensity of rainfall cells is relatively higher. Based on the common echo intensity range of rainfall cells, when the echo intensity of a monitoring point is greater than a preset monitoring threshold (preferably 30 dBZ in this embodiment), the monitoring point can be considered as a monitoring point within a rainfall cell. Clustering the monitoring points within the rainfall cell, if the area of ​​the resulting cluster is greater than a preset area threshold, then the cluster is considered a valid rainfall cell.

[0046] The rainfall cell trajectory determination module 12 is used to determine the proximity of any rainfall cell at the current monitoring time to each rainfall cell at the previous monitoring time, and to determine the movement trajectory of the target cell for a set time period based on the proximity of the target cells.

[0047] Then, based on the changes in the location and echo intensity of rain cells at continuous monitoring times, the movement trend of rain cells is analyzed. Because rain cells may undergo changes such as growth, splitting, and merging, the shape and size of the same rain cell may not be exactly the same at different monitoring times, making it impossible to directly identify the same rain cell at different monitoring times based on its shape. However, the movement direction of rain cells usually has certain regularity. Therefore, based on the correlation of the locations of rain cells at different monitoring times, the position of the same rain cell in the scan data at different monitoring times can be obtained, thereby obtaining the movement trend of the rain cell, that is, determining the movement trajectory of the rain cell.

[0048] The analysis focuses on a specific rainfall cell at the current moment, which is also the last moment in the acquired historical data.

[0049] First, the proximity of each rainfall cell at the previous monitoring time to a rainfall cell at the next monitoring time is determined. Taking the determination of the proximity between a target cell at the current monitoring time and a rainfall cell to be judged at the previous monitoring time as an example, this embodiment does not use the centroid of the rainfall cell to obtain the proximity of the rainfall cells, but rather uses the distance between the same position on the edge of the rainfall cells at adjacent times to determine the proximity. The process is as follows: First, on the edge of the target cell, that is, among the edge monitoring points of the target cell, a predetermined number (preferably 20 in this embodiment) of edge monitoring points that are closest to the rainfall cell to be judged at the previous monitoring time are continuously selected to form a target edge segment. Then, the corresponding edge segment on the rainfall cell to be judged at the previous monitoring time that corresponds to the target edge segment is determined. The proximity between the rainfall cell to be judged at the previous monitoring time and the target cell is determined based on the average distance between the target edge segment and the corresponding edge monitoring points in the corresponding edge segment.

[0050] The method for determining the corresponding edge segment is as follows:

[0051] In the previous monitoring time, a preset number of 20 consecutive edge monitoring points are randomly selected from the edge monitoring points of the rainfall cell to be judged to form the edge segment to be judged. The distance between the target edge segment and the edge segment to be judged is determined based on the average distance between the corresponding edge monitoring points in the target edge segment and the edge segment to be judged. The edge segment to be judged corresponding to the maximum distance value is taken as the corresponding edge segment of the rainfall cell to be judged in the previous monitoring time. The reason for selecting the edge segment corresponding to the maximum distance value as the corresponding edge segment of the target edge segment in the previous monitoring time is that the purpose here is to determine the proximity of the rain cells at adjacent times. Therefore, it is only necessary to consider the movement distance of the same position on the rain cells at adjacent times, such as the movement distance of the left side of the rain cells at two adjacent times. It does not consider the possibility that the rain cells may rotate or deform during movement, causing (partial or all) edge points on the corresponding edge segment to be different from the original edge points on the target edge segment. In other words, the process of determining the corresponding edge segment is not about finding the actual corresponding edge points on the rain cells at two adjacent times, but only about determining the edge segments at the same position on the rain cells at two adjacent times. Since this embodiment first selects the edge segment on the target cell that is closest to the rain cell to be judged at the previous time as the target edge segment, the distance between the target edge segment and the corresponding edge segment should be greater than the distance between the target edge segment and other edge segments on the edge of the rain cell to be judged at the previous time, thus achieving accurate determination of the corresponding edge segment.

[0052] The specific implementation method for determining the corresponding edge segment can be:

[0053] Using 20 consecutive monitoring points as the window length, the system iterates through all edge monitoring points of the target cell at the previous monitoring time, calculating the average distance between each edge monitoring point within the window and the target edge segment of the target cell at the current monitoring time.

[0054]

[0055] Among them, s w This represents the average distance between the edge monitoring point and the target edge segment in the w-th window, where n is the average distance between the edge monitoring point and the target edge segment. w This represents the number of edge monitoring points contained in the w-th window (where n is the number of edge monitoring points). w =20), d w,j This represents the distance between the j-th edge monitoring point in the w-th window and the corresponding j-th edge monitoring point in the target edge segment.

[0056] Where s wThe edge monitoring point in the window corresponding to the maximum value is considered to be the monitoring point in the corresponding edge segment of the target individual at the current monitoring time. According to the arrangement order of the edge monitoring points in the window and the arrangement order of the monitoring points in the target edge segment, such as clockwise or counterclockwise, the edge monitoring points in the two edge segments can be matched one by one to form 20 sets of edge monitoring points.

[0057] Based on the obtained target edge segment and the 20 sets of edge monitoring points formed on its corresponding edge segment, the positional proximity between the rainfall cell in the corresponding edge segment and the target cell in the target edge segment is determined according to the average distance between the target edge segment and the corresponding edge monitoring point in the corresponding edge segment (i.e., the 20 sets of edge monitoring points). This determines the positional proximity between each rainfall cell and the target cell at the previous monitoring time.

[0058]

[0059] Among them, c u This indicates the proximity between the u-th rainfall cell and the target cell at the previous monitoring time, where n is the distance between them. c This represents the number of edge monitoring points in the target edge segment, that is, the number of pairs of edge monitoring points between the target edge segment and its corresponding edge segment (here, n). c =20), Δd u,j This represents the distance between the j-th edge monitoring point in the corresponding edge segment of the u-th rainfall cell at the previous monitoring time and the j-th edge monitoring point in the corresponding edge segment of the target edge segment. This indicates that the smaller the distance, the closer the two rainfall cells are located. `norm()` represents a normalization function, such as norm normalization, etc.

[0060] It should be noted that if two rainfall cells completely overlap or one completely covers the other, it indicates that the rainfall cell has not moved and may have only grown or partially dissipated. Therefore, they are considered to be the same rainfall cell, and the maximum value of their corresponding positional proximity is taken (here, c). u The maximum value is 1).

[0061] Because the distance between different rainfall cells is usually significantly greater than the movement distance of a single rainfall cell between two adjacent monitoring times, the target cell is first selected as the cell to be analyzed. Then, at the previous monitoring time, a set number of rainfall cells (preferably 3 in this embodiment) that are closest to the cell to be analyzed are selected as the pending rainfall cells. That is, at the previous monitoring time, the top 3 rainfall cells that are closest to the current rainfall cell to be analyzed are selected, and it is assumed that these 3 rainfall cells may actually be the same rainfall cell as the current rainfall cell to be analyzed.

[0062] Then, after the determined undetermined rainfall cells are used as new undetermined rainfall cells, at the previous monitoring time (the monitoring time before the previous monitoring time), select the number of rainfall cells (3) that are closest in location to the new undetermined rainfall cells as new undetermined rainfall cells. Repeat the process of determining the undetermined rainfall cells corresponding to the undetermined rainfall cells until all monitoring times within the set time period are traversed.

[0063] Each combination of the target cell and the corresponding undetermined rainfall cells at each monitoring time is determined. The angular change and position change of the rainfall cells in any two adjacent monitoring times in the current cell combination are determined. The cell combination with the smallest angular change and position change is taken as the movement trajectory.

[0064] Specifically, traditional methods for analyzing rainfall cell trajectories rely on the centroid position of the individual rainfall cells. However, when rainfall cells merge or break up, their centroids shift significantly, affecting the analysis of rainfall movement trends. Therefore, based on the relative positions of the 20 closest edge monitoring points obtained above, the direction of movement of the current rainfall cell under analysis relative to each suspected identical rainfall cell is calculated.

[0065]

[0066] Where, θ u n represents the azimuth angle of the cell to be analyzed at the later monitoring time relative to the u-th undetermined rainfall cell at the previous monitoring time. c This represents the number of edge monitoring point groups used to calculate the proximity of locations (here, n). c =20), θ u,j α represents the angle between the line connecting the j-th pair of edge monitoring points, projected onto the same horizontal plane, and the positive x-axis. u α represents the elevation angle of the cell to be analyzed at the later monitoring time relative to the u-th undetermined rainfall cell at the previous monitoring time. u,j This represents the angle of inclination of the line connecting the j-th pair of edge monitoring points relative to the horizontal plane. This indicates the relative movement direction between the unit to be analyzed at the later monitoring time and the u-th rainfall unit at the later monitoring time. Since the radar monitoring points are distributed in three dimensions, the movement direction is represented by a vector composed of its corresponding azimuth and elevation angles.

[0067] The movement of the same rainfall cell exhibits a certain regularity, meaning that the change in the direction of movement of the rainfall cell is relatively small between adjacent monitoring times. After combining the suspected identical rainfall cells to obtain multiple cell combinations, the relative movement direction feature vector corresponding to adjacent monitoring times is calculated within any cell combination. The angle change was determined by using the least squares method to obtain a fitting curve based on the time sequence of each monitoring moment within a set duration (15 minutes). A better fit indicates a stronger regularity of the rainfall cell within the cell combination, suggesting that the group of rainfall cells is more likely to belong to the same rainfall cell. The angle change was calculated as follows:

[0068]

[0069] Where β represents the change in angle. and These represent the relative movement direction vectors of the rain-producing cells at two adjacent monitoring times within the cell combination. The modulus of the vector obtained by the cross product of two vectors is represented by , and atan2() represents the arctangent function, which outputs angle values ​​with positive and negative values.

[0070] Therefore, by combining the root mean square error of the fitted curve of the angle change of the rainfall cells in each cell combination, and the proximity of the rainfall cells between adjacent monitoring times under that combination, the probability that the rainfall cells in each cell combination are the same rainfall cell is calculated. The formula is as follows:

[0071]

[0072] Among them, Y m σ represents the probability that the rainfall cells in the m-th cell combination are the same rainfall cell. m This represents the root mean square error of the fitted curve for the change in the angle of the rainfall cells in the m-th cell combination. This indicates that the smaller the root mean square error, the better the fit of the fitted curve, meaning that the movement direction of the rainfall cells in this cell combination is more regular. m c represents the number of monitoring times included in the m-th individual combination. m,u This indicates the proximity of the u-th rainfall cell in the m-th cell combination to the rainfall cell at the previous moment. `norm()` represents the normalization function.

[0073] The most probable combination of individual cells is considered the movement path of the cell currently being analyzed. Therefore, the movement trajectories of all rainfall cells at the current monitoring time can be obtained.

[0074] The prediction value impact determination module 13 is used to determine the predicted movement direction and predicted movement distance of each monitoring point in the target unit at the next monitoring time based on the movement trajectory, thereby determining the predicted movement direction and predicted movement distance of the monitoring points in each rainfall unit at the current monitoring time, and further determining the degree of influence of any monitoring point on other monitoring points at the current monitoring time.

[0075] During the acquisition of the movement trajectory of each rainfall cell at the current monitoring time, the movement direction and angle change of the rainfall cell relative to the previous monitoring time at each historical monitoring time were obtained, and the angle change fitting function was obtained. Based on the angle change fitting function, the value of the angle change fitting function at the next monitoring time after the current monitoring time can be used as the predicted angle change, and combined with the movement direction of the rainfall cell at the current monitoring time, the predicted value of the movement direction of the rainfall cell at the next monitoring time can be obtained.

[0076] Furthermore, during the acquisition of the movement trajectory of each rainfall cell at the current monitoring time, the position of that rainfall cell at each historical monitoring time can be determined. Therefore, based on the average movement distance of that rainfall cell at any two adjacent monitoring times in the movement trajectory, the predicted movement distance at the next monitoring time after the current monitoring time can be determined.

[0077]

[0078] Where D represents the predicted distance traveled by the single entity to be analyzed at the next monitoring time after the current monitoring time, and n D This indicates the number of monitoring time intervals included in the current combination of rainfall cells to be analyzed, which is also the number of sampling intervals between various monitoring times within the set duration. This represents the distance the rainfall cell traveled within the a-th interval.

[0079] Since the internal structure and meteorological characteristics of a rainfall cell are uniform, the predicted direction and distance of movement of that rainfall cell at the next monitoring time after the current monitoring time can be used as the predicted direction and distance of movement for all monitoring points within that rainfall cell at the current monitoring time. Therefore, the predicted direction and distance of movement for each monitoring point within each rainfall cell can be determined at the current monitoring time.

[0080] After obtaining the predicted movement direction and distance of each monitoring point in each rainfall cell at the current moment, it is possible to infer the location where the echo intensity of each monitoring point in the rainfall cell will affect other monitoring points at the next monitoring moment. Therefore, combining the relative position of each monitoring point with other monitoring points and the echo intensity, the degree of influence on each monitoring point at the next monitoring moment can be calculated. The underlying concept is as follows:

[0081] Taking any monitoring point at the current monitoring time as the target monitoring point, determine the direction angle from any monitoring point in any rainfall cell at the current monitoring time to the target monitoring point, and calculate the direction deviation between the direction angle and the predicted value of the movement direction of any monitoring point in any rainfall cell at the current monitoring time at the next monitoring time;

[0082] Based on the distance between the target monitoring point and any monitoring point in any rainfall cell at the current monitoring time, and the predicted movement distance of any monitoring point in any rainfall cell at the current monitoring time at the next monitoring time, the distance deviation between the target monitoring point and any monitoring point in any rainfall cell at the current monitoring time at the next monitoring time is determined.

[0083] The degree of influence is determined based on the directional deviation, the distance deviation, and the monitoring value of any monitoring point in a rainfall cell at the current monitoring time.

[0084] The formula for construction is:

[0085]

[0086] Among them, H i N represents the degree to which the i-th monitoring point is affected by other monitoring points at the current monitoring time and at the next monitoring time. i N represents the total number of remaining monitoring points within the monitoring area excluding the rainfall cell containing the i-th monitoring point (because rainfall cells have a certain structure, the influence here does not include monitoring points within the same rainfall cell). It's easy to understand that if the i-th monitoring point is not within a rainfall cell at the current monitoring time, then N... i This indicates the total number of other monitoring points within the monitoring area besides this one. This represents the relative movement direction vector composed of the azimuth and elevation angles of the line connecting the e-th monitoring point to the i-th monitoring point at the current monitoring time. It also represents the direction angle between the e-th monitoring point and the i-th monitoring point at the current monitoring time. This represents the predicted movement direction of the e-th monitoring point at the current monitoring time at the next monitoring time. express and The directional deviation between them, d i,e D represents the distance between the i-th monitoring point and the e-th monitoring point at the current monitoring time. e d represents the predicted distance traveled by the e-th monitoring point at the current monitoring time at the next monitoring time. i,e -D e |d represents the distance deviation between the i-th monitoring point and the e-th monitoring point at the next monitoring time after the current monitoring time. i,e -D e | indicates the calculation of d. i,e -D e The absolute value of Z, ε, is used to ensure that the fraction is meaningful; here, ε is defined as 0.01. eThis represents the monitoring value of the e-th monitoring point at the current monitoring time. The larger this value is, the greater its influence on the echo intensity of the i-th monitoring point. norm() represents the normalization function. In this embodiment, the normalization in each place can be implemented using the same method, such as linear normalization. This means that when the position, direction, and distance of the i-th monitoring point relative to the e-th monitoring point are similar to the rainfall movement trend of the e-th monitoring point, the i-th monitoring point is more significantly affected by the rainfall of the e-th monitoring point.

[0087] If the e-th monitoring point is not a monitoring point within a rainfall cell, then there is no movement trend, and it is considered that it will not affect the i-th monitoring point. In this case, the setting is... It is 0.

[0088] The prediction correction and early warning module 14 is used to determine the corrected prediction value of any monitoring point at the current monitoring time at the next monitoring time based on the predicted value of any monitoring point at the current monitoring time at the next monitoring time and the degree of influence, so as to complete the rainfall early warning.

[0089] Based on the echo intensity data (monitored values) of each monitoring point 15 minutes prior to the current time, the predicted value Z of that monitoring point at the next monitoring time is obtained using the ARIMA algorithm. i,pre Because direct prediction using the forecasting algorithm cannot predict rainfall movement at other monitoring points, the predicted value is corrected by considering the degree of impact on the current monitoring point.

[0090] Z i,pre '=Z i,pre ×(1+H i )

[0091] Among them, Z i,pre Z represents the corrected predicted echo intensity value of the i-th monitoring point at the next monitoring time after the current monitoring time. i,pre This represents the predicted echo intensity value of the i-th monitoring point at the next monitoring time after the current monitoring time.

[0092] By correcting the predicted values ​​based on the echo intensity of each monitoring point at the next monitoring time after the current monitoring time, an early warning of rainfall in the current monitoring area can be issued. In this embodiment, preferably, when 30dBZ≤Z i,pre A heavy rain warning is issued when Z < 40 dBZ; a heavy rain warning is issued when 40 dBZ ≤ Z. i,pre A rainstorm warning is issued when the Z value is less than 50 dBZ; when the Z value is less than 50 dBZ, a rainstorm warning is issued. i,pre When the value is ≥50dBZ, a severe convective weather warning will be issued, as hail, thunderstorms, etc. may occur.

[0093] This invention first determines the movement trajectory of individual rainfall cells to predict the movement direction and distance of each monitoring point at the next predicted time. By combining the deviations in movement direction and distance between the monitoring point under investigation and other monitoring points, as well as the current monitoring values ​​of other monitoring points, the influence of other monitoring points on the monitoring point under investigation at the next monitoring time can be determined. This influence level characterizes the impact of other monitoring points on the future weather changes of the monitoring point under investigation, which is something that traditional prediction methods cannot characterize under rapidly changing weather conditions. This allows for more accurate prediction of the monitoring values ​​of the monitoring point under investigation, improving the accuracy of rainfall monitoring and early warning by phased array radar in rapidly changing weather scenarios.

[0094] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A smart rainfall monitoring and early warning system based on phased array radar, characterized in that, The system includes: The historical data acquisition module is used to acquire the monitoring values ​​of each monitoring point in the monitoring area by the phased array radar within a set time period, and to determine the rainfall cells at each monitoring moment; The rainfall cell trajectory determination module is used to take any rainfall cell at the current monitoring time as the target cell, determine the proximity of the target cell to each rainfall cell at the previous monitoring time, and determine the movement trajectory of the target cell at the set time period based on the proximity of the positions. The module for determining the degree of influence of the predicted value is used to determine the predicted value of the movement direction and the predicted value of the movement distance of each monitoring point in the target unit at the next monitoring time based on the movement trajectory, thereby determining the predicted value of the movement direction and the predicted value of the movement distance of the monitoring points in each rainfall unit at the current monitoring time, and further determining the degree of influence of any monitoring point on other monitoring points at the current monitoring time. The prediction correction and early warning module is used to determine the corrected prediction value of any monitoring point at the current monitoring time at the next monitoring time based on the predicted value of any monitoring point at the current monitoring time at the next monitoring time and the degree of impact, and to complete the rainfall early warning.

2. The intelligent rainfall monitoring and early warning system based on phased array radar according to claim 1, characterized in that, The determination of rainfall cells at each monitoring time includes: The monitoring points in the monitoring area whose monitored values ​​are greater than a preset monitoring threshold are clustered, and the clusters with an area greater than a preset area threshold are taken as the rainfall units.

3. The intelligent rainfall monitoring and early warning system based on phased array radar according to claim 1, characterized in that, Determining the proximity of the target cell to the location of each rainfall cell at the previous monitoring time includes: A predetermined number of edge monitoring points that are closest to the target rainfall cell at the previous monitoring time are continuously selected from the edge monitoring points of the target cell to be judged to form a target edge segment. The corresponding edge segment on the target rainfall cell at the previous monitoring time is determined. The proximity of each rainfall cell to the target cell at the previous monitoring time is determined based on the average distance between the target edge segment and the corresponding edge monitoring points in the corresponding edge segment.

4. The intelligent rainfall monitoring and early warning system based on phased array radar according to claim 3, characterized in that, Determining the corresponding edge segment on the rainfall cell to be judged at the previous monitoring time that corresponds to the target edge segment includes: A predetermined number of consecutive edge monitoring points are randomly selected from the edge monitoring points of the rainfall cell to be judged at the previous monitoring time to form the edge segment to be judged. The distance value between the target edge segment and the edge segment to be judged is determined according to the average distance between the target edge segment and the corresponding edge monitoring point in the edge segment to be judged. The edge segment to be judged corresponding to the maximum distance value is taken as the corresponding edge segment in the rainfall cell to be judged at the previous monitoring time.

5. The intelligent rainfall monitoring and early warning system based on phased array radar according to claim 1, characterized in that, Determining the movement trajectory of the target unit within the set time period based on the proximity of its location includes: The target cell is taken as the cell to be analyzed. At the previous monitoring time, a set number of rainfall cells that are closest to the target cell in location are selected as the undetermined rainfall cells. The undetermined rainfall cells at the previous time are each selected as new cells to be analyzed. At the monitoring time before the previous monitoring time, a set number of rainfall cells that are closest to the new cells to be analyzed are selected as new undetermined rainfall cells. The process of determining the undetermined rainfall cells corresponding to the target cell is repeated until all monitoring times within the set time period are traversed. Each combination of the target cell and the corresponding undetermined rainfall cells at each monitoring time is determined. The angular change and position change of the rainfall cells in any two adjacent monitoring times in the current cell combination are determined. The cell combination with the smallest angular change and position change is taken as the movement trajectory.

6. The intelligent rainfall monitoring and early warning system based on phased array radar according to claim 1, characterized in that, The method for determining the predicted movement direction and predicted movement distance of each monitoring point in the target unit at the next monitoring time based on the movement trajectory includes: Based on the angle change of the target unit at any two adjacent monitoring times in the movement trajectory, predict the predicted angle change of the target unit at the next monitoring time relative to the current monitoring time. Determine the predicted value of the movement direction based on the movement direction of the target unit at the current monitoring time compared to the previous monitoring time and the predicted angle change. Determine the predicted value of the movement distance based on the average movement distance of the target unit at any two adjacent monitoring times in the movement trajectory.

7. The intelligent rainfall monitoring and early warning system based on phased array radar according to claim 6, characterized in that, The step of predicting the change in the angle of the target unit relative to the current monitoring time at the next monitoring time based on the change in angle of the target unit at any two adjacent monitoring times in the movement trajectory includes: The angle change of the target unit at any two adjacent monitoring times in the movement trajectory is arranged into an angle change sequence according to time, and curve fitting is performed to obtain an angle change fitting curve. The value of the angle change fitting curve at the next monitoring time is used as the predicted angle change.

8. The intelligent rainfall monitoring and early warning system based on phased array radar according to claim 1, characterized in that, The further determination of the degree to which any monitoring point is affected by other monitoring points at the current monitoring time includes: Taking any monitoring point at the current monitoring time as the target monitoring point, determine the direction angle from any monitoring point in any rainfall cell at the current monitoring time to the target monitoring point, and calculate the direction deviation between the direction angle and the predicted value of the movement direction of any monitoring point in any rainfall cell at the current monitoring time at the next monitoring time; Based on the distance between the target monitoring point and any monitoring point in any rainfall cell at the current monitoring time, and the predicted movement distance of any monitoring point in any rainfall cell at the current monitoring time at the next monitoring time, the distance deviation between the target monitoring point and any monitoring point in any rainfall cell at the current monitoring time at the next monitoring time is determined. The degree of influence is determined based on the directional deviation, the distance deviation, and the monitoring value of any monitoring point in a rainfall cell at the current monitoring time.

9. The intelligent rainfall monitoring and early warning system based on phased array radar according to claim 8, characterized in that, The degree of impact is as follows: Among them, H i N represents the degree to which the i-th monitoring point is affected by other monitoring points at the current monitoring time and at the next monitoring time. i This represents the total number of remaining monitoring points within the monitoring area, excluding the rainfall cell containing the i-th monitoring point. This represents the direction angle from the e-th monitoring point to the i-th monitoring point at the current monitoring time. This represents the predicted movement direction of the e-th monitoring point at the current monitoring time at the next monitoring time. express and The directional deviation between them, d i,e D represents the distance between the i-th monitoring point and the e-th monitoring point at the current monitoring time. e d represents the predicted distance traveled by the e-th monitoring point at the current monitoring time at the next monitoring time. i,e -D e |d represents the distance deviation between the i-th monitoring point and the e-th monitoring point at the next monitoring time after the current monitoring time. i,e -D e | indicates the calculation of d. i,e -D e The absolute value of Z, ε is used to ensure that the fraction is meaningful. e This represents the monitoring value of the e-th monitoring point at the current monitoring time, and norm() represents the normalization function.

10. The intelligent rainfall monitoring and early warning system based on phased array radar according to claim 1, characterized in that, Determining the predicted value of any monitoring point at the current monitoring time at the next monitoring time includes: Based on the monitoring values ​​of any monitoring point at the current monitoring time at each monitoring time within the set time period, the ARIMA algorithm is used to obtain the predicted value of any monitoring point at the current monitoring time at the next monitoring time.

Images