A wind shear early warning method and system based on radar technology

Abstract

The application provides a wind shear early warning method and system based on radar technology, and relates to the technical field of meteorological detection and aviation safety guarantee, and the method comprises the following steps: deploying multiple X-band Doppler weather radars to form a detection network, covering the airport runway and the surrounding low-altitude airspace and performing periodic scanning to obtain original radar echo data; transmitting the original radar echo data to an edge computing node in real time, obtaining three-dimensional airflow data through point cloud reconstruction, filtering and noise reduction processing; performing spatial interpolation based on the three-dimensional airflow data to generate wind field data; calculating the wind speed gradient and the air vorticity characteristic parameter based on the wind field data, and identifying the region meeting the initial early warning condition as a potential wind shear region through mutation detection and threshold comparison. Through the collaborative calibration and intelligent fusion of multi-source radar data, the application realizes the whole-process accurate and real-time early warning from data acquisition, dynamic risk assessment to hierarchical early warning and airport linkage.

Description

Technical Field

[0001] This invention relates to the fields of meteorological detection technology and aviation safety assurance technology, and in particular to a wind shear early warning method and system based on radar technology. Background Technology

[0002] In the field of airport aviation safety, low-level wind shear is a major threat during takeoff and landing. Currently, Doppler weather radar-based early warning systems are the mainstream monitoring method, but their core data processing flow has inherent flaws, which may limit the accuracy and reliability of early warnings.

[0003] Existing methods suffer from the following main shortcomings at the data processing level: First, the multi-radar data fusion approach is simplistic and crude. The system typically performs post-processing superposition of alarms generated independently by each radar, i.e., hard fusion, failing to establish a unified spatiotemporal reference for collaborative calibration. Due to differences in radar installation locations, scanning sequences, and beam obstruction, the detection of the same physical event inherently suffers from temporal asynchrony and spatial perspective bias, potentially leading to false alarms, ambiguous positioning, or amplified deviations in the fusion results, and failing to effectively compensate for the detection blind spots of individual radars. Second, the signal processing algorithms lack specificity. General filtering and noise reduction algorithms struggle to distinguish and retain clear-sky wind shear when suppressing interference such as ground clutter. Weak meteorological signals associated with the edge of a microburst may cause critical signals to be filtered out, resulting in missed reports. At the same time, the algorithm's ability to identify specific environmental disturbances such as airport building bypasses is insufficient, which may misjudge such non-meteorological disturbances as wind shear, leading to false alarms. Finally, the extraction and assessment of situation features are too isolated and static. Existing methods rely on threshold judgments of instantaneous, isolated parameters (such as maximum wind speed difference), lacking comprehensive modeling and dynamic analysis of the three-dimensional spatial structure, complete intensity gradient distribution, movement vector, and evolution trend of wind shear events. This may result in coarse output warning information, which cannot provide accurate and predictive analysis of the impact range, movement path, and development stage for flight decisions. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a wind shear early warning method and system based on radar technology. By performing collaborative calibration and intelligent fusion of multi-source radar data, it can achieve accurate and real-time early warning throughout the entire process from data acquisition and dynamic risk assessment to graded early warning and airport linkage.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows: Firstly, a wind shear early warning method based on radar technology, the method comprising: Multiple X-band Doppler weather radars are deployed to form a detection network, covering the airport runway and surrounding low-altitude airspace and conducting periodic scans to acquire raw radar echo data. The raw radar echo data is transmitted to the edge computing node in real time, and three-dimensional spatial airflow data is obtained through point cloud reconstruction, filtering and noise reduction. Spatial interpolation is performed based on three-dimensional airflow data to generate wind field data; wind speed gradient and air vorticity characteristic parameters are calculated based on wind field data, and areas that meet the initial warning conditions are identified as potential wind shear areas by abrupt change detection and threshold comparison. Based on the distribution of potential wind shear areas and the spatial relationship of multiple radars in the detection network, the airspace covered by the potential wind shear areas is divided into multiple calibration sub-regions, and spatial calibration coefficients are calculated for each calibration sub-region. By integrating potential wind shear region data with spatial calibration coefficients, comprehensive wind shear situation data is obtained through partitioned weighted calibration. Based on the comprehensive wind shear situation data, a risk assessment is conducted using a preset set of decision rules combined with dynamically adjusted judgment thresholds to identify risk areas and their risk levels. Based on the aforementioned risk level and comprehensive wind shear situation data, graded early warning information is generated and disseminated through a low-latency communication link, simultaneously triggering the corresponding airport operation control command set.

[0006] Secondly, a wind shear early warning system based on radar technology includes: The acquisition module is used to deploy multiple X-band Doppler weather radars to form a detection network, covering the airport runway and surrounding low-altitude airspace and performing periodic scans to acquire raw radar echo data. The reconstruction module is used to transmit the raw radar echo data to the edge computing node in real time, and obtain three-dimensional spatial airflow data through point cloud reconstruction, filtering and noise reduction. The identification module is used to perform spatial interpolation based on three-dimensional airflow data to generate wind field data; calculate wind speed gradient and air vortex characteristic parameters based on wind field data, and identify areas that meet the initial warning conditions as potential wind shear areas by abrupt change detection and threshold comparison. The calibration module is used to divide the airspace covered by the potential wind shear area into multiple calibration sub-regions based on the distribution of the potential wind shear area and the spatial positional relationship of multiple radars in the detection network, and to calculate the spatial calibration coefficient for each calibration sub-region. The fusion module is used to fuse potential wind shear area data with spatial calibration coefficients, and obtain comprehensive wind shear situation data through partitioned weighted calibration. The assessment module is used to conduct risk assessment based on the comprehensive wind shear situation data, using a preset set of decision rules combined with dynamically adjusted judgment thresholds, to identify risk areas and their risk levels. The early warning module is used to generate graded early warning information based on the risk level and comprehensive wind shear situation data, and to release it through a low-latency communication link, simultaneously triggering the corresponding airport operation control command set.

[0007] Thirdly, a computing device, comprising: One or more processors; A storage device for storing one or more programs that, when executed by one or more processors, cause the one or more processors to implement the method.

[0008] Fourthly, a computer-readable storage medium storing a program that, when executed by a processor, implements the method.

[0009] The above-described solution of the present invention has at least the following beneficial effects: A detection network composed of multiple radars is deployed to cover the airport runway and surrounding low-altitude airspace and periodically scan to acquire raw radar echo data, providing a comprehensive and real-time foundational data source for all subsequent data processing stages. The raw radar echo data is transmitted in real-time to edge computing nodes, where point cloud reconstruction, filtering, and noise reduction are performed to obtain three-dimensional spatial airflow data. Invalid interference data is removed, and the data format is standardized to provide data support for wind field data generation. Wind field data is generated based on the three-dimensional spatial airflow data, calculating wind speed gradients and air vorticity characteristic parameters. Potential wind shear regions are identified through abrupt change detection and threshold comparison, improving the wind field characteristic characterization and screening effective target areas to provide a benchmark for subsequent zonal calibration. Finally, multiple calibration sub-regions are divided based on the distribution of potential wind shear regions and the spatial relationship between the multiple radars. The system calculates spatial calibration coefficients to quantify the consistency of multi-radar data in each sub-region, providing a targeted calibration basis for subsequent data fusion. It integrates potential wind shear area data with spatial calibration coefficients, and obtains comprehensive wind shear situation data through regional weighted calibration, taking into account the differences in data from each sub-region and integrating multi-dimensional wind field information to form standardized and complete situation data. Based on the comprehensive wind shear situation data, it conducts risk assessments using a preset decision rule set and dynamically adjusted judgment thresholds, standardizing the risk assessment logic, quantifying risk details, and determining risk areas and their risk levels. Based on the risk level and comprehensive wind shear situation data, it generates tiered early warning information, which is then disseminated through a low-latency communication link and simultaneously triggers the corresponding airport operation control command set, standardizing the early warning dissemination process and achieving efficient connection between early warning information and airport operation control. Attached Figure Description

[0010] Figure 1 This is a schematic flowchart of a wind shear early warning method based on radar technology provided by an embodiment of the present invention.

[0011] Figure 2This is a schematic diagram of a wind shear early warning system based on radar technology provided in an embodiment of the present invention. Detailed Implementation

[0012] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0013] like Figure 1 As shown, an embodiment of the present invention proposes a wind shear early warning method based on radar technology, the method comprising the following steps: Step 100: Deploy multiple X-band Doppler weather radars to form a detection network, covering the airport runway and surrounding low-altitude airspace and periodically scanning to obtain raw radar echo data. Step 200: The raw radar echo data is transmitted to the edge computing node in real time. Through point cloud reconstruction, filtering and noise reduction processing, three-dimensional spatial airflow data is obtained. Step 300: Spatial interpolation is performed based on three-dimensional airflow data to generate wind field data; wind speed gradient and air vorticity characteristic parameters are calculated based on wind field data, and areas that meet the initial warning conditions are identified as potential wind shear areas by abrupt change detection and threshold comparison. Step 400: Based on the distribution of potential wind shear areas and the spatial positional relationship of multiple radars in the detection network, the airspace covered by the potential wind shear areas is divided into multiple calibration sub-regions, and spatial calibration coefficients are calculated for each calibration sub-region. Step 500: Integrate potential wind shear area data with spatial calibration coefficients, and obtain comprehensive wind shear situation data through partitioned weighted calibration; Step 600: Based on the comprehensive wind shear situation data, a risk assessment is conducted using a preset set of decision rules combined with dynamically adjusted judgment thresholds to identify risk areas and their risk levels. Step 700: Generate graded early warning information based on the risk level and comprehensive wind shear situation data, and release it through a low-latency communication link, simultaneously triggering the corresponding airport operation control instruction set.

[0014] In this embodiment of the invention, multiple X-band Doppler weather radars are networked and deployed to achieve comprehensive coverage of the airport runway and surrounding low-altitude airspace. Periodic scanning ensures the continuity and timeliness of the raw radar echo data, providing a basic data source for subsequent data processing. Edge computing nodes enable real-time and rapid processing of the raw radar echo data, reducing latency. Point cloud reconstruction transforms the two-dimensional echoes into three-dimensional airflow data, filtering and noise reduction eliminates clutter, improving the purity of the airflow data and reducing interference with subsequent wind field generation. Spatial interpolation compensates for the limitations of radar detection resolution, making wind field data coverage more uniform. Calculation of wind speed gradient and air vortex characteristic parameters captures airflow dynamics. Sudden change detection and threshold comparison quickly screen initial warning areas, ensuring the accuracy of subsequent calibration and evaluation. Targeted approach: Based on the distribution of potential wind shear areas and the locations of multiple radars, calibration sub-regions are divided, and a dedicated spatial calibration coefficient is set for each sub-region to provide accurate basis for subsequent data fusion and solve the spatial bias of multi-radar data; partitioned weighted calibration and fusion of potential wind shear data and spatial calibration coefficients leverage the advantages of multi-radar networking, weaken the impact of data bias, and integrate to form unified comprehensive wind shear situation data; preset decision rule set provides standard basis for risk assessment, dynamically judges thresholds to adapt to real-time airflow changes, achieves reasonable division of risk areas and levels, and ensures reliable assessment; graded early warning information facilitates accurate response by relevant departments, low-latency communication links ensure rapid transmission of early warnings, and simultaneously triggers airport operation control commands, improving the efficiency of wind shear handling.

[0015] In a preferred embodiment of the present invention, step 100 above, which involves deploying multiple X-band Doppler weather radars to form a detection network, covering the airport runway and surrounding low-altitude airspace and performing periodic scans to acquire raw radar echo data, includes: Step 101: Based on the airport runway configuration and the distribution of surrounding terrain obstacles, determine the deployment locations and number of radar stations to form a detection network with cross-coverage capability for the runway ends, climb corridor, and approach corridor. Specifically, this includes: First, conducting a comprehensive survey of the airport runway configuration, using on-site measurements to determine the runway's direction, length, and the specific coordinates of both ends. Simultaneously, thoroughly investigating the distribution of terrain obstacles around the runway, recording the height and distribution range of each type of obstacle. Then, combining the beam propagation characteristics of X-band Doppler weather radar, by statistically analyzing the correspondence between obstacle height and radar beam elevation angle, calculating the range and coordinates of the radar detection obstruction areas that various obstacles may form. Based on the above survey and calculation results, further planning and calculation of the radar station deployment locations and number are carried out, starting with statistical analysis. The total area of ​​the critical flight airspace formed by the runway end, climb corridor, and approach corridor, combined with the actual detection radius of a single X-band Doppler weather radar in low-altitude airspace, and after deducting the area of ​​the obstruction area caused by terrain obstacles, is used to calculate the minimum number of radars required. Then, based on the geometric distribution of the critical flight airspace, the optimal deployment coordinates of each radar station are calculated. This detection network adopts a distributed internal structure, with each radar acting as an independent detection node. It establishes a two-way communication connection with the core control node through a high-speed fiber optic link. Each radar has signal acquisition and preliminary preprocessing functions. The core control node is responsible for the unified management of the operation status of all detection nodes, realizing the synchronous acquisition of signals from each radar and the preliminary aggregation of data. At the same time, it reserves data interfaces with subsequent edge computing nodes to ensure that the raw radar echo data can be transmitted efficiently.

[0016] After the detection network was established, on-site testing was conducted for training and debugging. Typical scenarios under different time periods and weather conditions at the airport were selected for testing. Detection data from each radar was collected, and the differences between the actual terrain obstacle distribution and the radar detection obstruction area were compared. The parameter configuration of the radar detection nodes was adjusted, and the background signal characteristics in key flight airspace without weather interference were statistically analyzed as benchmark data for subsequent quality control. Through multiple rounds of on-site testing and training, the collaborative operation accuracy of the detection network was optimized, ensuring that the detection accuracy of the detection network reached the preset standard. This ensured that the deployed multiple X-band Doppler weather radars could form cross-coverage capability for the runway tip, climb corridor, and approach corridor, so that each low-altitude airspace in the above-mentioned key flight areas could be detected simultaneously by at least two radars. This effectively compensated for the detection blind spots caused by terrain obstruction of a single radar, laying the foundation for solving the problem of spatial perspective deviation of multiple radars in the future. At the same time, it ensured that the detection network could fully cover the key airspace during the aircraft takeoff and landing phases, and specifically avoid the safety threats caused by low-altitude wind shear.

[0017] Step 102: Based on the deployment location and number of the detection network, set scanning modes and parameters for each radar in the detection network. These parameters include the scanning elevation sequence, azimuth range, and scanning period to perform periodic volumetric scanning of the low-altitude airspace. Specifically, this includes: based on the deployment location and number of radar sites, and considering the overall coverage requirements of the detection network and the deployment environment of each radar site, setting a unique scanning mode and specific scanning parameters for each radar in the detection network. The scanning parameters specifically include the scanning elevation sequence, azimuth range, and scanning period. The specific calculation process for each parameter is as follows: The scanning elevation sequence is set by combining the key altitude range for aircraft takeoff, climb, and landing approaches, statistically analyzing the minimum and maximum values ​​within this altitude range, and combining this with the radar beamwidth to calculate the required starting angle, ending angle, and interval angle between adjacent elevation angles, forming a complete scanning elevation sequence. To ensure comprehensive coverage of airflow signals at key altitude levels, the azimuth range is set by first determining the boundary coordinates of the detection area for each radar unit, then calculating the starting and ending azimuth values ​​for that area using coordinate conversion. Simultaneously, the overlapping areas of azimuth coverage for adjacent radars are calculated to ensure effective connection between their scanning azimuth ranges and prevent detection gaps. The scanning cycle is set by first statistically analyzing the typical rate of change of low-altitude airflow around the airport, then combining this with the maximum processing load of subsequent data processing stages to calculate a scanning cycle that can capture dynamic airflow changes without causing data accumulation. This balances detection timeliness and data processing load, ensuring continuous periodic volumetric scanning of the low-altitude airspace. This allows the radar to continuously capture dynamic changes in airflow within the airspace, providing support for resolving the issue of time synchronization between multiple radars, and ensuring that radar detection data can promptly reflect the real-time status of low-altitude wind shear.

[0018] Step 103: Based on the scanning mode and parameters, control each radar to perform synchronous scanning and receive radar echo signals from various range and azimuth angles. Specifically, this includes: based on the scanning mode and various scanning parameters, using preset synchronization control commands, the preset synchronization control commands are combined with the scanning mode and various scanning parameters. First, a unified reference start sequence for synchronous scanning of each radar is determined. Combined with parameters such as scanning period and azimuth angle range, the standard time node for scanning start is determined. Then, time deviation data of each radar system collected during the construction of the detection network and the initial debugging are retrieved. The time deviation fluctuation range of each radar under different operating conditions is statistically analyzed to determine the calculation logic and adaptation criteria for the synchronization compensation. Subsequently, the reference start sequence, the synchronization compensation calculation logic, and the scanning parameter adaptation requirements of each radar are integrated to form a complete synchronization control command, which is preset to the core control node of the detection network to ensure that the command can accurately match the scanning mode of each radar and can dynamically adjust the compensation based on the subsequent real-time time deviation data. The synchronization compensation is used to control the synchronous scanning actions of each radar in the detection network. The specific calculation process for synchronization is as follows: First, the system time of each radar is collected, and the time deviation value between each radar is counted. The synchronization compensation is calculated based on the time deviation value and integrated into the synchronization control command to ensure that all radars can perform scanning operations in the same time dimension. This effectively avoids the time deviation of detection data caused by the asynchronous scanning sequence of each radar, thereby solving the time asynchrony problem in multi-radar detection. During the radar synchronous scanning process, radar echo signals from each radar range database and different azimuth angles are received in real time. At the same time, the actual detection distance corresponding to each range database is calculated. Combined with the azimuth angle parameters, the airspace coordinates corresponding to each radar echo signal are calculated to ensure that the signal characteristics of different distances and azimuths in the detection airspace can be completely captured without missing any signal information that may be related to low-altitude wind shear. This provides complete and continuous raw signal support for subsequent data processing and ensures the comprehensiveness of subsequent signal processing and feature extraction.

[0019] Step 104: Perform analog-to-digital conversion and preliminary quality control on the radar echo signal to remove non-meteorological echoes such as ground clutter, and obtain raw radar echo data containing radial velocity, spectral width, and reflectivity. Specifically, this includes: based on the received radar echo signal, firstly, performing analog-to-digital conversion processing on the acquired analog radar echo signal. The specific calculation process is as follows: firstly, obtain the peak and valley amplitudes of the analog radar echo signal through signal acquisition equipment, calculate the complete amplitude range of the analog radar echo signal, then determine the allowable error range for subsequent digital signal processing, and determine the accuracy level of digital signal processing based on this error range. The required sampling rate and quantization bits are calculated by considering the amplitude range and accuracy level. The sampling rate calculation must meet the Nyquist sampling criterion, ensuring that the sampling rate is not less than twice the highest frequency of the analog radar echo signal. The quantization bits are calculated by working backward from the amplitude range and accuracy requirements. The required quantization level is obtained by calculating the ratio of the amplitude range to the accuracy level, and then the corresponding quantization bits are determined based on the quantization level. According to the calculated sampling rate and quantization bits, the analog radar echo signal is sampled and quantized point by point, converting the analog signal into a digital signal that is easy for subsequent data processing, reducing the difficulty of subsequent data processing and improving the convenience of data processing.

[0020] Subsequently, preliminary quality control operations were performed on the converted digital radar echo signals. The specific calculation process is as follows: First, typical ground clutter areas around the airport were selected, and ground clutter signals were continuously collected for multiple scanning cycles. The signal amplitude and frequency parameters of the ground clutter in each cycle were extracted. The collected amplitude and frequency parameters were statistically analyzed to calculate the maximum, minimum, and average values ​​of the ground clutter amplitude. 1.2 times the average amplitude was taken as the ground clutter amplitude threshold. At the same time, the fluctuation range of the ground clutter frequency was calculated, and the upper and lower frequency limits were determined as the ground clutter frequency thresholds, forming a complete ground clutter signal characteristic threshold. Then, the amplitude and frequency parameters of the converted digital radar echo signals were extracted one by one. Each set of extracted parameters was compared with the ground clutter amplitude threshold and frequency threshold. If the signal amplitude exceeded the amplitude threshold or the frequency was within the ground clutter frequency threshold range, it was judged as ground clutter and other non-meteorological interference echoes and was removed. This specifically addresses the difficulty of general filtering algorithms in distinguishing clutter from weak meteorological signals. To address the shortcomings of traditional radar echo signals, this study selected weak meteorological signal samples related to historical clear-sky wind shear and microburst edges. Radial velocity, spectral width, and reflectivity parameters were extracted from these samples. Statistical calculations were performed on each parameter to obtain its average value and standard deviation of fluctuation. The lower limit retention threshold for each parameter was set by subtracting one standard deviation from the average value, and the upper limit retention threshold was set by adding one standard deviation to the average value. These three thresholds were combined to form the retention threshold for weak but valid signals. While eliminating invalid interference signals, the radial velocity, spectral width, and reflectivity parameters of the digital radar echo signals were checked one by one. Valid radar echo signals containing radial velocity, spectral width, and reflectivity, where all parameters are within their respective retention threshold ranges, were retained. These were then integrated to form the original radar echo data, providing a high-quality, high-purity raw data source for subsequent point cloud reconstruction, filtering and noise reduction, and wind field data generation. This ensures that subsequent data processing can capture weak meteorological signals related to low-level wind shear, reducing missed and false alarms.

[0021] In a preferred embodiment of the present invention, step 200 above involves transmitting the original radar echo data to an edge computing node in real time, and obtaining three-dimensional spatial airflow data through point cloud reconstruction, filtering, and noise reduction processing, including: Step 201 involves transmitting the raw radar echo data in real-time via a high-speed network to an edge computing node deployed locally at the airport, and decoding it into radar base data containing range, azimuth, elevation, radial velocity, and reflectivity. Specifically, this includes: first, retrieving the raw radar echo data, which contains effective meteorological parameters such as radial velocity, spectral width, and reflectivity; then, transmitting the raw radar echo data in real-time to the edge computing node via a high-speed fiber optic network deployed locally at the airport. Deploying the edge computing node locally at the airport effectively shortens the data transmission distance, avoids signal attenuation and delay caused by long-distance transmission, and ensures the real-time performance and integrity of the data transmission; after the data is transmitted to the edge computing node, the decoding operation is immediately initiated. During the decoding process, the encoding format of the raw radar echo data is first read, and this encoding format is consistent with that of the previous step. After step 104, the output format remains consistent with the analog-to-digital conversion and preliminary quality control. Then, the data is parsed step by step according to the encoding rules to extract the range, azimuth, elevation, radial velocity, and reflectivity parameters corresponding to each detection unit. The specific calculation process is as follows: first, the encoding bit length of the original radar echo data is counted to divide the encoding intervals corresponding to each parameter. Then, the actual measured values ​​of each parameter are restored through the numerical mapping of the encoding intervals. Abnormal data in the decoding process is eliminated. The parsed parameters are organized according to a unified data structure to form radar base data. This radar base data standardizes the data format, integrates the core parameters of airflow detection, and effectively solves the problem of inconsistent original data formats that are difficult to use directly for subsequent processing. This provides standardized basic data for subsequent coordinate transformation and point cloud reconstruction.

[0022] Step 202: Transform the range, azimuth, and elevation information of each measurement unit in the radar base data from the spherical coordinate system to the three-dimensional Cartesian coordinate system to obtain the three-dimensional rectangular coordinates of each measurement unit in space. Based on the three-dimensional rectangular coordinates of each measurement unit in space and their corresponding meteorological parameters, perform spatial position alignment and temporal interpolation fusion on the base data from multiple radars to generate an initial three-dimensional spatial airflow point cloud covering the low-altitude airspace. Specifically, this includes: First, transforming the spatial position parameters of each measurement unit in the radar base data, reading the range, azimuth, and elevation information corresponding to each measurement unit. This information is all measured data collected and decoded during radar synchronous scanning. Transform the parameters in the spherical coordinate system to the three-dimensional Cartesian coordinate system. The specific calculation process is as follows: First, obtain the range, azimuth, and elevation information of each measurement unit. The measured distance, azimuth, and elevation angles of each measurement unit are converted into their x, y, and z coordinates in a three-dimensional Cartesian coordinate system using trigonometric functions, ensuring precise spatial positioning of each unit. To further improve the accuracy of multi-radar data fusion and define the effective detection range of each measurement unit from a single radar, a conical lateral surface area algorithm is introduced. This algorithm calculates the lateral surface area of ​​the cone formed by the radar beam to define the effective detection area of ​​each measurement unit. Its core is to combine the radar beamwidth and detection range to determine the effective coverage area of ​​each measurement unit in three-dimensional space, providing a basis for spatial alignment and overlapping area determination during multi-radar data fusion, and avoiding data interference from invalid detection areas in the fusion results.

[0023] A specific application example of the conical lateral area algorithm is as follows: A measurement unit of an X-band Doppler weather radar in a detection network is selected. The decoded detection range of this measurement unit is 500 meters. Given that the radar's beamwidth is 1.5 degrees, the effective detection area of ​​this measurement unit is calculated using the conical lateral area algorithm. First, the apex angle of the cone is determined based on the beamwidth, using the 1.5-degree beamwidth as the apex angle. Then, the detection range of 500 meters is used as the generatrix length of the cone. Through actual measurement and statistical analysis, the correspondence between the equivalent radius of the radar beam and the generatrix length is obtained. The conical lateral area corresponding to this measurement unit is then calculated, defining the effective coverage area of ​​the measurement unit in three-dimensional space. Measurement data within this range is considered valid data, while data outside this range is considered invalid and discarded. Each measurement unit calculates its own effective detection area using the conical lateral area algorithm in this manner, ensuring that subsequent multi-radar data fusion only integrates data within the effective detection areas of each radar, improving the effectiveness of the fused data.

[0024] After completing coordinate transformation and defining the effective detection area, the baseline data from multiple radars are further fused. First, based on the deployment coordinates of each radar, the spatial position deviation between different radar measurement units is calculated. Then, combined with the timing data after synchronization control in step 103, the baseline data from multiple radars are spatially aligned. Simultaneously, based on the effective detection area defined by the conical lateral area algorithm for each measurement unit, effective fused data is selected. Interpolation calculations are used to compensate for spatial and timing gaps between different radar detections. Specifically, the spatial coverage overlap area of ​​the baseline data from multiple radars is statistically analyzed, and the overlap is addressed... The parameters of the measurement units within the region are calculated using a weighted average. The weighting is determined by the size of the conical lateral area of ​​the corresponding measurement unit of each radar. The larger the lateral area, the wider the effective detection range and the higher the weighting percentage. For the blank parts of the non-overlapping areas, interpolation is performed based on the parameters of adjacent measurement units to gradually integrate the base data from each radar. Finally, an initial three-dimensional spatial airflow point cloud covering the airport runway and surrounding low-altitude airspace is generated. This point cloud fully presents the spatial distribution of airflow in the low-altitude airspace, effectively compensating for the detection blind spots of a single radar and mitigating the adverse effects of spatial perspective deviation caused by multiple radars.

[0025] Step 203 involves performing point cloud gridding and statistical filtering on the initial three-dimensional spatial airflow point cloud to remove sparsely distributed isolated noise points and elevation clutter points with abnormal reflectivity, resulting in a filtered point cloud. Specifically, this includes: First, performing point cloud gridding on the initial three-dimensional spatial airflow point cloud. Based on the detection range and accuracy requirements of the low-altitude airspace, a uniform three-dimensional grid is divided. The spatial coordinate range of each grid is calculated. Each measurement unit in the initial three-dimensional spatial airflow point cloud is assigned to its corresponding grid based on its three-dimensional rectangular coordinates, ensuring that the point cloud data within each grid corresponds to a specific spatial region, improving the regularity of the point cloud data, and facilitating subsequent statistical filtering. After gridding, statistical filtering is immediately performed. Statistical analysis is conducted on the point cloud data within each grid, calculating the number of points in each grid, calculating the point cloud density, and setting a reasonable point cloud density threshold. The threshold is determined by statistically analyzing the point cloud density in areas free of clutter interference. Point clouds within grids with density below this threshold are identified as sparsely distributed isolated noise points and removed. Simultaneously, the reflectivity parameters of each measurement unit are read, and the average reflectivity and fluctuation range of the entire initial three-dimensional spatial airflow point cloud are statistically analyzed. Anomaly thresholds for reflectivity are calculated, and measurement units with reflectivity parameters exceeding these thresholds are identified as elevation clutter points with abnormal reflectivity and removed. During the filtering process, the reflectivity parameter range of weak meteorological signals is simultaneously analyzed to ensure that while removing noise and clutter, weak and effective point cloud data related to clear-sky wind shear and microburst edges are retained. After the above processing, the filtered point cloud is finally obtained. This point cloud effectively filters out invalid interference data, improves the purity of the point cloud data, and avoids interference from noise and clutter on subsequent noise reduction processing and airflow data generation.

[0026] Step 204 involves performing temporal moving average and spatial wavelet denoising on the filtered point cloud to suppress random noise and atmospheric background clutter, obtaining three-dimensional spatial airflow data. Specifically, this includes: First, performing temporal moving average processing on the filtered point cloud. Based on the radar scanning cycle set in step 102, determining the size of the sliding window. The duration of the sliding window matches the radar scanning cycle to ensure coverage of point cloud data within a complete scanning cycle. The filtered point cloud is then sorted according to time sequence, and the sliding window is moved sequentially. The average value of each measurement unit parameter within each sliding window is calculated, and this average value replaces the corresponding original parameter value within the window, gradually smoothing the temporal fluctuations of the airflow data and ensuring the accuracy of the airflow data. The temporal stability of the data is ensured. After completing the temporal moving average processing, spatial wavelet denoising is performed. The spatial distribution and parameter characteristics of the filtered point cloud are read, and wavelet decomposition is performed on the point cloud parameters in each grid to decompose the signal components of different frequencies. The frequency ranges corresponding to atmospheric background clutter and random noise are statistically analyzed, and the signal components within the frequency range are suppressed, retaining the frequency components of the corresponding effective airflow signals. Then, wavelet reconstruction is used to integrate the retained effective signal components to restore the pure airflow parameters. Through the coordinated processing of temporal moving average and spatial wavelet denoising, the interference of random noise and atmospheric background clutter is effectively suppressed, further purifying the airflow data, and finally obtaining three-dimensional spatial airflow data.

[0027] In this embodiment of the invention, the raw radar echo data is transmitted in real time to the airport's local edge computing node via a high-speed network, shortening the data transmission distance and latency, and ensuring the real-time performance of data processing. The raw radar echo data is decoded into radar base data containing range, azimuth, elevation, radial velocity, and reflectivity, standardizing the data format and extracting core data features to provide standardized and highly available base data for subsequent data processing steps such as coordinate transformation and point cloud reconstruction. The spherical coordinate system parameters of each measurement unit in the radar base data are transformed to a three-dimensional Cartesian coordinate system to determine the spatial positional relationship of each measurement unit, facilitating spatial comparison and integration of multiple radar data. Based on the three-dimensional rectangular coordinates and corresponding meteorological parameters of each measurement unit, spatial position alignment and temporal interpolation fusion are performed on multiple radar base data to compensate for spatial perspective deviations and temporal gaps between multiple radar data, generating a comprehensive coverage. The initial three-dimensional spatial airflow point cloud covering the low-altitude airspace fully presents the spatial distribution of airflow within the airspace. The initial three-dimensional spatial airflow point cloud is then gridded to standardize its spatial distribution and improve data regularity. Statistical filtering removes sparsely distributed isolated noise points and elevation clutter points with abnormal reflectivity, selectively filtering out invalid interference data while retaining valid airflow point cloud information, thus improving the purity of the point cloud data. Temporal moving average processing is applied to the filtered point cloud to smooth temporal fluctuations in the airflow data and ensure its temporal stability. Combined with spatial wavelet denoising, random noise and atmospheric background clutter are suppressed, further purifying the airflow data and reducing the impact of irrelevant interference on airflow feature extraction. Ultimately, high-quality three-dimensional spatial airflow data is obtained, providing reliable data support for subsequent wind field data generation and wind shear region identification.

[0028] In a preferred embodiment of the present invention, step 300 above involves spatial interpolation based on three-dimensional airflow data to generate wind field data; calculating wind speed gradient and air vortex characteristic parameters based on the wind field data; and identifying areas that meet the initial warning conditions as potential wind shear areas through abrupt change detection and threshold comparison. Step 301 involves performing three-dimensional spatial interpolation on the wind speed vectors in the three-dimensional airflow data to generate three-dimensional gridded wind field data with uniform spatial resolution. Specifically, this includes: first, retrieving the three-dimensional airflow data, which has already undergone filtering and noise reduction to retain effective airflow characteristics; then, performing three-dimensional spatial interpolation on the wind speed vectors in the data. Specifically, this involves first calculating the existing spatial sampling density of the three-dimensional airflow data, and then, based on the accuracy requirements for airport low-altitude airspace detection, determining the spatial spacing of the three-dimensional grid points. The airport runway and surrounding low-altitude airspace are then divided into three uniform grids according to this spacing. For each blank grid point, the nearest multiple valid airflow data points are selected, and the average wind speed vector of these data points is calculated. This average value is used as the wind speed vector of the blank grid point. This process is repeated to fill all blank grid points. For areas with excessively high spatial sampling density, the wind speed vectors of adjacent data points are weighted and averaged to normalize the data distribution. Through the above interpolation operations, a three-dimensional gridded wind field data with uniform spatial resolution is finally generated. This data effectively fills the spatial gaps in the original airflow data, normalizes the data distribution, and ensures that each subsequent grid point and its adjacent grid points have complete wind speed data.

[0029] Step 302: Based on the three-dimensional gridded wind field data, calculate the wind speed vector difference between each grid point and its adjacent grid points to obtain the wind speed gradient tensor describing the rate and direction of wind speed change, and calculate the vorticity scalar value describing the intensity of air rotation as a wind field characteristic parameter. Specifically, this includes: First, reading the wind speed vector information of each grid point in the three-dimensional gridded wind field data, which includes the wind speed magnitude and direction. Based on this, for each grid point, select its six adjacent grid points (front, back, left, right, up, and down), calculate the wind speed vector difference between this grid point and each of its adjacent grid points, and count the wind speed vector differences of all adjacent grid points. By integrating these parameters, a wind speed gradient tensor is formed that describes the rate and direction of wind speed change at each grid point, clearly presenting the spatial variation characteristics of wind speed at each grid point. Simultaneously, by combining the wind speed direction and magnitude at each grid point and its adjacent grid points, the wind speed rotation trend between adjacent grid points is statistically analyzed, and the average intensity of wind speed rotation around each grid point is calculated, yielding a vorticity scalar value describing the intensity of air rotation. The wind speed gradient tensor and the vorticity scalar value are used together as wind field characteristic parameters. Through the coordinated calculation of these two parameters, the representational dimensions of wind field characteristics are improved, avoiding the problem that a single parameter cannot fully reflect wind field anomalies, and providing comprehensive and reliable parameter support for subsequent wind field abrupt change detection.

[0030] Step 303: Based on the wind speed gradient tensor and vorticity scalar values, perform sliding window statistics in three-dimensional space to detect the abrupt changes in wind speed gradient magnitude and vorticity values ​​in space, and obtain the wind field abrupt change feature vector. Specifically, this includes: First, determining the size of the sliding window based on the grid spacing of the three-dimensional gridded wind field data. The size of the sliding window is set in conjunction with the conventional scale of low-altitude airspace wind shear to ensure that the window can completely cover the abnormal area of ​​small-scale wind shear, while taking into account both detection accuracy and processing efficiency. Then, moving the sliding window grid by grid within the three-dimensional grid, and statistically analyzing the wind speed gradient magnitude and vorticity values ​​of all grid points within each window, calculating the average wind speed gradient magnitude within the window. The system calculates the average, maximum, and fluctuation range of vorticity values ​​within a window. It then compares the statistical results of the current window with those of adjacent windows, calculating the difference between the wind speed gradient magnitude and vorticity values ​​between them. If the difference exceeds a preset reasonable fluctuation range, a wind field abrupt change is identified in that area. The location of the abrupt change and the corresponding wind speed gradient magnitude and vorticity value are recorded. All abrupt change locations and intensity information are integrated to obtain a wind field abrupt change feature vector. Through region-by-region statistical analysis using a sliding window, comprehensive detection of wind field abrupt changes in three-dimensional space is achieved, avoiding the omission of local abrupt change areas. This determines the specific location and intensity of abnormal wind field changes, ensuring the relevance of subsequent threshold comparisons.

[0031] Step 304: The wind field abrupt change feature vector is compared point by point with the preset initial warning threshold, and all spatial grid points that meet the initial warning threshold conditions are marked. Specifically, this includes: First, determining the preset initial warning threshold. The initial warning threshold is a criterion used to distinguish between normal and abnormal wind fields and to screen suspected wind shear grid points. It includes two warning thresholds: wind speed gradient modulus warning threshold and vorticity warning threshold. The two work together to ensure accurate and reliable judgment of wind field anomalies. The specific preset process is carried out step by step in combination with the airport's historical wind shear event data and airport flight safety requirements. First, all wind shear event records that have occurred at the airport in the past five years are retrieved, and the measured wind speed gradient modulus and vorticity values ​​corresponding to each event are extracted. Data: The extracted historical measured data are classified and statistically analyzed, and the specific values ​​of the two parameters in each event are recorded one by one; then the historical average value and fluctuation range of the two types of parameters are calculated, and the historical maximum value, minimum value and average fluctuation range of the wind speed gradient modulus are statistically obtained. At the same time, the historical maximum value, minimum value and average fluctuation range of the vorticity value are calculated. Based on the airport flight safety level requirements, the judgment standard is appropriately increased on the basis of the historical average value, taking into account the prevention and control needs of missed and false alarms, and the specific values ​​of the warning thresholds for the wind speed gradient modulus value and the vorticity value are determined. The initial warning threshold is preset to ensure that the threshold can accurately distinguish between normal wind fields and abnormal wind fields, without missing key wind field abnormal signals or misjudging normal wind field fluctuations as abnormal.

[0032] After the initial warning threshold is preset, the wind field abrupt change feature vector is compared point by point with the preset initial warning threshold. The wind speed gradient magnitude and vorticity value corresponding to each wind field abrupt change location are checked one by one. If both parameters reach or exceed the corresponding initial warning threshold, the spatial grid point corresponding to the abrupt change location is determined to be a suspected wind shear grid point, and the spatial coordinates of the grid point are recorded. If only one parameter reaches the threshold or neither parameter reaches the threshold, the grid point is determined to be a normal wind field grid point and is not included in subsequent processing. Through the above point-by-point comparison operation, all spatial grid points that meet the initial warning threshold conditions are marked to form a spatial grid point set, which provides an accurate grid point basis for the subsequent integrated identification of potential wind shear areas, and further improves the efficiency and reliability of subsequent area identification.

[0033] Step 305 involves performing three-dimensional spatial clustering and region growing on the spatial grid point set to form a continuous spatial region, which is then identified as a potential wind shear region. Specifically, this includes: First, performing three-dimensional spatial clustering on all grid points in the spatial grid point set. Based on the spatial coordinates of the grid points, calculating the spatial distance between any two grid points, and grouping grid points with a spatial distance less than a preset interval into the same cluster group. This process gradually divides all discrete grid points into multiple cluster groups, with each cluster group corresponding to a local suspected wind shear region. Subsequently, region growing is performed on each cluster group, using the core grid points within the cluster group as... Starting from the core grid point, each blank grid point around it is examined. If the blank grid point is close to the core grid point in space and its corresponding wind speed gradient tensor and vorticity scalar value are close to the initial warning threshold, the blank grid point is included in the cluster group, and the range of the cluster group is gradually expanded. At the same time, adjacent and spatially overlapping cluster groups are merged to avoid splitting the same potential wind shear region. Through the above-mentioned three-dimensional spatial clustering and region growth, discrete suspected wind shear grid points are integrated into a continuous spatial region. Finally, this continuous spatial region is identified as a potential wind shear region, and the complete spatial range of the potential wind shear region is determined.

[0034] In this embodiment of the invention, three-dimensional spatial interpolation is performed on the wind speed vector in the three-dimensional airflow data to generate three-dimensional gridded wind field data with uniform spatial resolution. This regularizes the wind field data distribution, fills data gaps, and provides regular and continuous basic data for subsequent feature parameter calculations. Based on the three-dimensional gridded wind field data, the wind speed vector difference between each grid point and its adjacent grid points is calculated to obtain the wind speed gradient tensor and vorticity scalar values. This accurately captures the wind field change rate, direction, and air rotation intensity, improves the wind field feature characterization, and provides reliable parameter support for abrupt change detection. The wind speed gradient tensor and vorticity scalar are statistically analyzed using a sliding window. The system accurately detects the location and intensity of sudden wind field changes, extracts feature vectors of these changes, and identifies core information about abnormal wind field variations, ensuring the relevance of subsequent threshold comparisons. It then compares the feature vectors of sudden wind field changes with initial warning thresholds point by point, marking sets of spatial grid points that meet the criteria. This accurately filters suspected wind shear grid point data, reducing interference from invalid data and improving the efficiency of subsequent region identification. Finally, it performs 3D clustering and region growing on the grid point sets to form continuous spatial regions and identify them as potential wind shear regions. By integrating discrete grid points, it determines the complete range of potential wind shear regions, providing a clear regional benchmark for subsequent calibration and evaluation.

[0035] In a preferred embodiment of the present invention, step 400, based on the distribution of potential wind shear regions and the spatial positional relationships of multiple radars in the detection network, divides the airspace covered by the potential wind shear regions into multiple calibration sub-regions, and calculates spatial calibration coefficients for each calibration sub-region, including: Step 401: Based on all radars participating in the detection network that are involved in detecting the potential wind shear region, obtain the three-dimensional spatial coordinates of their sites, and define each radar site as a generator based on the three-dimensional spatial coordinates. Specifically, this includes: First, calling the spatial range data of the potential wind shear region, and simultaneously retrieving the deployment location information of all radars in the detection network. This information is the three-dimensional spatial coordinate data measured and recorded during radar deployment in step 101. Filter out all radars whose detection range intersects with the potential wind shear region, i.e., radars participating in the detection of the potential wind shear region; then, retrieve the deployment location information of each of these participating radars one by one. The three-dimensional spatial coordinates of the radar stations are obtained, and the retrieved coordinate data is checked and calibrated to eliminate abnormal coordinate data caused by equipment errors. The actual measured values ​​of the x-axis, y-axis, and z-axis coordinates of each radar station are statistically analyzed to ensure the accuracy of the coordinate data. Based on this, each radar station participating in the detection is defined as a generator, and each generator corresponds to a unique three-dimensional spatial coordinate. The number of generators is consistent with the number of radars participating in the detection. Through the generators and their coordinates, a clear and accurate spatial benchmark is provided for the subsequent spatial area division, ensuring that the divided calibration sub-areas can accurately correspond to the detection range of multiple radars.

[0036] Step 402: Based on each generator and its three-dimensional spatial coordinates, perform spatial region partitioning, dividing the airspace covered by the potential wind shear region into multiple three-dimensional spatial regions. Each spatial location within a three-dimensional spatial region is closest to its corresponding generator. Each partitioned three-dimensional spatial region is defined as a calibration sub-region. Specifically, this includes: first, reading the complete spatial range of the potential wind shear region identified in step 305, determining the extreme values ​​of the x-axis, y-axis, and z-axis coordinates of the airspace, and determining the overall boundary of the partition; then, based on the three-dimensional spatial coordinates of each generator, performing spatial region partitioning on the airspace covered by the potential wind shear region. Specifically, the partitioning process involves selecting any spatial location point within the potential wind shear region and calculating... Calculate the spatial straight-line distance from the location point to each generator, count the distance values ​​between all generators and the location point, and assign the location point to the region corresponding to the nearest generator. Following the above method, calculate and classify all spatial location points within the potential wind shear region one by one, gradually dividing the entire potential wind shear region into multiple independent, non-overlapping, and completely covering three-dimensional spatial regions. Finally, define each of the divided three-dimensional spatial regions as a calibration sub-region, and each calibration sub-region corresponds to a unique generator, ensuring that all spatial locations within each calibration sub-region are closest to their corresponding generator. This division method enables each calibration sub-region to be highly compatible with the detection range of the corresponding radar, facilitating subsequent targeted calculation of spatial calibration coefficients.

[0037] Step 403: For each calibration sub-region, determine its corresponding generator radar; based on the spatial range of the calibration sub-region, filter out all radars whose detection ranges spatially intersect with that calibration sub-region. Specifically, this includes: First, for each calibration sub-region, according to the partitioning logic of step 402, determine its corresponding generator. The radar corresponding to this generator is the generator radar of that calibration sub-region, and the generator radar is the reference radar for the calibration calculation of that calibration sub-region; then, retrieve the spatial range data of the calibration sub-region to determine its x-axis, y-axis, and z-axis coordinate ranges, and simultaneously retrieve the effective detection range data of all radars participating in the detection of potential wind shear areas in the detection network. This data is the radar detection range calculated and recorded in step 101; compare each... The effective detection range of the radar and the spatial range of the current calibration sub-region are compared, and the spatial intersection area between them is calculated. If the intersection area is greater than a preset minimum effective area threshold, it is determined that the radar's detection range and the calibration sub-region have a spatial intersection, and the radar is included in the list of radars participating in the calibration calculation of the calibration sub-region. If the intersection area is less than or equal to the preset minimum effective area threshold, it is determined that the radar's detection range and the calibration sub-region have no effective intersection, and the radar is removed. Through the above filtering operation, all radars whose detection ranges have a spatial intersection with the calibration sub-region are retained, ensuring that all radars participating in the calibration calculation of the calibration sub-region can effectively detect the wind field characteristic parameters within the sub-region, reducing the interference of irrelevant radar data on the calibration calculation, and improving the accuracy of the calibration coefficient calculation.

[0038] Step 404: Obtain the detection values ​​of the generator radar and the selected radars for the wind field characteristic parameters within the calibration sub-region. The wind field characteristic parameters include wind speed and wind direction. Specifically, this includes: First, for each calibration sub-region, determine its corresponding generator radar and other participating radars selected in step 403, and compile them into a radar list for that calibration sub-region; then, retrieve the three-dimensional gridded wind field data generated in step 302, which contains the wind speed and wind direction parameters for each spatial grid point. Based on the spatial range coordinates of the current calibration sub-region, filter out the detection values ​​of the wind field characteristic parameters corresponding to all spatial grid points within that sub-region; extract the wind speed and wind direction detection values ​​for these spatial grid points for each radar in the radar list one by one, ensuring that the detection value of each radar accurately corresponds to the corresponding spatial grid point during the extraction process to avoid data misalignment; perform a preliminary check on the extracted detection values, remove invalid detection data with missing or abnormal values, and count the number of valid detection grid points for each radar within the calibration sub-region.

[0039] If the number of effective detection grid points is lower than the preset threshold, the radar's detection data is re-extracted to ensure data integrity. The preset threshold is the minimum number of effective detection grid points required to determine whether radar detection data meets the conditions for effective calibration calculation. The specific preset process is carried out step-by-step, considering radar detection performance, calibration sub-area spatial scale, and data calibration accuracy requirements. First, the rated detection accuracy and effective detection range of each participating radar are retrieved, and the percentage of effective detection grid points for each radar in a typical detection scenario is calculated. Then, combined with the total number of spatial grid points in the current calibration sub-area, the initial value of the minimum number of effective detection grid points required to ensure calibration accuracy is calculated. Next, historical data from similar calibration scenarios are retrieved, and the calibration result errors corresponding to different initial thresholds are statistically analyzed. Based on the acceptable error range, the initial threshold is fine-tuned. Finally, the rationality of the threshold is verified through multiple sets of measured data, eliminating problems such as insufficient calibration accuracy and data waste caused by excessively high or low thresholds, and the final preset threshold is determined.

[0040] Finally, the extracted and verified values ​​from the generator radar and other radars are categorized and organized according to radar type and spatial grid coordinates to form a set of wind field characteristic parameter detection values ​​for the calibration sub-region. This set contains the wind speed and wind direction detection data of each radar for each grid point in the calibration sub-region, providing reliable support for subsequent difference calculations.

[0041] Step 405: For each calibration sub-region, the wind field characteristic parameter detection values ​​of the generator radar are compared pairwise with the corresponding detection values ​​of other radars. The differences in wind speed and direction for each pair of detection values ​​are calculated to obtain the differences in wind speed and direction for each group. Specifically, this includes: First, for each calibration sub-region, retrieving the organized set of detection values ​​and determining the detection data of the generator radar and other participating radars, using the detection value of the generator radar as the baseline detection value to ensure the uniformity of the calibration calculation; then, the wind field characteristic parameter detection values ​​of the generator radar are compared pairwise with the corresponding detection values ​​of other radars. The process is carried out point by point in the spatial grid. Each grid point corresponds to a set of generator radar values ​​compared with those of other radars. For the comparison data of each grid point, the wind speed difference and wind direction difference are calculated separately. The wind speed difference is obtained by subtracting the wind speed detection value of the generator radar from the corresponding wind speed detection value of other radars and taking the absolute value. The wind direction difference is obtained by subtracting the wind direction detection value of the generator radar from the corresponding wind direction detection value of other radars and taking the absolute value. If the wind direction difference is greater than a preset angle threshold, the wind direction difference is corrected to ensure that the difference value is within a reasonable range. The preset angle threshold is used for... The angle standard for determining whether wind direction differences need correction and ensuring they remain within a reasonable range is determined through a pre-set process that integrates radar wind direction detection accuracy, wind shear calibration requirements, and historical detection data. First, the rated wind direction detection accuracy parameters of each participating radar are retrieved, and the wind direction detection error range of each radar under conventional detection scenarios is statistically analyzed. Combined with the accuracy requirements for wind direction difference representation under wind shear calibration, an initial angle threshold value is set. Then, data on the wind direction difference correction effect and calibration coefficient calculation accuracy corresponding to different angle thresholds in similar past calibration scenarios are retrieved to analyze the impact of the threshold setting on subsequent calibrations. The impact of the process is considered; subsequently, the initial threshold is fine-tuned based on a reasonable range of wind direction differences to avoid problems such as invalid correction due to excessively high thresholds or over-correction due to excessively low thresholds; finally, the rationality of the threshold is verified through multiple sets of measured wind direction detection data and calibration simulation scenarios to determine the final preset angle threshold; the wind speed and wind direction differences corresponding to each grid point are treated as a set of data, and the difference data of all comparison groups are recorded one by one to obtain the wind speed and wind direction differences of each group. These difference data fully present the deviation of multiple radar detection values, providing core data for subsequent deviation quantification and calibration coefficient calculation.

[0042] Step 406: Construct a two-dimensional difference vector for each group of wind speed and direction differences, and calculate the Euclidean distance of each two-dimensional difference vector to form a set of distance values. Specifically, this includes: First, for each group of wind speed and direction differences, the wind speed difference of the group is taken as the first component of the two-dimensional difference vector, and the wind direction difference of the group is taken as the second component of the two-dimensional difference vector. Two-dimensional difference vectors corresponding to each group of differences are constructed one by one to ensure that each two-dimensional difference vector can completely represent the deviation characteristics of a set of multi-radar detection values. Then, each two-dimensional difference vector is quantized, and the Euclidean distance of the two-dimensional difference vector is calculated. The specific calculation process is as follows: First, for each group of wind speed and direction differences, the wind speed difference of the group is taken as the first component of the two-dimensional difference vector, and the wind direction difference of the group is taken as the second component of the two-dimensional difference vector. The two components of the two-dimensional difference vector are squared, and the two squared values ​​are added together. The square root of the sum is then taken, which is the Euclidean distance corresponding to the two-dimensional difference vector. In this way, the two-dimensional deviation is transformed into a single distance value, realizing the quantification and unification of deviation data. The Euclidean distances corresponding to all two-dimensional difference vectors are calculated one by one. All the calculated Euclidean distances are sorted in order, and abnormal distance values ​​that occur during the calculation process are removed. Finally, a set of distance values ​​is formed. This set quantifies the multi-radar detection deviation into a single statistical and analyzable value, standardizes the data format, and provides a regular basic data for subsequent calculation of the root mean square error.

[0043] Step 407: Calculate the root mean square error (RMSE) of the distance value set, and normalize the reciprocal of the RMSE to a preset weight interval to obtain the spatial calibration coefficient characterizing the consistency of multi-radar data within the calibration sub-region. Specifically, this includes: First, retrieving the distance value set, counting the total number of distance values ​​in the set, and verifying the completeness of the distance values ​​to ensure no missing or abnormal values; then, calculating the RMSE of the distance value set. The specific calculation process is as follows: first, calculate the square of each distance value in the set, then add all the squares together, and divide the sum of squares by the total number of distance values ​​to obtain the average value. The square root of this average value is the statistical root mean square error (RMSE) corresponding to this set of distance values. This error effectively characterizes the overall level of multi-radar detection bias; the smaller the error, the better the consistency of the multi-radar data. Then, the reciprocal of this RMSE is taken to obtain the reciprocal error value. This value is positively correlated with the consistency of the multi-radar data; the larger the value, the better the data consistency. Finally, the reciprocal error is normalized to a preset weighted interval. The preset weighted interval is set based on the multi-radar data fusion requirements, and its specific preset process is carried out gradually based on the accuracy requirements of multi-radar data fusion, historical calibration experience, and radar detection capabilities. First, historical calibration records of multi-radar data fusion at the airport are retrieved. The actual distribution range of the reciprocal errors of different calibration sub-regions in history is extracted, and the maximum, minimum, and average distribution density of these reciprocals are calculated. Then, combined with the current fusion accuracy requirements for low-altitude airspace detection at the airport, a reasonable range for weight allocation is defined, balancing the accuracy of data consistency representation with the convenience of subsequent fusion calculations. The upper and lower limits of the weight interval are determined, typically preset to a range of 0 to 1, to ensure that the normalized calibration coefficients can intuitively reflect the consistency of multi-radar data and facilitate subsequent fusion processes based on calibration. The coefficients are assigned corresponding weights to complete the setting of the preset weight range. The specific normalization process is as follows: first, the maximum and minimum values ​​of the reciprocals of the errors of all calibration sub-regions are counted, the difference between the reciprocal of the error and the minimum value is calculated, and then divided by the difference between the maximum and minimum values ​​to obtain the normalized value, ensuring that the value is within the preset weight range. The normalized value is the spatial calibration coefficient that characterizes the consistency of multi-radar data in the calibration sub-region. Each calibration sub-region corresponds to a unique spatial calibration coefficient. This coefficient quantifies the consistency level of multi-radar data in the sub-region and provides a targeted calibration basis for subsequent multi-radar data fusion.

[0044] In this embodiment of the invention, the three-dimensional spatial coordinates of radar stations participating in the detection of potential wind shear areas are obtained, and each radar is defined as a generator, providing a spatial reference for subsequent spatial region division. Spatial regions are divided based on the generators and their three-dimensional coordinates, with each region being closest to its corresponding generator and defined as a calibration sub-region. This achieves regularization and targeted division of the calibration sub-regions, providing a regional basis for subsequent zonal calibration. The generator radars corresponding to each calibration sub-region are determined, and radars whose detection range intersects with the sub-region are selected. Invalid radar data is eliminated to ensure that the radars participating in the calibration are targeted and to reduce interference from irrelevant data. The wind speed, etc., of the generator radars and the selected radars within the calibration sub-regions are obtained. The system collects wind direction detection values ​​and wind field characteristic parameters from multiple radars within a sub-region, providing complete basic data for subsequent difference calculations. It then compares the detection values ​​of the generator radar with those of other radars, calculating wind speed and direction differences to accurately capture the details of deviations in the multi-radar detection data and determine the data deviation distribution. A two-dimensional difference vector is constructed, and Euclidean distance is calculated to quantify the multi-radar detection deviations into statistically significant distance values, forming a standardized set of distance values ​​for subsequent error analysis. Finally, the statistical root mean square error of the distance value set is calculated, and its reciprocal is normalized to obtain the spatial calibration coefficient. This quantifies the consistency of multi-radar data within the calibration sub-region, generating accurate zonal calibration criteria to support subsequent data fusion processing.

[0045] In a preferred embodiment of the present invention, step 500, which integrates potential wind shear region data and spatial calibration coefficients to obtain comprehensive wind shear situation data through partitioned weighted calibration, includes: Step 501: Obtain the spatial extent data of each potential wind shear region and the spatial calibration coefficients of each corresponding calibration sub-region. Specifically, this includes: First, retrieving the spatial extent data of all potential wind shear regions. This data includes the three-dimensional Cartesian coordinate system coordinate range, spatial coverage area, and core area location information of each potential wind shear region. Verify the spatial extent data of each potential wind shear region one by one, and eliminate abnormal coordinate data caused by region identification deviation to ensure that the data can accurately represent the actual spatial distribution of each potential wind shear region. To further improve the efficiency and accuracy of subsequent association and matching between calibration sub-regions and potential wind shear regions, and to avoid invalid spatial comparison operations, a spatial indexing algorithm is introduced here. This algorithm is an algorithm that quickly realizes spatial object search, association, and matching by establishing an index structure for spatial data, thereby improving the efficiency of spatial data processing. Its core is to reduce the comparison of irrelevant spatial data by dividing the spatial grid and establishing index mapping relationships, thus ensuring the accuracy and efficiency of spatial data association.

[0046] The specific implementation and application examples of the spatial indexing algorithm are as follows: First, the low-altitude airspace of the airport is divided into uniform three-dimensional grids according to a preset size. The side lengths of the x, y, and z axes of each grid are all set to 50 meters. A unique index ID is assigned to each three-dimensional grid, and the index IDs are encoded sequentially according to the spatial position of the grid to ensure that the index IDs of each grid are not repeated. Then, the three-dimensional coordinate range of each potential wind shear region is compared with the spatial range of the three-dimensional grid to determine all the three-dimensional grids covered by each potential wind shear region. The index IDs corresponding to these grids are recorded to form the index ID set of each potential wind shear region. At the same time, the spatial calibration system of each calibration sub-region calculated in step 407 is retrieved. The coefficient, normalized to a preset weight range, characterizes the consistency level of multi-radar data within the corresponding calibration sub-region. Similarly, the 3D coordinate range of each calibration sub-region is compared with the 3D grid to determine the 3D grid covered by each calibration sub-region and its corresponding index ID, forming an index ID label for each calibration sub-region. By comparing the set of index IDs of potential wind shear regions with the index ID labels of calibration sub-regions, calibration sub-regions with matching index IDs are quickly selected, completing the initial association between the two. Further verification is made to ensure that the spatial range of the calibration sub-region is completely within the corresponding potential wind shear region, ensuring the accuracy of the association and matching. This example does not require complex equipment support and can be directly implemented by combining existing spatial coordinate data.

[0047] After completing the association matching assisted by the spatial indexing algorithm, the calibration sub-regions covered by each potential wind shear region and their corresponding spatial calibration coefficients are further determined according to the division of potential wind shear regions. Finally, the retrieved spatial range data and spatial calibration coefficients are classified and organized, archived one by one according to the potential wind shear region number, and the number of calibration sub-regions corresponding to each potential wind shear region and the calibration coefficient values ​​of each sub-region are counted. The integrity and correlation of the data are checked to avoid problems such as data misalignment and missing data, laying a solid foundation for subsequent partitioned weighted average fusion. At the same time, the application of the spatial indexing algorithm and accurate data retrieval and organization improve the efficiency of spatial data association, providing a prerequisite guarantee for optimizing the multi-radar data fusion effect.

[0048] Step 502: For each potential wind shear region, the spatial calibration coefficients of all calibration sub-regions within its coverage area are used as the fusion weights for the wind field characteristic parameter data of each sub-region. The wind speed, wind direction, and air vorticity parameters detected by multiple radars within the calibration sub-regions are weighted and averaged to obtain the fused wind field characteristics of the region. Specifically, this includes: First, for each potential wind shear region, based on the association matching results of step 501, all calibration sub-regions within the coverage area of ​​the region are determined. The spatial calibration coefficient of each calibration sub-region is directly used as the fusion weight for the wind field characteristic parameter data within the corresponding sub-region. The larger the calibration coefficient value, the better the consistency of the multi-radar data within the corresponding sub-region, and the higher the weight ratio of its detected data in the fusion process, thereby achieving a match between the fusion weight and data consistency. Then, the wind field characteristic parameter detection values ​​of each radar within each calibration sub-region collected in step 404 are retrieved, focusing on extracting the three core parameters: wind speed, wind direction, and air vorticity. Detection data for the corresponding parameters are extracted one by one according to the calibration sub-region, ensuring that the detection data within each calibration sub-region comes from the valid radars selected in step 403. For each calibration sub-region, using the spatial calibration coefficient of that sub-region as the weight, a weighted average fusion is performed on the detected values ​​of the same wind field characteristic parameter from multiple radars. The fusion calculation is performed separately for each wind speed, wind direction, and air vortex parameter. Specifically, the calculation process involves first counting all valid detected values ​​of a certain parameter within the sub-region, multiplying each detected value by the corresponding radar's calibration coefficient for that sub-region to obtain a weighted value for each detected value, then summing all weighted values ​​to obtain a weighted sum. Simultaneously, the sum of the calibration coefficients corresponding to the number of radars participating in the fusion is calculated, and the weighted sum is divided by... The fused value of the parameter in the sub-region is obtained by summing the calibration coefficients. The weighted average fusion of the three wind field characteristic parameters in each calibration sub-region is completed in the above manner. Then, the parameter fusion values ​​of all calibration sub-regions in the potential wind shear region are integrated, and the overall average value of the three parameters in the region is calculated. Finally, the fused wind field characteristics of the potential wind shear region are obtained. The fused wind field characteristics can take into account the consistency differences of multi-radar data in each sub-region, effectively weaken the adverse effects of multi-radar spatial perspective deviation, and improve the accuracy and reliability of wind field characteristic parameter fusion.

[0049] Step 503: For each potential wind shear region, integrate its fused wind field characteristics with the region's spatial extent, location, and timestamp to generate comprehensive wind shear situation data for that region. Specifically, this includes: First, for each potential wind shear region, retrieve the fused wind field characteristics, which include the region's average wind speed, prevailing wind direction, and average air vorticity parameters. Verify the fused value of each parameter one by one, eliminating abnormal values ​​caused by fusion calculation deviations to ensure the fused wind field characteristics can represent the actual wind field state of the region. Then, retrieve the spatial extent data and core location information of the potential wind shear region to determine the region's three-dimensional spatial coordinate boundaries, spatial coverage, and the specific location of the core area. Perform preliminary integration of this spatial information with the fused wind field characteristics. Finally, retrieve the timestamp of the radar synchronous scan. The timestamp is consistent with the time-series data of the radar synchronous scan in step 103, and can characterize the detection time of the wind field characteristics of the potential wind shear area. The timestamp information is added to the integrated data to realize the correlation between wind field characteristics and time dimension. Finally, the four core information of fused wind field characteristics, spatial range, location and timestamp are fully integrated, archived one by one according to the preset data structure, the data format is standardized, the correspondence of each piece of information is determined, and a unique comprehensive wind shear situation data is generated for each potential wind shear area. This data fully covers the wind field characteristics, spatial distribution and time-series information of the potential wind shear area, improves the characterization dimension of wind shear situation, and provides comprehensive and standardized multi-dimensional data basis for the accurate release of subsequent wind shear warnings and flight decisions, further optimizing the completeness and standardization of warning information.

[0050] In this embodiment of the invention, spatial range data of each potential wind shear region and spatial calibration coefficients of corresponding calibration sub-regions are acquired, and core basic data are integrated to provide complete support for subsequent zonal weighted calibration and data fusion. Using the spatial calibration coefficients of the calibration sub-regions as fusion weights, the wind speed, wind direction, and air vortex parameters of multiple radars are weighted and averaged to ensure consistency of data from multiple radars in each sub-region and improve the reliability of the fused wind field characteristics. The fused wind field characteristics, spatial range, location, and timestamps of potential wind shear regions are integrated to standardize the data format and form complete comprehensive wind shear situation data, providing a regular multi-dimensional data basis for subsequent early warning applications.

[0051] In a preferred embodiment of the present invention, step 600, based on the comprehensive wind shear situation data, utilizes a preset decision rule set combined with dynamically adjusted judgment thresholds to perform risk assessment, confirming the risk area and its risk level, including: Step 601: Analyze the comprehensive wind shear situation data for each potential wind shear region, extracting its wind shear intensity index, impact range parameter, spatial persistence parameter, and movement trend. Specifically, this includes: First, retrieving the comprehensive wind shear situation data for all potential wind shear regions. This data integrates core information such as fused wind field characteristics, spatial range, location, and timestamps. The situation data for each region is verified one by one, eliminating invalid data such as missing values ​​and abnormal coordinates caused by data integration deviations to ensure data completeness and accuracy. Then, the comprehensive wind shear situation data for each potential wind shear region is analyzed one by one, focusing on extracting four core evaluation parameters: wind shear intensity index, impact range parameter, spatial persistence parameter, and movement trend. Among these, the wind shear intensity index is extracted based on wind speed and air vorticity parameters from the fused wind field characteristics, and statistics for the region are statistically analyzed. The average wind speed gradient within the region is used as an indicator of wind shear intensity. The influence range parameter is extracted based on the spatial extent data of the region, statistically analyzing the three-dimensional spatial coverage area and the radius of the core area. The spatial persistence parameter is extracted by combining timestamp information, statistically analyzing the duration of continuous occurrence of the potential wind shear region, and simultaneously analyzing the fluctuation amplitude of wind field characteristic parameters in different time series, thus obtaining the spatial persistence parameter. The movement trend is determined by comparing the spatial coordinates of the region in different time series, calculating the spatial distance between the core locations of the region in adjacent time series, statistically analyzing the distance changes in value and direction, and thus determining its movement trend. Finally, the four extracted core parameters are archived one by one according to the potential wind shear region number, verifying the correlation between the parameters and the corresponding regions to avoid parameter misalignment issues, laying a solid foundation for subsequent rule matching.

[0052] Step 602 involves matching the wind shear intensity index, impact range parameter, spatial persistence parameter, and movement trend with the logical conditions defined in a preset decision rule set to determine whether to trigger the primary alarm flags for various wind shear events. Specifically, this includes: first, retrieving the preset decision rule set. The preset decision rule set is a set of logical rules pre-set to determine different types of wind shear events and trigger primary alarm flags, based on the actual characteristics of wind shear in the airport's low-altitude airspace, historical alarm data, and flight safety requirements. Its core function is to provide judgment criteria for parameter matching and standardize the primary alarm triggering logic. Specifically, it includes judgment logic conditions for different types of wind shear events. Each type of wind shear event corresponds to a clear parameter threshold range and logical association conditions, covering common event types such as mild, moderate, severe wind shear, and microbursts.

[0053] The specific pre-setting process of this set of decision rules is carried out step by step in conjunction with the actual needs of airport wind shear prevention and control. First, all wind shear event alarm records and corresponding measured data that have occurred at the airport in the past five years are retrieved. This data is then classified and organized according to differences in wind shear intensity, impact range, spatial persistence, and movement trend, classifying wind shear events into common types such as mild, moderate, severe wind shear, and micro-downbursts. Subsequently, the measured data of each type of wind shear event are classified and statistically analyzed, recording the specific values ​​of wind shear intensity indicators, impact range parameters, spatial persistence parameters, and movement trends when each type of event occurs, and calculating the average value and fluctuation amplitude of each parameter. Based on the degree and critical values, and in conjunction with airport flight safety level requirements, the threshold ranges of various parameters corresponding to each type of wind shear event are determined. Then, according to the formation mechanism and actual hazard level of various wind shear events, the logical association conditions of the four core parameters are determined, the order of parameter matching and judgment criteria are defined, and the parameter thresholds and logical conditions of various events are integrated to form an initial decision rule set. Finally, the initial decision rule set is verified and debugged through multiple sets of measured data and simulated wind shear scenarios. Rules with unreasonable matching logic and large deviations in threshold setting are eliminated, and the parameter thresholds and logical association conditions are adjusted to ensure the matching accuracy of the rule set. Finally, the setting of the preset decision rule set is completed.

[0054] After the preset decision rule set is retrieved and confirmed, for each potential wind shear area, the wind shear intensity index, impact range parameter, spatial persistence parameter, and movement trend are matched one by one with the logical conditions of various wind shear events defined in the decision rule set. The matching process is carried out one by one according to the event type in the rule set. First, it is compared whether the wind shear intensity index meets the intensity threshold of the corresponding event. Then, it is compared whether the impact range parameter and spatial persistence parameter meet the logical requirements of the event. Finally, the rationality of the event type is verified by combining the movement trend. If all four parameters of a potential wind shear area meet the logical conditions of a certain type of wind shear event, the primary alarm flag of that type of wind shear event is triggered, and the triggered event type and the corresponding parameter matching status are recorded. If none of the four parameters meet the logical conditions of any type of wind shear event, the primary alarm flag is not triggered and the area is not included in the subsequent risk assessment process. Through the above rule matching operation, the triggering logic of the primary alarm is standardized, potential wind shear areas that need further risk assessment are screened out, ineffective assessment operations are reduced, and the efficiency of risk assessment is improved.

[0055] Step 603: For the potential wind shear area of ​​the primary alarm sign, based on the current meteorological background and airport operation status, adaptively adjust the dynamic judgment threshold corresponding to the triggered event type to obtain the real-time judgment threshold. Specifically, this includes: First, for each potential wind shear area that triggers the primary alarm sign, determine the type of wind shear event it triggers, and retrieve the initial dynamic judgment threshold corresponding to that event type. This initial threshold matches the logical conditions in the decision rule set of step 602 and serves as the benchmark for threshold adjustment. Then, retrieve the current real-time meteorological background data and airport operation status data of the airport. The real-time meteorological background data includes parameters such as the airport's current near-surface wind speed, wind direction, air pressure, and cloud cover. The airport operation status data includes: the current takeoff and landing busyness and runway usage. Information such as flight density is collected; the initial dynamic judgment threshold is adaptively adjusted based on the above real-time data. Specifically, if the current meteorological background is complex and the airport is operating during a busy period, the judgment threshold is tightened appropriately, the threshold value is reduced, and the alarm sensitivity is improved; if the current meteorological background is stable and the airport is operating during an idle period, the judgment threshold is relaxed appropriately, the threshold value is increased, and unnecessary alarms are reduced. During the adjustment process, the change range of real-time meteorological parameters and operational status parameters is statistically analyzed, and the adjustment ratio of the threshold is determined based on the change range to ensure that the adjusted threshold is highly adapted to the real-time scenario. Finally, the adjusted threshold is used as the real-time judgment threshold for the corresponding event type in the area. Each area that triggers the primary alarm flag corresponds to a unique real-time judgment threshold, providing a judgment standard for subsequent quantitative comparison.

[0056] Step 604 involves quantifying and comparing the parameters of the potential wind shear area of ​​the primary alarm sign with the real-time judgment threshold to calculate its exceedance range and comprehensive risk index. Specifically, this includes: First, for each potential wind shear area triggering the primary alarm sign, retrieving the wind shear intensity index, impact range parameters, spatial persistence parameters, and movement trend, while simultaneously retrieving the corresponding real-time judgment threshold obtained in step 603, ensuring accurate correspondence between the event type and area location of the parameters and thresholds. Then, each core evaluation parameter is compared with its corresponding real-time judgment threshold one by one to quantify the exceedance range of each parameter. Specifically, the actual value of each parameter is subtracted from the corresponding real-time judgment threshold; if the difference is positive, it indicates that the parameter exceeds the limit, and the difference represents the exceedance. The absolute value is calculated, and then the absolute value exceeding the standard is divided by the real-time judgment threshold to obtain the percentage of exceeding the standard. If the difference is negative or zero, it means that the parameter has not exceeded the standard, and the exceeding standard is recorded as zero. After calculating the exceeding standard of each parameter, the comprehensive risk index of the region is further calculated. Different weights are set according to the importance of the four parameters, with the wind shear intensity index having the highest weight, followed by the influence range parameter, the spatial persistence parameter, and the movement and change trend having the lowest weight. The percentage of exceeding the standard of each parameter is multiplied by the corresponding weight to obtain the risk score of each parameter. Then, the risk scores of the four parameters are added together to obtain the comprehensive risk index of the region. Finally, the exceeding standard of each parameter and the comprehensive risk index are recorded to verify the accuracy of the values ​​in the calculation process and avoid calculation deviations.

[0057] Step 605: Based on the comprehensive risk index, determine the final risk level of the area according to the preset risk level mapping table, and confirm the area that meets the highest risk threshold condition as the final risk area. Specifically, this includes: First, retrieving the preset risk level mapping table. The preset risk level mapping table is a mapping table that is pre-set in combination with flight safety requirements and the degree of wind shear hazard. It is used to correlate the comprehensive risk index with specific risk levels. Its core function is to provide a unified standard for determining the level of the comprehensive risk index and to define the basis for determining the final risk area. The mapping table clearly divides different comprehensive risk index intervals, each interval corresponding to a unique risk level, usually divided into three levels: mild risk, moderate risk, and severe risk. At the same time, a highest risk threshold is set, and the area that meets the threshold condition is determined as the core risk area. The table clearly marks the upper and lower limits of the comprehensive risk index, risk characteristics, and corresponding prevention and control prompts for each risk level to ensure the standardization and consistency of the level determination.

[0058] The specific process of pre-setting this risk level mapping table is carried out step by step, taking into account the actual needs of airport flight safety, the degree of wind shear hazard, and historical risk assessment data. First, it retrieves the airport's wind shear risk assessment records from the past five years, comprehensive risk index calculation data, and the actual hazards of corresponding wind shear events. This data is then categorized and organized, and the degree of hazard, impact range, and impact on flight safety of wind shear events corresponding to different comprehensive risk indices are statistically analyzed. Subsequently, based on the airport's flight safety level requirements, the criteria for classifying risks as mild, moderate, and severe are determined. According to the distribution patterns of historical comprehensive risk index data, the critical values ​​of the comprehensive risk index corresponding to different hazard levels are calculated, dividing the airport into three non-overlapping intervals that completely cover all possible comprehensive risk indices. Each interval corresponds to a risk level, with the lowest comprehensive risk index among them. The risk level map is structured as follows: a range with a low index corresponds to mild risk, a range with a medium index corresponds to moderate risk, and a range with a high index corresponds to severe risk. Next, considering the core requirements of airport flight safety control, a maximum risk threshold is set. This threshold is typically selected as the lower limit or median of the severe risk range to ensure that the core risk areas posing the greatest threat to flight safety are identified. Subsequently, the defined comprehensive risk index ranges, corresponding risk levels, and maximum risk thresholds are integrated to create an initial risk level mapping table. Finally, the initial mapping table is verified and debugged using multiple sets of historical comprehensive risk index data and simulated risk scenarios. The critical values ​​of the index ranges and the maximum risk threshold are adjusted to ensure that the mapping table's judgment results highly match the actual severity of wind shear hazards. Unreasonable range divisions are eliminated, and the accuracy of the mapping table is optimized, ultimately completing the setting of the preset risk level mapping table.

[0059] After the preset risk level mapping table is retrieved and confirmed, the comprehensive risk index is retrieved for each potential wind shear area that triggers the primary alarm flag. This comprehensive risk index is then compared with the index ranges in the risk level mapping table. If the comprehensive risk index is in the mild risk range, the area is classified as mild risk; if it is in the moderate risk range, it is classified as moderate risk; and if it is in the severe risk range, it is classified as severe risk. After determining the risk level, the comprehensive risk index of each area is further checked to determine if it meets the preset maximum risk threshold. If the comprehensive risk index reaches or exceeds the maximum risk threshold, the area is confirmed as the final risk area; otherwise, it is not included in the final risk area range. Finally, the risk level and whether it is a final risk area for each potential wind shear area are archived and integrated into a risk assessment report. This report determines the spatial location, impact range, and risk level of the final risk area, providing a clear basis for subsequent wind shear warning issuance and flight decisions, and improving the wind shear situation assessment system.

[0060] In this embodiment of the invention, comprehensive wind shear situation data is analyzed to extract parameters such as wind shear intensity, impact range, spatial persistence, and movement trend. Core assessment indicators are identified to provide targeted data support for subsequent risk assessments. The extracted parameters are matched with the logical conditions of a preset decision rule set to determine primary alarm flags, standardize alarm triggering logic, and screen areas requiring further assessment, reducing invalid assessment operations. Combined with the current meteorological background and airport operational status, dynamic judgment thresholds are adaptively adjusted to obtain real-time thresholds, adapting to real-time scenario changes and ensuring judgment standards align with actual operational needs. The parameters are quantitatively compared with the real-time judgment thresholds to calculate the exceedance magnitude and comprehensive risk index, quantifying risk details and improving the objectivity and analyzability of risk assessments. Based on the comprehensive risk index and a preset mapping table, risk levels are determined and risk areas are identified, standardizing risk level judgment criteria and determining core risk areas, providing a clear basis for subsequent early warning applications.

[0061] In a preferred embodiment of the present invention, step 700 above, which generates graded early warning information based on the risk level and comprehensive wind shear situation data, and publishes it through a low-latency communication link, synchronously triggering the corresponding airport operation control instruction set, includes: Step 701: Based on the risk level and comprehensive wind shear situation data, determine the warning level, warning text template, and target terminals to be issued. Specifically, this includes: First, verifying the risk level and corresponding situation data for each risk area according to the risk level and comprehensive wind shear situation data to ensure their correlation and accuracy, avoiding misalignment between levels and data; then, determining the corresponding warning level based on the risk level, with a one-to-one correspondence between risk level and warning level: low-level warning for mild risk, medium-level warning for moderate risk, and high-level warning for severe risk, ensuring that the warning level classification matches the risk severity; simultaneously, selecting a preset corresponding warning text template based on the wind shear event types included in the comprehensive wind shear situation data. This preset warning text template is based on wind shear event types, warning levels, and airport warning issuance specifications. The specific preset process is carried out step-by-step based on the actual warning needs of the airport. First, common wind shear event types at the airport are retrieved, including mild, moderate, and severe wind shear and microbursts, and classified according to event type and corresponding warning level; then, combined with the airport's... Based on airport flight safety control requirements, early warning information release standards, and the information acquisition needs of relevant control entities, we designed text template structures for each type and level of early warning, determining the core information modules that each template must include. Subsequently, we reserved spaces within the templates for key information such as risk areas, intensity, movement trends, and expected durations, standardizing the expression format of each information module to ensure template uniformity and relevance. Finally, through multiple simulated early warning release scenarios, we verified the templates' applicability, adjusted the template structure and expression methods, eliminated redundant content, optimized the information presentation order, and completed the setting of all preset early warning text templates. Different types of wind shear events correspond to different text templates, and the reserved information spaces within the templates ensure the relevance of the early warning texts. Finally, considering the spatial location and impact range of the risk area, we determined the target terminals for the early warning information release. Target terminals include airport tower control terminals, flight crew receiving terminals, ground handling dispatch terminals, and airport operations monitoring center terminals, ensuring that early warning information accurately reaches all relevant control entities, laying the foundation for subsequent early warning releases and command triggering.

[0062] Step 702: Extract the geometric location, spatial extent, wind shear event type, intensity, movement vector, and estimated duration based on the evolution trend of the risk area from the comprehensive wind shear situation data. Specifically, this includes: First, retrieving the comprehensive wind shear situation data, focusing on the situation data corresponding to each risk area, analyzing the data content one by one, removing redundant information irrelevant to the warning issuance, and retaining the core effective content; then, extracting various key information of the risk area from the comprehensive wind shear situation data, specifically including: geometric location, spatial extent, wind shear event type, intensity, movement vector, and estimated duration information; among which, the geometric location is extracted from the three core points of the risk area. The system extracts the three-dimensional coordinate boundaries and coverage area of ​​the risk area using spatial coordinates. The wind shear event type is directly derived from the type defined in the situation data. The intensity is extracted from the average wind speed gradient value in the situation data. The movement vector is extracted from the movement direction and speed of the risk area recorded in the situation data. The expected duration information is calculated based on the spatial persistence parameters and movement change trends extracted from the situation data. The current duration of the wind shear event is statistically analyzed, and its subsequent expected duration is calculated by combining the movement speed and intensity change patterns. During the extraction process, the accuracy of each type of information is checked one by one to avoid problems such as missing information or abnormal values, ensuring that the extracted information can completely support the filling of the warning text.

[0063] Step 703 involves filling the risk level, geometric location, spatial range, wind shear event type, intensity, movement vector, and expected duration information into the warning text template according to a predetermined format, generating a standardized hierarchical warning information data package. Specifically, this includes: First, retrieving the corresponding warning text template selected in step 701 and verifying whether the template type matches the wind shear event type and warning level to ensure template applicability; then determining the predetermined format of the template, which is based on airport warning information release specifications, target terminal parsing requirements, and key information presentation logic. The specific pre-determining process is carried out step by step, first prioritizing the core information to be presented in the warning information, determining the warning level and wind shear event type as the primary information to be presented, and then determining the presentation order of geometric location, spatial range, intensity, movement vector, and expected duration in sequence; then standardizing the expression format of various types of information, unifying the presentation specifications of coordinates, values, and time to ensure consistent warning information expression for different risk areas; and simultaneously, considering the parsing capabilities of various target terminals, setting the layout format, character spacing, and identification rules for information filling. The format is designed to facilitate rapid parsing and recognition by terminals. Multiple template filling and terminal parsing tests are conducted to verify the applicability and standardization of the format, fine-tuning the information sorting and expression to complete the predetermined format setting. Following this predetermined format, various key information and risk levels are filled into the corresponding reserved positions in the templates, adhering to the predetermined format requirements and standardizing the expression of information to ensure that all information is arranged in an orderly, clear, and easy-to-understand manner. For example, the warning level and wind shear event type are marked first, followed by the geometric location, spatial range, intensity, movement vector, and expected duration of the risk area. After filling, the warning text undergoes preliminary verification to check the accuracy, completeness, and format standardization of the filled information, eliminating redundant expressions and numerical errors, and correcting format deviations. Finally, the verified warning text is packaged into a standardized hierarchical warning information data package. The data package uses a unified data format to facilitate rapid parsing and recognition by various target terminals, while also marking the risk area number, warning level, and generation time of the data package to ensure traceability and prepare for subsequent warning releases.

[0064] Step 704 involves transmitting the tiered early warning information data packets to the target terminals in real time via a low-latency communication link. This includes: First, retrieving the encapsulated tiered early warning information data packets and verifying the target terminal list corresponding to each data packet to ensure the data packets match the target terminals and avoid mistransmissions or omissions. Then, determining the preset mode of the low-latency communication link. This low-latency communication link is preset based on the early warning information transmission requirements, the distribution of target terminals at the airport, and anti-interference requirements. The specific preset process is carried out step by step: First, based on the real-time transmission requirements of the early warning information, determining the minimum transmission rate and maximum latency threshold required for the link to ensure rapid delivery of the early warning information. Then, based on the distribution location and number of target terminals at the airport, planning a dedicated transmission channel to avoid public networks and reduce the impact of network congestion and signal interference on transmission delay. Next, setting the highest priority for early warning information transmission to ensure that early warning data packets take precedence over other ordinary data transmissions, further ensuring transmission timeliness. Simultaneously, configuring the adaptation parameters for real-time monitoring and retransmission mechanisms. To ensure rapid response in case of link anomalies, redundant channels are reserved to handle sudden transmission failures. Multiple field transmission tests verify the link's transmission rate, stability, and anti-interference capabilities. Transmission parameters and channel planning are fine-tuned to optimize link performance and complete the pre-set low-latency communication link. This pre-set low-latency communication link, using a dedicated transmission channel, prioritizes the transmission rate of warning information and avoids transmission delays caused by network congestion. During transmission, warning information data packets are monitored in real time, tracking their transmission status. If transmission interruptions or loss occur, a retransmission mechanism is immediately activated to resend the corresponding data packets, ensuring successful reception by each target terminal. Simultaneously, the release time, receiving terminal, and receiving status of each data packet are recorded to form a release log for subsequent traceability and verification. Through the low-latency communication link, tiered warning information data packets are released to all target terminals in real time, ensuring that relevant control entities can promptly obtain warning information and quickly prepare for response.

[0065] Step 705: Based on the risk level and the geometric location of the risk area, automatically match and trigger the corresponding airport operation control instruction set. Specifically, this includes: First, verifying the correspondence between the risk level, geometric location, and the preset airport operation control instruction set for each risk area to ensure accurate instruction matching. The preset airport operation control instruction set is based on wind shear risk level, risk area geometric location, and airport operation management specifications. The preset process is carried out step-by-step, first identifying common combinations of wind shear risk levels and risk area geometric locations at airports, covering scenarios such as mild risk (runway not covered), mild risk (runway close), moderate risk (runway not covered), moderate risk (runway close), severe risk (runway close), and severe risk (runway covered). Various scenarios are considered; then, based on the actual degree of hazard and scope of impact of each scenario, combined with airport flight safety control requirements, runway operation specifications, and ground handling scheduling standards, the corresponding control command content for each scenario is designed, and the specific commands included in each command set are determined, such as early warning and reminder commands, routine monitoring commands, flight deceleration commands, ground handling standby commands, flight grounding commands, runway closure commands, and personnel evacuation commands, etc.; subsequently, the execution order, triggering conditions, and corresponding responsible parties of various commands are determined, and the one-to-one correspondence between different command sets and risk levels and geometric locations is sorted out to form an initial command set and matching rules; finally, through multiple sets of airport operation simulation scenarios, the rationality and executability of the command set are verified, the command content, execution order, and triggering conditions are adjusted, the matching rules are optimized, and the setting of the preset airport operation control command set is completed.

[0066] Subsequently, based on the risk level and the geometric location of the risk area, the system automatically matches the corresponding airport operation control instruction set. Different risk levels and different geometric locations correspond to different control instruction sets. For example, areas with low risk and not covered by the runway trigger early warning reminders and routine monitoring instruction sets; areas with medium risk and close to the runway trigger flight deceleration and ground crew standby instruction sets; and areas with high risk and covered by the runway trigger flight grounding, runway closure, and personnel evacuation instruction sets. After the instruction set is matched, the corresponding instruction is automatically executed. At the same time, the instruction triggering status, execution content, and corresponding risk area information are synchronously fed back to the airport operation monitoring center terminal, facilitating real-time monitoring of instruction execution by staff. If any abnormality occurs during instruction triggering, an alert signal is immediately issued to remind staff to intervene in a timely manner, ensuring that the instruction set can be executed smoothly. This achieves efficient linkage between early warning information and airport operation control, ensuring airport operation safety.

[0067] In this embodiment of the invention, based on the risk level and comprehensive wind shear situation data, the warning level, warning text template, and target terminal are determined, defining the core elements of the warning and providing preliminary support for the standardized generation and accurate release of subsequent warning information. Various key information about the risk area is extracted from the comprehensive wind shear situation data, integrating the core content of the warning and providing comprehensive data support for filling in the warning text. Various information is filled into the warning text template according to a predetermined format to generate a standardized warning information data package, standardizing the form of the warning data and improving the uniformity and standardization of the warning information. The warning information data package is released to the target terminal in real time through a low-latency communication link, ensuring efficient transmission of the warning information and ensuring that the information accurately reaches the corresponding terminal. Based on the risk level and the geometric location of the risk area, the corresponding airport operation control instruction set is automatically matched and triggered, realizing the linkage between warning information and operation control, standardizing the instruction triggering logic, and improving the airport operation response efficiency.

[0068] like Figure 2 As shown, embodiments of the present invention also provide a wind shear early warning system based on radar technology, comprising: The acquisition module is used to deploy multiple X-band Doppler weather radars to form a detection network, covering the airport runway and surrounding low-altitude airspace and performing periodic scans to acquire raw radar echo data. The reconstruction module is used to transmit the raw radar echo data to the edge computing node in real time, and obtain three-dimensional spatial airflow data through point cloud reconstruction, filtering and noise reduction. The identification module is used to perform spatial interpolation based on three-dimensional airflow data to generate wind field data; calculate wind speed gradient and air vortex characteristic parameters based on wind field data, and identify areas that meet the initial warning conditions as potential wind shear areas by abrupt change detection and threshold comparison. The calibration module is used to divide the airspace covered by the potential wind shear area into multiple calibration sub-regions based on the distribution of the potential wind shear area and the spatial positional relationship of multiple radars in the detection network, and to calculate the spatial calibration coefficient for each calibration sub-region. The fusion module is used to fuse potential wind shear area data with spatial calibration coefficients, and obtain comprehensive wind shear situation data through partitioned weighted calibration. The assessment module is used to conduct risk assessment based on the comprehensive wind shear situation data, using a preset set of decision rules combined with dynamically adjusted judgment thresholds, to identify risk areas and their risk levels. The early warning module is used to generate graded early warning information based on the risk level and comprehensive wind shear situation data, and to release it through a low-latency communication link, simultaneously triggering the corresponding airport operation control command set.

[0069] It should be noted that this system is a system corresponding to the above method. All implementation methods in the above method embodiments are applicable to this embodiment and can achieve the same technical effect.

[0070] 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. A wind shear early warning method based on radar technology, characterized in that, The method includes: Step 100: Deploy multiple X-band Doppler weather radars to form a detection network, covering the airport runway and surrounding low-altitude airspace and periodically scanning to obtain raw radar echo data. Step 200: The raw radar echo data is transmitted to the edge computing node in real time. Through point cloud reconstruction, filtering and noise reduction processing, three-dimensional spatial airflow data is obtained. Step 300: Spatial interpolation is performed based on three-dimensional airflow data to generate wind field data; wind speed gradient and air vorticity characteristic parameters are calculated based on wind field data, and areas that meet the initial warning conditions are identified as potential wind shear areas by abrupt change detection and threshold comparison. Step 400: Based on the distribution of potential wind shear areas and the spatial positional relationship of multiple radars in the detection network, the airspace covered by the potential wind shear areas is divided into multiple calibration sub-regions, and spatial calibration coefficients are calculated for each calibration sub-region. Step 500: Integrate potential wind shear area data with spatial calibration coefficients, and obtain comprehensive wind shear situation data through partitioned weighted calibration; Step 600: Based on the comprehensive wind shear situation data, a risk assessment is conducted using a preset set of decision rules combined with dynamically adjusted judgment thresholds to identify risk areas and their risk levels. Step 700: Generate graded early warning information based on the risk level and comprehensive wind shear situation data, and release it through a low-latency communication link, simultaneously triggering the corresponding airport operation control instruction set.

2. The wind shear early warning method based on radar technology according to claim 1, characterized in that, Step 100 includes: Based on the airport runway configuration and the distribution of surrounding terrain obstacles, the deployment locations and number of radar stations are determined to form a detection network with cross-coverage capability for the runway end, climb corridor and approach corridor. Based on the deployment location and number of the detection network, a scanning mode and parameters are set for each radar in the detection network. The parameters include the scanning elevation angle sequence, azimuth angle range and scanning period, so as to perform periodic volume scanning on the low-altitude airspace. According to the scanning mode and parameters, control each radar to perform scanning synchronously and receive radar echo signals from each range and azimuth angle. The radar echo signal is subjected to analog-to-digital conversion and preliminary quality control to remove non-meteorological echoes such as ground clutter, and raw radar echo data including radial velocity, spectral width and reflectivity are obtained.

3. The wind shear early warning method based on radar technology according to claim 2, characterized in that, Step 200 includes: The raw radar echo data is transmitted in real time via a high-speed network to an edge computing node deployed locally at the airport, and decoded into radar base data containing range, azimuth, elevation, radial velocity, and reflectivity. The range, azimuth, and elevation information of each measurement unit in the radar base data are transformed from the spherical coordinate system to the three-dimensional Cartesian coordinate system to obtain the three-dimensional rectangular coordinates of each measurement unit in space. Based on the three-dimensional rectangular coordinates of each measurement unit in space and their corresponding meteorological parameters, the base data from multiple radars are spatially aligned and temporally interpolated and fused to generate an initial three-dimensional spatial airflow point cloud covering the low-altitude airspace. The initial three-dimensional spatial airflow point cloud is gridded and statistically filtered to remove sparsely distributed isolated noise points and elevation clutter points with abnormal reflectivity, so as to obtain the filtered point cloud. The filtered point cloud is subjected to temporal moving average and spatial wavelet denoising to suppress random noise and atmospheric background clutter, thus obtaining three-dimensional spatial airflow data.

4. The wind shear early warning method based on radar technology according to claim 3, characterized in that, Step 300 includes: Three-dimensional spatial interpolation is performed on the wind speed vector in the three-dimensional airflow data to generate three-dimensional gridded wind field data with uniform spatial resolution. Based on the three-dimensional gridded wind field data, the wind speed vector difference between each grid point and its adjacent grid points is calculated to obtain the wind speed gradient tensor describing the rate and direction of wind speed change, and the vorticity scalar value describing the intensity of air rotation is calculated as a wind field characteristic parameter. Based on the wind speed gradient tensor and vorticity scalar value, a sliding window statistical analysis is performed in three-dimensional space to detect the abrupt change location and intensity of the wind speed gradient magnitude and vorticity value in space, thereby obtaining the wind field abrupt change feature vector. The wind field abrupt change feature vector is compared point by point with the preset initial warning threshold, and all spatial grid points that meet the initial warning threshold conditions are marked. The spatial grid set is subjected to three-dimensional spatial clustering and region growing to form a continuous spatial region, which is then identified as a potential wind shear region.

5. The wind shear early warning method based on radar technology according to claim 4, characterized in that, Step 400 includes: Based on all radars participating in the detection of the potential wind shear region in the detection network, the three-dimensional spatial coordinates of their sites are obtained, and each radar site is defined as a generator based on the three-dimensional spatial coordinates. Based on each generator and its three-dimensional spatial coordinates, spatial region division is performed, so that the airspace covered by the potential wind shear region is divided into multiple three-dimensional spatial regions, wherein the spatial position in each three-dimensional spatial region is closest to its corresponding generator, and each of the divided three-dimensional spatial regions is defined as a calibration sub-region. For each calibration sub-region, determine its corresponding generator radar; based on the spatial range of the calibration sub-region, select all radars whose detection ranges have spatial intersection with the calibration sub-region; The detection values ​​of wind field characteristic parameters in the calibration sub-region are obtained by the generator radar and each selected radar, and the wind field characteristic parameters include wind speed and wind direction.

6. The wind shear early warning method based on radar technology according to claim 5, characterized in that, Step 400 further includes: For each calibration sub-region, the wind field characteristic parameter detection values ​​of the generator radar are compared with the corresponding detection values ​​of other radars one by one, and the differences in wind speed and wind direction of each pair of detection values ​​are calculated to obtain the differences in wind speed and wind direction for each group. Each group of wind speed and wind direction differences is constructed as a two-dimensional difference vector, and the Euclidean distance of each two-dimensional difference vector is calculated to form a set of distance values; The statistical root mean square error is calculated for the set of distance values, and the reciprocal of the root mean square error is normalized to a preset weight interval to obtain the spatial calibration coefficient that characterizes the consistency of multi-radar data within the calibration sub-region.

7. The wind shear early warning method based on radar technology according to claim 6, characterized in that, Step 500 includes: Acquire spatial extent data for each potential wind shear region and spatial calibration coefficients for each corresponding calibration sub-region; For each potential wind shear region, the spatial calibration coefficients of all calibration sub-regions within its coverage area are used as the fusion weights of the wind field characteristic parameter data of each sub-region. The wind speed, wind direction and air vorticity parameters detected in the calibration sub-regions from multiple radars are weighted and averaged to obtain the fused wind field characteristics of the region. For each potential wind shear region, its fused wind field characteristics are integrated with the region's spatial extent, location, and timestamp to generate comprehensive wind shear situation data for that region.

8. The wind shear early warning method based on radar technology according to claim 7, characterized in that, Step 600 includes: Analyze the comprehensive wind shear situation data for each potential wind shear region to extract its wind shear intensity index, influence range parameter, spatial persistence parameter, and movement trend; The wind shear intensity index, influence range parameter, spatial persistence parameter and movement trend are matched with the logical conditions defined in the preset decision rule set to determine whether the primary alarm flags of various wind shear events are triggered. For the potential wind shear area of ​​the primary alarm sign, the dynamic judgment threshold corresponding to the triggered event type is adaptively adjusted according to the current meteorological background and airport operation status to obtain the real-time judgment threshold. The parameters of the potential wind shear area of ​​the primary alarm sign are quantitatively compared with the real-time judgment threshold to calculate the exceedance range and comprehensive risk index. Based on the comprehensive risk index, the final risk level of the region is determined according to the preset risk level mapping table, and the region that meets the highest risk threshold condition is identified as the final risk region.

9. The wind shear early warning method based on radar technology according to claim 8, characterized in that, Step 700 includes: Based on the aforementioned risk level and comprehensive wind shear trend data, determine the warning level, warning text template, and target terminals to be issued the warning; From the comprehensive wind shear situation data, extract the geometric location, spatial range, wind shear event type, intensity, movement vector, and estimated duration based on the evolution trend of the risk area; The risk level, geometric location, spatial range, wind shear event type, intensity, movement vector, and expected duration information are filled into the warning text template according to a predetermined format to generate a standardized graded warning information data package. The graded early warning information data packets are transmitted to the target terminal in real time via a low-latency communication link; Based on the risk level and the geometric location of the risk area, the corresponding airport operation control instruction set is automatically matched and triggered.

10. A wind shear early warning system based on radar technology, wherein the system implements the method as described in any one of claims 1 to 9, characterized in that, include: The acquisition module is used to deploy multiple X-band Doppler weather radars to form a detection network, covering the airport runway and surrounding low-altitude airspace and performing periodic scans to acquire raw radar echo data. The reconstruction module is used to transmit the raw radar echo data to the edge computing node in real time, and obtain three-dimensional spatial airflow data through point cloud reconstruction, filtering and noise reduction. The identification module is used to perform spatial interpolation based on three-dimensional airflow data to generate wind field data; calculate wind speed gradient and air vortex characteristic parameters based on wind field data, and identify areas that meet the initial warning conditions as potential wind shear areas by abrupt change detection and threshold comparison. The calibration module is used to divide the airspace covered by the potential wind shear area into multiple calibration sub-regions based on the distribution of the potential wind shear area and the spatial positional relationship of multiple radars in the detection network, and to calculate the spatial calibration coefficient for each calibration sub-region. The fusion module is used to fuse potential wind shear area data with spatial calibration coefficients, and obtain comprehensive wind shear situation data through partitioned weighted calibration. The assessment module is used to conduct risk assessment based on the comprehensive wind shear situation data, using a preset set of decision rules combined with dynamically adjusted judgment thresholds, to identify risk areas and their risk levels. The early warning module is used to generate graded early warning information based on the risk level and comprehensive wind shear situation data, and to release it through a low-latency communication link, simultaneously triggering the corresponding airport operation control command set.

Images