Zenith rotation dual-polarization radar calibration diagnosis method, system, device and medium

Abstract

This invention relates to the field of meteorological radar, and more particularly to a zenith rotating dual-polarization radar calibration and diagnostic method, system, device, and medium. The method includes: if preset triggering conditions are met, controlling the radar antenna to perform a vertical pointing operation towards the zenith; driving the radar antenna to perform azimuth rotation scanning, acquiring polarization base data with azimuth full-circumference distribution characteristics and radial height-level characteristics; screening and cleaning the polarization base data to filter out interference echoes and non-liquid precipitation echoes, obtaining a benchmark analysis sample; performing multi-dimensional modeling on the benchmark analysis sample to separate fluctuation errors, obtaining calibration characteristic quantities for online data compensation, and mapping and identifying the radar hardware component status based on the comparison results of the standard deviation profile of the benchmark analysis sample and theoretical accuracy indicators to output diagnostic information. Therefore, this invention significantly enhances the accurate diagnostic capability for the degradation risk of radar underlying components while improving the calibration accuracy of dual-polarization radar.

Description

Technical Field

[0001] This invention relates to the field of meteorological radar technology, and in particular to a zenith rotating dual polarization radar calibration and diagnostic method, system, device, and medium. Background Technology

[0002] Dual-polarization weather radar is a core piece of equipment for modern meteorological monitoring. Its differential reflectivity is a crucial parameter for identifying the phase of precipitation particles and for quantitative precipitation estimation. According to meteorological observation specifications, the differential reflectivity system deviation of radar must typically be strictly controlled within ±0.2 dB. Because precipitation estimation is highly sensitive to this parameter, a deviation exceeding 0.2 dB can lead to an error of more than 20% in liquid precipitation estimation.

[0003] However, during long-term operational use, radar systems are subject to unpredictable dynamic drift due to a combination of factors, including ambient temperature fluctuations, water accumulation or aging of the radome, transmitter power drift, and the stability of the dual-channel receiver link. Therefore, achieving high-frequency, high-precision online calibration and system status diagnosis has become a crucial issue in ensuring the quality of radar observations.

[0004] In existing calibration techniques, the built-in signal calibration method primarily involves injecting electrical signals internally for link detection. However, because this closed-loop process doesn't cover both antenna radiation and reception, it fails to reflect the physical losses and polarization changes of the radome, feed network, and related mechanical interfaces during actual operation, resulting in significant localization of the calibration results. In contrast, external source calibration methods attempt to achieve full link coverage, but methods like the solar method are limited by the incoherent characteristics of natural sources, only monitoring the receiving link. Furthermore, the calibration frequency is easily affected by weather conditions, making it difficult to comprehensively assess the overall performance of the radar system. Using metal spheres or towers for calibration is problematic due to severe multipath interference in the field and a high degree of manual intervention, leading to lengthy calibration processes that are difficult to automate and achieve high-frequency calibration.

[0005] Furthermore, while there are schemes for zenith observations utilizing the physical properties of precipitation particles, current implementations generally focus on linear averaging of observations at a single moment or specific angle. Due to a lack of deeper statistical analysis of the data, existing analytical models struggle to establish a logical connection between the fluctuations in the data and the operational status of the radar's underlying hardware. Therefore, when the accuracy of radar data decreases due to performance degradation of underlying components (such as phase-locked loops and receiver links), current technologies often only provide the deviation of the observed values, failing to offer effective diagnostic support for underlying hardware failures. Summary of the Invention

[0006] (a) Technical problems to be solved In view of the above-mentioned shortcomings and deficiencies of the prior art, the present invention provides a zenith rotating dual polarization radar calibration and diagnostic method, system, device and medium, which solves the technical problems of the prior art, which cannot take into account the comprehensiveness of radar full-link dynamic performance monitoring and calibration accuracy, and lacks real-time mapping of observation data statistical characteristics to hardware operating status, thus making it difficult to meet the needs of high-frequency continuous calibration for operational purposes and the diagnostic requirements for the risk of degradation of underlying components.

[0007] (II) Technical Solution To achieve the above objectives, the main technical solutions adopted by the present invention include: In a first aspect, embodiments of the present invention provide a zenith rotating dual-polarization radar calibration and diagnostic method, comprising: In response to the coordinated fulfillment of preset triggering conditions by the acquired environmental detection parameters and meteorological echo characteristics, the radar antenna is controlled to perform an operation of pointing vertically towards the zenith; In the vertical pointing state, the radar antenna is driven to perform azimuth rotation scanning, and polarization base data with azimuth full-circumference distribution characteristics and radial height level characteristics are collected based on the preset sampling configuration. Spatial domain constraint screening and physical parameter threshold cleaning were performed on the polarization basis data to filter out interference echoes and non-liquid precipitation echoes, and obtain benchmark analysis samples. Multidimensional modeling is performed on the benchmark analysis samples to separate the periodic fluctuation error caused by the asymmetry of the radar structure, obtain calibration characteristic quantities and use them for real-time compensation of polarization basis data, and map and identify the status of the radar hardware components based on the comparison results of the standard deviation profile of the benchmark analysis samples and the theoretical accuracy index to output diagnostic information.

[0008] Optionally, in response to the coordinated fulfillment of preset triggering conditions by the acquired environmental detection parameters and meteorological echo characteristics, the radar antenna is controlled to perform an operation of pointing vertically towards the zenith, including: Real-time acquisition of environmental detection parameters and meteorological echo characteristics, including ground wind speed and wind speed at a preset altitude, and meteorological echo characteristics including reflectivity factor distribution; The ground wind speed and the wind speed at a preset height are compared with preset wind speed thresholds, and the period of vertical fall of precipitation particles is determined when both the ground wind speed and the wind speed at a preset height are less than the preset wind speed threshold. During the vertical fall of precipitation particles, meteorological echoes with intensity within a preset intensity range are selected from the reflectivity factor distribution as candidate echoes. Identify the zero-degree layer height in the reflectivity factor distribution, and combine it with the preset antenna far-field region to delineate a pure liquid precipitation area located above the lower limit of the antenna far-field region and below the zero-degree layer height in the spatial distribution of candidate echoes. A calibration task window is constructed using a defined pure liquid precipitation area. In response to the opening of the calibration task window, the radar's current detection task is paused and the scanning parameters are saved, driving the radar antenna to adjust to a state of vertical pointing towards the zenith.

[0009] Optionally, in the vertical pointing state, the radar antenna is driven to perform azimuth rotation scanning, and polarization basis data with azimuth full-circle distribution characteristics and radial height hierarchical characteristics are acquired based on a preset sampling configuration, including: Keep the radar antenna in a vertical pointing state and drive the radar antenna to perform a full or multiple rotational scan in the azimuth direction at a preset rotation speed; The radar is controlled to transmit detection waveforms at a preset pulse repetition frequency so that the number of independent samples in each azimuth step interval reaches a preset statistical significance threshold during the continuous rotation scanning of the radar antenna. During the continuous rotating scan and detection waveform transmission of the radar antenna, the original echo signals corresponding to the full azimuth angle and full detection range are acquired in real time. By utilizing the linear mapping relationship between the full detection range library and the vertical height under vertical pointing conditions, the original echo signal is converted into polarization-based data with azimuth full-circumference distribution characteristics and radial height-level characteristics; wherein, the polarization-based data includes: differential reflectivity factor, reflectivity factor, differential propagation phase shift and correlation coefficient.

[0010] Optionally, spatial domain constraint screening and physical parameter threshold cleaning are performed on the polarization basis data to filter out interference echoes and non-liquid precipitation echoes, obtaining benchmark analysis samples, including: The polarization-based data is constrained within a pure liquid precipitation region defined in a vertically pointing state. The lower boundary limit of the pure liquid precipitation region is used to eliminate the blind zone and near-field interference introduced by the transient delay of the transceiver switch. The upper boundary limit of the pure liquid precipitation region is used to eliminate the echo interference of non-spherical ice phase particles above the zero-degree layer, thus obtaining the polarization-based data that has been spatially constrained and filtered. The signal-to-noise ratio, signal quality index, and correlation coefficient in the polarization basis data that have been filtered by spatial domain constraints are compared with a preset set of physical quantity thresholds to remove noise echoes and non-meteorological target interference, thus obtaining polarization basis data after physical parameter threshold cleaning. The total number of polarization basis data after physical parameter threshold cleaning is counted, and it is determined whether the total number of samples has reached the preset minimum calibration sample size threshold. In response to the total number of samples reaching the preset minimum calibration sample size threshold, the polarization basis data after physical parameter threshold cleaning is used as the benchmark analysis sample for performing calibration diagnosis.

[0011] Optionally, multidimensional modeling is performed on the benchmark analysis samples to separate the periodic fluctuation errors caused by radar structural asymmetry, obtain calibration characteristic quantities, and use them for real-time compensation of polarization basis data, including: The differential reflectance factor values ​​in the benchmark analysis sample are denoised or outlier removed, and the estimated system bias is obtained through statistical calculation. The benchmark analysis samples are binned and statistically analyzed in the azimuth domain according to a preset angular resolution. The confidence evaluation index for each azimuth bin is constructed by combining the reflectivity factor, signal-to-noise ratio, correlation coefficient, signal quality index and sample number contained in the benchmark analysis samples. Based on the confidence evaluation index and the system bias estimate, the periodic error decoupling model is iteratively solved to achieve confidence constraints on bin samples with different azimuth angles during the parameter fitting process. After the periodic error decoupling model converges, the solution result of the DC bias component is obtained. The fitting objective of the periodic error decoupling model includes the DC bias component to be solved, as well as the multi-order harmonic disturbance component adaptively determined based on the convergence of the fitting residual, model complexity constraints, and the energy ratio of the periodic component. The solution results are used as the differential reflectivity system deviation after removing multi-period composite fluctuation errors. At the same time, the amplitude, phase and energy ratio of multi-order harmonic disturbance components in the periodic error decoupling model are extracted as auxiliary diagnostic information to characterize the state of radar hardware structure. The differential reflectivity system deviation is converted into a calibration compensation constant and fed back in real time to the compensation parameter table of the preset signal processor for online automatic calibration of polarization basis data.

[0012] Optionally, based on the comparison between the standard deviation profile of the benchmark analysis sample and the theoretical accuracy index, the hardware component status of the radar is mapped and identified to output diagnostic information, including: Calculate the measured standard deviation profiles of the distribution of each parameter with height in the benchmark analysis sample. The measured standard deviation profiles include the differential reflectivity factor standard deviation profile, the differential propagation phase shift standard deviation profile, the correlation coefficient standard deviation profile, and the reflectivity factor standard deviation profile. Each profile in the measured standard deviation profile is compared with the corresponding preset theoretical standard deviation profile, and the deviation characteristic value of the measured standard deviation profile relative to the corresponding theoretical standard deviation profile is analyzed. In response to the deviation of the eigenvalue characterizing the differential reflectivity factor standard deviation profile being higher than the corresponding theoretical standard deviation profile in the pure liquid precipitation region, the stability of the radar's dual-channel receiving link is determined to be reduced and the first diagnostic information is output. In response to abnormal fluctuations in the differential propagation phase shift standard deviation profile that deviates from the eigenvalue, which exceeds the preset fluctuation range relative to the corresponding theoretical standard deviation profile, the radar transmitter phase-locked loop phase noise is determined to be degraded or the radar coherence is reduced, and second diagnostic information is output. In response to the deviation of the eigenvalue characterization correlation coefficient standard deviation profile being higher than the corresponding theoretical standard deviation profile in the pure liquid precipitation region, or the dispersion abrupt change exceeding the preset jump threshold between adjacent height layers, the radar dual polarization channel consistency is determined to be reduced, the polarization isolation is deteriorated, or the feed and polarization separation network performance is abnormal, and third diagnostic information is output. In response to the deviation of the standard deviation profile of the reflectivity factor, which represents the deviation of the eigenvalue, from the corresponding theoretical standard deviation profile, the radar stability is determined to have changed and the fourth diagnostic information is output. By integrating differential reflectivity system deviation, first diagnostic information, second diagnostic information, third diagnostic information, fourth diagnostic information, and auxiliary diagnostic information, a radar health status inspection report is generated.

[0013] Optionally, the measured standard deviation profiles of each parameter's distribution with height in the benchmark analysis sample are calculated, including: The pure liquid precipitation area is divided into multiple height-level sampling windows according to the preset height resolution, and the benchmark analysis samples mapped to each height-level sampling window are extracted. The correlation coefficients of the benchmark analysis samples within the sampling window at each altitude level are statistically analyzed, and the correlation time of the pulse sequence within the sampling window at each altitude level is calculated by combining the obtained radar antenna rotation speed, pulse repetition frequency and dwell time. The total number of samples within each height level sampling window is corrected using the relevant time, and the number of independent samples corresponding to the differential reflectivity factor and reflectivity factor is calculated. The equivalent number of samples corresponding to the differential propagation phase shift and correlation coefficient is also calculated. The correlation coefficient, number of independent samples, and number of equivalent samples within each height level sampling window are used as feature variables and substituted into the standard deviation analytical model constructed based on the parameter statistical coherence characteristics. By decoupling the influence of signal coherence time on sampling capacity, the measured standard deviation of each parameter corresponding to each height level sampling window is obtained in real time. The measured standard deviations of each parameter are processed continuously according to the vertical height order to generate a measured standard deviation profile that includes the differential reflectivity factor standard deviation profile, the differential propagation phase shift standard deviation profile, the correlation coefficient standard deviation profile, and the reflectivity factor standard deviation profile.

[0014] Secondly, embodiments of the present invention provide a zenith-rotating dual-polarization radar calibration and diagnostic system, comprising: a zenith detection control module, used to control the radar antenna to perform vertical pointing towards the zenith in response to the coordinated satisfaction of preset triggering conditions by acquired environmental detection parameters and meteorological echo characteristics; a scanning sampling execution module, used to drive the radar antenna to perform azimuth rotation scanning in the vertical pointing state, and to collect polarization base data with azimuth full-circumference distribution characteristics and radial height level characteristics based on preset sampling configuration; a sample cleaning and quality control module, used to perform spatial domain constraint screening and physical parameter threshold cleaning on the polarization base data, filter out interference echoes and non-liquid precipitation echoes, and obtain benchmark analysis samples; and a calibration and diagnostic integration module, used to perform multi-dimensional modeling on the benchmark analysis samples to separate the periodic fluctuation errors caused by radar structural asymmetry, obtain calibration feature quantities and use them for real-time compensation of polarization base data, and map and identify the state of radar hardware components based on the comparison results of the standard deviation profile of the benchmark analysis samples and theoretical accuracy indicators to output diagnostic information.

[0015] Thirdly, embodiments of the present invention provide a zenith rotating dual-polarization radar calibration and diagnostic device, comprising: at least one controller; and a memory communicatively connected to the at least one controller; wherein the memory stores instructions executable by the at least one controller, the instructions being executed by the at least one controller to enable the at least one controller to perform the zenith rotating dual-polarization radar calibration and diagnostic method as described above.

[0016] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a controller, implement the zenith rotating dual-polarization radar calibration and diagnostic method as described above.

[0017] (III) Beneficial Effects The beneficial effects of this invention are: First, this invention controls the radar antenna to point vertically towards the zenith by responding to the coordinated fulfillment of preset triggering conditions by the acquired environmental detection parameters and meteorological echo characteristics. This environmentally adaptive triggering logic enables the radar to automatically perform calibration logic during volume scan intervals during operational use, achieving high-frequency, real-time online monitoring without manual intervention or field auxiliary equipment. This overcomes the drawbacks of traditional calibration methods, such as discontinuity and the inability to automate inspections.

[0018] Secondly, to address azimuth asymmetry errors, this invention drives the radar antenna to perform azimuth rotation scanning in a vertical pointing state, acquiring polarization base data with azimuth full-circle distribution characteristics and radial height-level characteristics. It utilizes the physical symmetry of natural rain particles under vertical observation to construct a true end-to-end calibration path, which not only compensates for the gain difference between the transmit and receive channels but also eliminates systematic errors caused by radome wetting, aging, and antenna feed deviation, ensuring that the calibration accuracy of the differential reflectivity system deviation is stably controlled within ±0.2dB.

[0019] Next, this invention effectively filters out interference echoes and non-liquid precipitation echoes by performing spatial domain constraint screening and physical parameter threshold cleaning on the polarization basis data. This process ensures that the benchmark analysis samples used for calibration and diagnosis have extremely high physical purity, eliminating measurement deviations caused by non-target objects from the data source, and providing data support for high-precision calibration.

[0020] Furthermore, by performing multi-dimensional modeling on the benchmark analysis sample, this invention separates the periodic fluctuation errors caused by the asymmetry of the radar structure (such as the fluctuations in the polarization loss of the rotating joint and feeder with the azimuth angle). By obtaining the calibration characteristic quantity and using it for real-time compensation, this invention can offset the sinusoidal oscillation error generated by the radar azimuth scanning mechanism during continuous rotation, eliminate the influence of azimuth asymmetry on the calibration results, and further improve the dynamic calibration accuracy.

[0021] Finally, this invention establishes a real-time mapping logic between the fluctuation characteristics of the detection data and the operating status of the radar's underlying hardware by mining the standard deviation profile of the benchmark analysis samples and comparing it with theoretical accuracy indicators. This medical "CT scan-like" diagnostic method can map, identify, and output diagnostic information of underlying hardware components such as the stability of the dual-channel receiver link, the degradation of the transmitter phase-locked loop phase noise, and the decrease in system coherence based on abnormal deviations in the standard deviation, thus realizing a closed loop of calibration and hardware inspection.

[0022] Therefore, through scientific process design and multi-dimensional data mining, this invention not only improves the calibration accuracy of dual-polarization radar, but also significantly enhances the ability to accurately diagnose the degradation risk of radar underlying components. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating the method provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the specific process of method step S1 provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the zenith scanning mode provided in an embodiment of the present invention; Figure 4This is a schematic diagram of the specific process of method step S2 provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of the specific process of step S3 in the embodiment of the present invention; Figure 6 This is a schematic diagram of the specific process of the first part of method step S4 provided in the embodiments of the present invention; Figure 7 This is a schematic flowchart of the second part of method step S4 provided in the embodiments of the present invention; Figure 8 This is a schematic flowchart of the method step S45 provided in an embodiment of the present invention; Figure 9 The standard deviation profile of the measured reflectivity factor Z provided in this embodiment of the invention; Figure 10 The measured differential reflectivity factor Z provided in this embodiment of the invention DR Standard deviation profile plot; Figure 11 The measured correlation coefficient ρ provided for the embodiments of the present invention hv Standard deviation profile plot; Figure 12 Measured differential propagation phase shift provided in the embodiments of the present invention Standard deviation profile plot; Figure 13 This is a planar position display diagram of S-band weather radar zenith observation data provided in an embodiment of the present invention. Detailed Implementation

[0024] To better explain and facilitate understanding of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0025] like Figure 1 As shown in the embodiment of the present invention, a zenith rotating dual-polarization radar calibration and diagnostic method includes: responding to the coordinated satisfaction of preset triggering conditions by the acquired environmental detection parameters and meteorological echo characteristics, controlling the radar antenna to perform vertical pointing towards the zenith; driving the radar antenna to perform azimuth rotation scanning in the vertical pointing state, and acquiring polarization base data with azimuth full-circle distribution characteristics and radial height level characteristics based on preset sampling configuration; performing spatial domain constraint screening and physical parameter threshold cleaning on the polarization base data to filter out interference echoes and non-liquid precipitation echoes to obtain benchmark analysis samples; performing multi-dimensional modeling on the benchmark analysis samples to separate the periodic fluctuation errors caused by radar structural asymmetry, obtaining calibration feature quantities and using them for real-time compensation of polarization base data, and mapping and identifying the state of radar hardware components based on the comparison results of the standard deviation profile of the benchmark analysis samples and theoretical accuracy indicators to output diagnostic information.

[0026] First, this invention controls the radar antenna to point vertically towards the zenith by responding to the coordinated fulfillment of preset triggering conditions by the acquired environmental detection parameters and meteorological echo characteristics. This environmentally adaptive triggering logic enables the radar to automatically perform calibration logic during volume scan intervals during operational use, achieving high-frequency, real-time online monitoring without manual intervention or field auxiliary equipment. This overcomes the drawbacks of traditional calibration methods, such as discontinuity and the inability to automate inspections.

[0027] Secondly, to address azimuth asymmetry errors, this invention drives the radar antenna to perform azimuth rotation scanning in a vertical pointing state, acquiring polarization base data with azimuth full-circle distribution characteristics and radial height-level characteristics. It utilizes the physical symmetry of natural rain particles under vertical observation to construct a true end-to-end calibration path, which not only compensates for the gain difference between the transmit and receive channels but also eliminates systematic errors caused by radome wetting, aging, and antenna feed deviation, ensuring that the calibration accuracy of the differential reflectivity system deviation is stably controlled within ±0.2dB.

[0028] Next, this invention effectively filters out interference echoes and non-liquid precipitation echoes by performing spatial domain constraint screening and physical parameter threshold cleaning on the polarization basis data. This process ensures that the benchmark analysis samples used for calibration and diagnosis have extremely high physical purity, eliminating measurement deviations caused by non-target objects from the data source, and providing data support for high-precision calibration.

[0029] Furthermore, by performing multi-dimensional modeling on the benchmark analysis sample, this invention separates the periodic fluctuation errors caused by the asymmetry of the radar structure (such as the fluctuations in the polarization loss of the rotating joint and feeder with the azimuth angle). By obtaining the calibration characteristic quantity and using it for real-time compensation, this invention can offset the sinusoidal oscillation error generated by the radar azimuth scanning mechanism during continuous rotation, eliminate the influence of azimuth asymmetry on the calibration results, and further improve the dynamic calibration accuracy.

[0030] Finally, this invention establishes a real-time mapping logic between the fluctuation characteristics of the detection data and the operating status of the radar's underlying hardware by mining the standard deviation profile of the benchmark analysis samples and comparing it with theoretical accuracy indicators. This medical "CT scan-like" diagnostic method can map, identify, and output diagnostic information of underlying hardware components such as the stability of the dual-channel receiver link, the degradation of the transmitter phase-locked loop phase noise, and the decrease in system coherence based on abnormal deviations in the standard deviation, thus realizing a closed loop of calibration and hardware inspection.

[0031] Therefore, through scientific process design and multi-dimensional data mining, this invention not only improves the calibration accuracy of dual-polarization radar, but also significantly enhances the ability to accurately diagnose the degradation risk of radar underlying components.

[0032] To better understand the above technical solutions, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that the present invention can be understood more clearly and thoroughly, and that the scope of the present invention can be fully conveyed to those skilled in the art.

[0033] Specifically, embodiments of the present invention provide a calibration and diagnostic method for a zenith rotating dual-polarization radar, which includes the following steps: S1. In response to the coordinated fulfillment of preset triggering conditions by the acquired environmental detection parameters and meteorological echo characteristics, control the radar antenna to perform vertical pointing towards the zenith.

[0034] In this step, the external detection environment and atmospheric physical state are continuously sensed and identified, and trigger logic is determined in real time. This process not only monitors atmospheric dynamic safety indicators in real time, but also simultaneously assesses the physical evolution characteristics of precipitation echoes. When the triggering conditions are met, an interrupt command is automatically issued to the radar servo control unit, ending the current routine volume scan mode (such as precipitation detection mode VCP 21D or clear sky detection mode VCP 31D), and driving the antenna beam to point at a 90° elevation angle towards the zenith, preparing a standard vertical detection environment for subsequent calibration and diagnostic procedures.

[0035] Furthermore, such as Figure 2 As shown, step S1 includes: S11. Real-time acquisition of environmental detection parameters and meteorological echo characteristics, wherein the environmental detection parameters include ground wind speed and wind speed at a preset altitude, and the meteorological echo characteristics include reflectivity factor distribution.

[0036] Specifically, in this embodiment of the invention, ground wind speed is obtained by real-time access to data from ground automatic weather stations via a communication interface; at the same time, radial velocity observations at preset height levels are directly extracted using Doppler velocity data generated by conventional volume scanning to obtain the corresponding upper-air horizontal wind speed, thereby constructing a complete dynamic environment profile covering the ground surface to the upper atmosphere.

[0037] Meanwhile, the meteorological echo characteristics were extracted from multiple dimensions, with a focus on obtaining the spatial distribution characteristics of the reflectivity factor Z in conventional volume scan data, providing basic data support for subsequent physical symmetry assumptions.

[0038] S12. Compare the ground wind speed and the wind speed at the preset height with the preset wind speed threshold respectively, and determine the period of vertical fall of precipitation particles when both the ground wind speed and the wind speed at the preset height are less than the preset wind speed threshold.

[0039] It is important to emphasize that this step ensures the physical symmetry of precipitation particles through strict wind field environmental constraints. In a still atmosphere, raindrops are affected by the balance of surface tension, hydrostatic pressure, and aerodynamic pressure. Their shape changes from a sphere to an oblate spheroid as their diameter increases, and their axis of symmetry always remains vertical. To utilize this natural physical property, a preset wind speed threshold is set to 3 m / s. By comparing the wind speed at the ground and at a preset low altitude, when both are less than this threshold, it can be ensured that the raindrops do not tilt significantly during their vertical descent. This ensures that the axis of symmetry of the precipitation particles coincides with the vertical observation axis of the radar, perfectly conforming to the physical symmetry assumption of vertical pointing observation.

[0040] S13. During the vertical descent of precipitation particles, meteorological echoes with intensities within a preset range are selected from the reflectivity factor distribution as candidate echoes. To ensure the purity of the calibration benchmark, precipitation types are further differentiated, with stratiform cloud precipitation or light rain selected as calibration targets. Specifically, meteorological echoes with reflectivity factor Z consistently between 15 dBZ and 35 dBZ are selected as candidate echoes. The scientific basis for setting this range is that if the echo is too strong (e.g., greater than 40 dBZ), asymmetric large particles such as hail may be mixed in the precipitation, disrupting the symmetry assumption; if the echo is too weak (e.g., less than 10 dBZ), the signal-to-noise ratio will be too low, leading to increased measurement errors. Through precise selection of this intensity, the stability of the target object characteristics is ensured.

[0041] S14. Identify the zero-degree layer height in the reflectivity factor distribution. Combined with the preset antenna far-field region, delineate a pure liquid precipitation region located above the lower limit of the antenna far-field region and below the zero-degree layer height in the spatial distribution of candidate echoes. To further eliminate interference from non-liquid precipitation (such as ice crystals and snowflakes) on the calibration, the height Z of the zero-degree layer bright band (Melting Layer) in the conventional volume scan data needs to be accurately identified. m Meanwhile, considering that the beam in the near-field region of the antenna is not yet fully formed and the measurement results are not representative, it is necessary to combine the antenna diameter D and the radar operating wavelength λ to calculate the lower limit of the antenna distance in the far-field region, L = (2 × D). 2 ) / λ.

[0042] The final defined effective observation interval must simultaneously meet the following conditions: above the lower limit L of the far-field region and below the zero-degree layer height Z. m This ensures that the observed target is entirely within a pure liquid precipitation area. When the radar antenna is pointed vertically to the zenith (90-degree elevation angle), regardless of the size or shape of the raindrops, the radar beam illuminates the raindrops directly below, and the observed cross-section of the raindrops is circular in both horizontal and vertical polarization. Under ideal conditions, the horizontal polarization reflectivity factor Z is... H With vertical polarization reflectivity factor Z V Theoretically, they are completely equal. Based on the differential reflectivity factor Z...DR Definition of Z DR It equals 10 multiplied by Z with base 10 H With Z V The logarithm of the ratio, therefore pointing downwards at the zenith (Z). DR The theoretical value should be strictly 0 dB. This step defines the deviation between the actual radar measurements and the theoretical 0 dB value within the pure liquid precipitation area.

[0043] S15. A calibration task window is constructed using the defined pure liquid precipitation area. In response to the opening of the calibration task window, the current radar detection task is paused and the scanning parameters are saved. Subsequently, a control command containing the working mode switching logic is generated and sent to the servo unit. The servo unit drives the radar antenna to adjust to a vertically oriented zenith operating state. After the above triggering conditions are met, the current volume scan task is automatically paused, and the radar is controlled by the servo unit to enter a new operating state. Figure 3 The zenith scanning mode is shown. This mechanism enables real-time online calibration during operation without manual intervention or the need for field equipment.

[0044] S2. Drive the radar antenna to perform azimuth rotation scanning in the vertical pointing state, and collect polarization base data with azimuth full-circumference distribution characteristics and radial height level characteristics based on the preset sampling configuration.

[0045] Furthermore, such as Figure 4 As shown, step S2 includes: S21. Maintain the radar antenna in a vertical pointing position and drive it to perform a full or multiple rotational scan in the azimuth direction at a preset rotation speed. Control the radar antenna to point vertically at 90° (zenith position) and lock the elevation angle. Initiate continuous low-speed rotation within the azimuth range of 0° to 360° (or set to multiple rotations, such as 0° to 720°). To acquire a large number of near-ocean samples covering all azimuth angles in a short time, the antenna rotation speed should be as fast as possible. Calculate and preset the scanning speed according to the current scanning mode parameters, ensuring the rotation speed does not exceed the radar's maximum allowable stable rotation speed (e.g., 36° / s). This design is not only for efficiency, but its core physical purpose is to use spatial symmetry to offset the sinusoidal periodic fluctuation errors caused by the radar feed support structure, waveguide rotation joint coaxiality error, and imperfect polarization isolation through 360° omnidirectional rotational spatial sampling, thereby obtaining a more robust estimate than static observation.

[0046] S22. The control radar transmits detection waveforms at a preset pulse repetition frequency to ensure that the number of independent samples within each azimuth step interval reaches a preset statistical significance threshold during continuous rotation scanning of the radar antenna. Specific waveform parameter configurations are used to set the pulse repetition frequency (PRF) to a high PRF mode above 1000Hz. The introduction of the high PRF mode aims to obtain an extremely high number of independent samples, ensuring that the number of effective samples within each azimuth step interval meets the statistical significance requirement (e.g., no less than 50 points). This high-density sampling effectively smooths the random fluctuations of precipitation echoes, providing a stable data substrate for subsequent high-precision standard deviation calculations.

[0047] S23. During the continuous rotating scan and waveform transmission of the radar antenna, the raw echo signals corresponding to the full azimuth angle and the full range database are acquired in real time. During the azimuth rotation, the raw base data stream of the full azimuth angle and the full range database (i.e., the set of all continuous sampling points from the near field to the maximum detection range when the radar is pointing vertically) is recorded in real time. Even if the wind speed is limited during the triggering phase, small turbulence may still cause raindrops to tilt slightly and deviate from the spherical symmetry assumption; by recording this omnidirectionally distributed signal, the positive and negative deviations caused by the tilting of raindrops at different azimuth angles can be effectively canceled by the rotating averaging strategy in subsequent processing, thereby eliminating the uncertainty caused by the fluctuation of the microphysical properties of precipitation particles.

[0048] S24. Utilizing the linear mapping relationship between the full detection range library and vertical height under vertical pointing conditions, the original echo signal is converted into polarization-based data with azimuth-wide distribution characteristics and radial height-level characteristics. Since the antenna is vertically pointed, the radial range library directly corresponds linearly to the vertical height. The acquired raw data is then analyzed into a differential reflectivity factor Z. DR Reflectivity factor Z, differential propagation phase shift and correlation coefficient ρ hv Equal base data samples provide a foundation for subsequent processing.

[0049] S3. Spatial domain constraint screening and physical parameter threshold cleaning are performed on the polarization basis data to filter out interference echoes and non-liquid precipitation echoes, and obtain the benchmark analysis sample.

[0050] Furthermore, such as Figure 5 As shown, step S3 includes: S31. The polarization basis data is constrained within a pure liquid precipitation region defined in a vertical pointing state. The lower boundary limit of the pure liquid precipitation region is used to eliminate the blind zone and near-field interference introduced by the transient delay of the transceiver switch. The upper boundary limit of the pure liquid precipitation region is used to eliminate the echo interference of non-spherical ice phase particles above the zero-degree layer, thus obtaining the polarization basis data that has been spatially constrained and filtered.

[0051] In this step, items with a value of 0 to (2×D) are automatically removed. 2 Near-field data within the range of λ / 2 is used to eliminate residual delay time intervals between transmit and receive switches and near-field coupling interference from the antenna. Specifically, considering that even when the radar main beam is vertically pointed towards the zenith, its low-gain sidelobes may still receive scattered signals from the ground, this step employs a strict altitude filter mechanism (e.g., removing near-field echoes below 2km) to effectively eliminate the influence of ground clutter. Simultaneously, data from the zero-degree layer height Z is removed. m The above non-spherical ice phase particle echoes, therefore, preferably (2×D) 2 ) / λ to Z m The upper layer serves as the core analysis area, ensuring that the signals involved in the calibration originate entirely from pure meteorological echoes.

[0052] S32. The signal-to-noise ratio (SNR), signal quality index (SQI), and correlation coefficient of the polarization basis data filtered by spatial domain constraints are compared with a preset set of physical quantity thresholds to remove noise echoes and non-meteorological target interference, resulting in polarization basis data cleaned by physical parameter thresholds. To ensure signal purity, multiple threshold cleaning is performed on the spatially filtered data. The hard filtering thresholds are set as follows: signal-to-noise ratio (SNR) > 20dB, signal quality index (SQI) > 0.8, and correlation coefficient (ρ) > 0.8. hv Hard thresholds such as >0.98 are used to effectively eliminate noise fluctuations and random interference from non-meteorological targets, ensuring the consistency of the baseline sample in statistical characteristics.

[0053] S33. Count the total number of polarization basis data samples after physical parameter threshold cleaning, and determine whether the total number of samples reaches the preset minimum calibration sample size threshold. Count the total number of valid samples after filtering in real time, and determine whether it reaches the preset minimum calibration sample size threshold (e.g., not less than 1000 points). The theoretical basis for this threshold setting is that the uncertainty of calibration accuracy is inversely proportional to the square root of the number of independent samples M (i.e., ...). By using the high pulse repetition frequency sampling and omnidirectional rotation mode described in the preceding steps, it is possible to ensure that a predetermined number of samples are obtained within each height level.

[0054] S34. In response to the total number of samples reaching a preset minimum calibration sample size threshold, the polarization basis data, after being cleaned by physical parameter thresholds, is used as the benchmark analysis sample for calibration diagnosis. When the number of effective independent samples reaches more than 1000 points, the statistical error caused by random signal fluctuations will be much smaller than the 0.2dB calibration accuracy requirement of radar differential reflectivity. Through this multi-dimensional quality control and sample point assurance mechanism, this invention can obtain highly representative and physically pure benchmark analysis samples without interrupting operations.

[0055] S4. Perform multi-dimensional modeling on the benchmark analysis sample to separate the periodic fluctuation error caused by the asymmetry of the radar structure, obtain the calibration characteristic quantity and use it for real-time compensation of the polarization basis data, and map and identify the status of the radar hardware components based on the comparison results of the standard deviation profile of the benchmark analysis sample and the theoretical accuracy index to output diagnostic information.

[0056] Furthermore, such as Figure 6 As shown, step S4 involves multidimensional modeling of the benchmark analysis samples to separate the periodic fluctuation error caused by the radar structural asymmetry, obtaining calibration characteristic quantities, and using them for real-time compensation of polarization basis data. This includes: S41. Perform noise reduction or outlier removal on the differential reflectance factor values ​​in the benchmark analysis sample, and obtain the system bias estimate through statistical calculation. Specifically, for all valid observation points within the omnidirectional angle, the differential reflectance factor Z... DR After the observations have undergone denoising or outlier removal, they are then subjected to arithmetic averaging or median filtering to obtain the system bias estimate. This is intended to filter out residual random impulse noise from massive samples, providing a robust initial search point for subsequent periodic error decoupling model fitting, thereby accelerating model convergence and avoiding getting trapped in local optima.

[0057] S42. Perform binning statistics on the benchmark analysis samples in the azimuth domain according to a preset angular resolution, and construct confidence evaluation indicators for each azimuth binning by combining the reflectivity factor, signal-to-noise ratio, correlation coefficient, signal quality index and sample quantity contained in the benchmark analysis samples.

[0058] S43. Based on the confidence evaluation index and the system bias estimate, the periodic error decoupling model is iteratively solved to achieve confidence constraints on the binned samples of different azimuth angles during the parameter fitting process. After the periodic error decoupling model converges, the solution result of the DC bias component is obtained. The fitting objective of the periodic error decoupling model includes the DC bias component to be solved, as well as the multi-order harmonic disturbance component adaptively determined based on the convergence of the fitting residual, the model complexity constraint, and the energy ratio of the periodic component.

[0059] Considering factors such as the feed support structure of the radar antenna, the coaxiality error of the waveguide rotation joint, imperfect polarization isolation, and local mechanical eccentricity during rotation, the observed values ​​will exhibit complex fluctuations with varying antenna azimuth angles, exceeding a single period. Therefore, this embodiment represents the differential reflectivity factor sequence as a circular periodic model composed of a DC bias term, multiple harmonic disturbance terms, and residual terms: ; Where az is the azimuth angle, Z DR(az) represents the observed sequence of differential reflectivity factors varying with antenna azimuth angle, B0 is the DC bias component, k is the harmonic order, and a k With b k Let a be the coefficient of the kth harmonic component. k With b k The numerical values ​​are obtained by fitting the benchmark analysis samples. N is the adaptively selected harmonic order, and ε(az) is the residual term, representing the remaining statistics after model fitting. The harmonic order N is adaptively determined based on the convergence of the fitting residuals, model complexity constraints, and the energy proportion of the periodic components to avoid underfitting or overfitting. Through the periodic error decoupling model, the multi-period coupling disturbances caused by radar structural asymmetry can be decoupled from the actual system deviation in mathematical space, while retaining the amplitude and phase information of each periodic component as subsequent diagnostic features.

[0060] In the specific solution process of the model, the benchmark analysis samples are first divided into bins and statistically analyzed in the azimuth domain according to a preset angular resolution. Then, the confidence evaluation index for each azimuth bin is constructed by combining the reflectivity factor, signal-to-noise ratio, correlation coefficient, signal quality index and sample number contained in the benchmark analysis samples in each bin.

[0061] Subsequently, the periodic error decoupling model is iteratively solved based on the confidence evaluation index and the previously obtained system deviation estimate. This confidence evaluation index is used to constrain the confidence of bin samples at different azimuth angles during the parameter fitting process. After the periodic error decoupling model converges, the calculation results of the DC bias component B0 and the harmonic coefficients of each order are obtained.

[0062] S44. The solution results are used as the differential reflectivity system deviation after removing multi-period composite fluctuation errors. At the same time, the amplitude, phase and energy ratio of multi-order harmonic disturbance components in the periodic error decoupling model are extracted as auxiliary diagnostic information to characterize the state of radar hardware structure.

[0063] Furthermore, the fitted circular periodic model is subjected to parameter analysis, and its DC bias component B0 is extracted as the final calibration feature. This B0 value represents the pure system deviation after eliminating the compound periodic errors caused by the asymmetry of the rotating structure, the periodic loss fluctuation of the polarization link, and the local azimuth disturbance.

[0064] Simultaneously, the amplitude, phase, and energy percentage of each harmonic component are jointly extracted, where the amplitude and phase of the k-th harmonic can be expressed as follows: ; ; S45. Convert the differential reflectivity system deviation into a calibration compensation constant and feed it back in real time to the compensation parameter table of the preset signal processor for online automatic calibration of polarization basis data.

[0065] Therefore, the DC component B0 can be used as the basis for calibration compensation and written into the compensation parameter table of the signal processor in real time to complete online automatic compensation. Meanwhile, the amplitude and phase characteristics of multi-order harmonics and residual statistics can be used as auxiliary diagnostic parameters to characterize feed support structure disturbances, rotary joint eccentricity, polarization isolation degradation or local directional anomalies. This achieves integrated processing of "calibration parameter extraction" and "mechanical / electrical condition diagnosis", improving the robustness of calibration results and the ability to identify system anomalies.

[0066] Because this invention uses zenith-pointing natural particles as a reference source, this calibration method not only covers the internal electrical signal links of the transmitter and receiver, but also incorporates all factors across the entire link, such as antenna radiation performance, dual-channel gain difference, radome water accumulation or aging, and feed thermal deformation. This dynamic update mechanism ensures that the radar differential reflectivity maintains a detection accuracy within 0.2 dB throughout long-term operation, significantly improving the accuracy of quantitative precipitation estimation.

[0067] Furthermore, in step S4, based on the comparison results between the standard deviation profile of the benchmark analysis sample and the theoretical accuracy index, the status of the radar's hardware components is mapped and identified to output diagnostic information, such as... Figure 7 As shown, it includes: Furthermore, such as Figure 8 As shown, step S45 includes; S451. Calculate the correlation coefficients of the benchmark analysis samples within each altitude level sampling window, and combine this with the acquired radar antenna rotation speed, pulse repetition frequency, and dwell time to calculate the correlation time of the pulse sequence within each altitude level sampling window. Within an extremely short scan period (e.g., within 1 minute), it can be assumed that the precipitation system is in a statistically stationary state. Based on this assumption, calculate the correlation coefficients ρ of the benchmark analysis samples within each altitude level sampling window. hv By combining the radar antenna rotation speed, pulse repetition frequency, and dwell time, the correlation time of the pulse sequence is calculated. The correlation time reflects the rate of signal change over time and is a core parameter for decoupling the influence of sampling capacity.

[0068] S452. Correct the total number of samples within each height level sampling window using correlation time, calculate the number of independent samples corresponding to the differential reflectivity factor and reflectivity factor, and calculate the equivalent number of samples corresponding to the differential propagation phase shift and correlation coefficient. Correct the total number of samples using correlation time. Specifically, calculate the number of independent samples corresponding to the reflectivity factor Z and differential reflectivity factor Z. DR Independent samples M1, and corresponding correlation coefficient ρhv Phase shift with differential propagation The equivalent sample size M' is obtained. By finely correcting the sample size, the interference of signal coherence caused by the low-speed rotation of the antenna on the statistical accuracy is eliminated.

[0069] S453. The correlation coefficient, number of independent samples, and number of equivalent samples within each height level sampling window are used as feature variables and substituted into the standard deviation analytical model constructed based on the parameter statistical coherence characteristics. By decoupling the influence of signal coherence time on the sampling capacity, the measured standard deviation of each parameter corresponding to each height level sampling window is obtained in real time.

[0070] ρ hv M1 and M' are used as feature variables and substituted into a pre-defined standard deviation analytical model. This invention constructs four types of diagnostic models based on the parametric statistical coherence characteristics, each corresponding to a different hardware evaluation dimension: (1) Differential reflectivity standard deviation used to evaluate dual-channel gain stability Computational model: ; (2) Differential propagation phase shift standard deviation used to evaluate the inter-pulse coherence of the transmitter and the stability of the phase-locked loop. Computational model: ; (3) Standard deviation of correlation coefficient used to evaluate system coherence and channel consistency Computational model: ; (4) Standard deviation of reflectivity factor used to evaluate the stability of the entire system, including transmitters and receivers, under horizontal polarization. Computational model: ; S454. The measured standard deviations of each parameter are processed continuously according to vertical height order to generate measured standard deviation profiles including the differential reflectivity factor standard deviation profile, the differential propagation phase shift standard deviation profile, the correlation coefficient standard deviation profile, and the reflectivity factor standard deviation profile. The standard deviations are processed continuously according to vertical height order to generate measured standard deviation profiles. For example... Figures 9-12 The figures shown are the measured reflectance factor Z and the differential reflectance factor Z, respectively. DR Correlation coefficient ρ hv and differential propagation phase shift The standard deviation profile of , where, Figures 9-12 In the diagram, the vertical axis represents height in meters (m), and the horizontal axis represents the standard deviation (dBZ) and differential reflectance factor Z, respectively. DRStandard deviation (in dB), correlation coefficient ρ hv Standard deviation (unitless) and differential propagation phase shift The standard deviation (in degrees) is then calculated. These profiles are subsequently compared in real time with preset theoretical limits for altitude variation, and the degree of deviation is used to quantitatively assess the health status of the radar hardware.

[0071] S46. Calculate the measured standard deviation profiles of each parameter's distribution with height in the benchmark analysis sample. These measured standard deviation profiles include the differential reflectivity factor standard deviation profile, the differential propagation phase shift standard deviation profile, the correlation coefficient standard deviation profile, and the reflectivity factor standard deviation profile. Calculate the measured standard deviation profiles of each parameter's distribution with height in the benchmark analysis sample and compare them with the theoretical standard deviation profiles corresponding to each parameter.

[0072] S47. Compare each profile in the measured standard deviation profile with the corresponding preset theoretical standard deviation profile, and analyze the deviation characteristic value of the measured standard deviation profile relative to the corresponding theoretical standard deviation profile.

[0073] S48. In response to the deviation of the eigenvalue characterizing the differential reflectivity factor standard deviation profile being higher than the corresponding theoretical standard deviation profile in the pure liquid precipitation region, the stability of the radar's dual-channel receiving link is determined to have decreased, and the first diagnostic information is output. In response to the deviation of the eigenvalue characterizing the differential propagation phase shift standard deviation profile exhibiting abnormal fluctuations exceeding the preset fluctuation range relative to the corresponding theoretical standard deviation profile, the phase noise of the radar's transmitter phase-locked loop is determined to have deteriorated or the radar coherence is reduced, and the second diagnostic information is output. In response to the deviation of the eigenvalue characterizing the correlation coefficient standard deviation profile being higher than the corresponding theoretical standard deviation profile in the pure liquid precipitation region, or to a discrete change exceeding the preset jump threshold between adjacent height layers, the consistency of the radar's dual polarization channels is determined to have decreased, the polarization isolation is deteriorated, or the performance of the feed and polarization separation network is abnormal, and the third diagnostic information is output. In response to the deviation of the eigenvalue characterizing the reflectivity factor standard deviation profile deviating from the corresponding theoretical standard deviation profile, the stability of the radar is determined to have changed, and the fourth diagnostic information is output.

[0074] (1) Considering the current signal-to-noise ratio, if Z DR If the measured standard deviation is significantly higher than the theoretical value (e.g., a sudden increase to more than 0.3dB, or a sustained increase to more than 0.2dB), it is identified as a deterioration in the stability of the dual-channel receiving link, which may indicate a risk of low-noise amplifier gain jitter, dual-channel gain imbalance, or increased temperature drift in the receiving link.

[0075] (2) If Abnormal fluctuations where the measured standard deviation exceeds the preset fluctuation range are identified as degradation of the transmitter phase-locked loop phase noise, decrease in inter-pulse coherence, or deterioration of system phase stability.

[0076] (3) If the correlation coefficient ρ hv If the measured standard deviation is significantly higher than the theoretical value, or if there is an unreasonable increase or abrupt change in the dispersion along the height direction (abrupt change in dispersion exceeding the preset jump threshold), then the mapping is identified as a decrease in the consistency of the dual polarization channels. This may be due to factors such as feed orthogonality deviation, deterioration of polarization isolation, abnormal performance of the polarization separation network, and imbalance of dual-channel amplitude matching, which may introduce additional depolarization effects.

[0077] (4) If the measured standard difference of reflectivity factor Z is abnormal, it is determined that the radar transmit / receive power stability has changed, the overall amplitude stability has decreased, or the system gain fluctuation has increased.

[0078] S49. By fusing differential reflectivity system bias, first diagnostic information, second diagnostic information, third diagnostic information, fourth diagnostic information, and auxiliary diagnostic information, a radar health status inspection report is generated. In specific implementation, the aforementioned decoupling-obtained DC bias component B0 and auxiliary diagnostic information (including the harmonic amplitudes A of each order of the periodic error decoupling model) are comprehensively utilized. k With phase It is used to assess mechanical structural deformation and asymmetric losses, as well as the first to fourth diagnostic information characterizing the status of electrical links, and generates fully automated online calibration and health inspection reports.

[0079] This multidimensional diagnostic system enables integrated processing of calibration parameter extraction and mechanical / electrical condition diagnosis. Because this invention utilizes zenith-pointing natural precipitation particles as a reference source, it eliminates the need for manual installation of auxiliary equipment such as standard metal spheres, achieving full automation and end-to-end coverage of the calibration process. This significantly improves the accuracy of quantitative precipitation estimation and reduces operation and maintenance costs.

[0080] In addition, this invention provides a zenith-rotating dual-polarization radar calibration and diagnostic system, comprising: a zenith detection control module, used to control the radar antenna to perform vertical pointing towards the zenith in response to the coordinated satisfaction of preset triggering conditions by acquired environmental detection parameters and meteorological echo characteristics; a scanning sampling execution module, used to drive the radar antenna to perform azimuth rotation scanning in the vertical pointing state, and to collect polarization base data with azimuth full-circumference distribution characteristics and radial height level characteristics based on preset sampling configuration; a sample cleaning and quality control module, used to perform spatial domain constraint screening and physical parameter threshold cleaning on the polarization base data, filter out interference echoes and non-liquid precipitation echoes, and obtain benchmark analysis samples; and a calibration and diagnostic integration module, used to perform multi-dimensional modeling on the benchmark analysis samples to separate the periodic fluctuation errors caused by radar structural asymmetry, obtain calibration feature quantities and use them for real-time compensation of polarization base data, and map and identify the status of radar hardware components based on the comparison results of the standard deviation profile of the benchmark analysis samples and theoretical accuracy indicators to output diagnostic information.

[0081] In a preferred embodiment of the present invention, the zenith rotating dual-polarization radar calibration and diagnostic system is primarily based on an S-band (operating wavelength typically 10 cm) dual-polarization Doppler weather radar system. Its core principles are also applicable to dual-polarization radars operating at other frequencies, such as C-band or X-band. The system first relies on a radar host with dual-polarization measurement capabilities, such as a CINRAD / S-type radar or a phased array weather radar. This host can simultaneously transmit and receive horizontally and vertically polarized electromagnetic waves and is equipped with a high-performance signal processor for real-time calculation and output of polarization basis data streams, including reflectivity factor, differential reflectivity factor, correlation coefficient, and differential propagation phase shift. Simultaneously, the system's antenna and servo system possess high-precision pointing control capabilities, supporting vertical zenith pointing at an elevation angle of 90° and continuous, stable, low-speed rotation scanning from 0° to 360° or multiple rotations in the azimuth direction. This ensures the accuracy of azimuth pointing and the symmetry of physical sampling during rotational sampling.

[0082] Specifically, to realize the functions of the zenith detection control module, the system software platform integrates a meteorological condition monitoring unit. This unit receives real-time reflectivity factor distribution, Doppler wind field data, and zero-degree layer height information provided by numerical weather prediction or radiosonde stations from the radar's conventional volume scan output. It also couples measured wind speeds from ground meteorological stations to collaboratively determine whether the atmospheric environment meets the preset automated triggering conditions of this invention. Once the triggering conditions are met, the scanning sampling execution module automatically instructs the antenna servo system to switch to zenith calibration mode via the scanning control unit. It then performs a full-circle continuous scan according to the preset sampling configuration (e.g., PRF > 1000Hz and a specific rotation speed), and utilizes a high-capacity data acquisition and storage unit to record the omnidirectional and full-range database raw data stream, including timestamps, azimuth angles, and range database information. Figure 13 As shown, the collected raw data can be represented as follows: with the radar as the center, the radius representing height, and the color representing Z. DR The Planar Position Display (PPI) plot visually illustrates the original state of the polarization basis data of precipitation particles at each altitude level as a function of azimuth under zenith observation mode.

[0083] Building upon this foundation, the sample cleaning and quality control module performs refined cleaning on the collected data. Through spatial domain constraints such as near-field blind zone removal and ice-phase particle removal above the zero-degree layer, coupled with threshold filtering of physical parameters such as signal-to-noise ratio, signal quality index, and correlation coefficient, it eliminates interference from ground clutter, transmit / receive delays, and non-meteorological echoes at the source, thereby obtaining a physically pure benchmark analysis sample. Subsequently, the calibration and diagnostic integration module performs multi-dimensional modeling on the benchmark analysis sample and decouples the periodic fluctuation errors caused by radar structural asymmetry, extracting the DC component as the final system bias.

[0084] Next, the statistical distribution characteristics of the benchmark analysis samples are deeply explored. The measured standard deviation profiles of each parameter are calculated and compared with the theoretical profiles constructed based on signal coherence characteristics. By analyzing the deviation characteristics of the measured profiles relative to the theoretical benchmark, the component status of the radar's underlying hardware can be accurately mapped and identified. For example, anomalies in the differential reflectivity standard deviation can determine the stability of the dual-channel receiver link and low-noise amplifier, or fluctuations in the differential propagation phase shift standard deviation can identify the degradation of the transmitter's phase-locked loop phase noise. Finally, the results display and early warning module visualizes the calibration results and diagnostic conclusions, and the calibration parameter storage and update module compensates for the calculated system deviations in real time within the signal processing flow, forming a fully automated, end-to-end closed-loop operation and maintenance system within the intervals of conventional volume scans.

[0085] Then, an embodiment of the present invention provides a zenith rotating dual polarization radar calibration and diagnostic device, which includes: at least one controller; and a memory communicatively connected to the at least one controller; wherein the memory stores instructions that can be executed by the at least one controller, and the instructions are executed by the at least one controller to enable the at least one controller to perform the zenith rotating dual polarization radar calibration and diagnostic method as described above.

[0086] Furthermore, embodiments of the present invention provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a controller, implement the zenith rotating dual-polarization radar calibration and diagnostic method as described above.

[0087] In summary, the embodiments of the present invention provide a zenith rotating dual polarization radar calibration and diagnostic method, system, device, and medium. Through in-depth analysis and improvement of the prior art, the present invention has achieved significant results in terms of technological advancement, calibration accuracy, and intelligent system operation and maintenance.

[0088] This invention takes real-time perception and identification of the detection environment as its logical starting point. By dynamically evaluating wind field stability and echo microphysical characteristics, it ensures accurate capture of the most suitable calibration window within the intervals of routine operational body scans. Subsequently, mode switching drives the antenna to perform a vertically pointing, omnidirectional, continuous rotational scan, acquiring a raw polarization data stream with high spatial resolution and high statistical independence. In data processing, this invention integrates multi-dimensional quality control and spatial domain purification algorithms to eliminate near-field interference and non-meteorological factors at the source, ensuring the physical purity of the calibration benchmark. Next, using multi-dimensional mathematical modeling and sine fitting techniques, it achieves precise decoupling between system deviation and mechanical structural fluctuation errors, and feeds the extracted calibration constants back to the signal processing terminal in real time for automatic compensation. Finally, through dynamic mapping of the measured standard deviation profile and the theoretical limit benchmark, a health check of the underlying hardware is completed, thus forming an intelligent radar closed-loop operation and maintenance system with self-perception, self-calibration, and self-diagnosis capabilities. It produces the following technical effects: First, this invention fully utilizes the physical symmetry of rain particles in natural precipitation to achieve true end-to-end calibration across the entire link. This method not only covers the electronic link between the transmitter and receiver but also includes key physical components such as the radome, feed, and antenna reflector, ensuring that the calibration accuracy of the differential reflectivity system deviation can be stably controlled within ±0.2 dB.

[0089] Secondly, this invention overcomes the limitations of traditional static pointing calibration by introducing a 360-degree azimuth continuous rotation scanning strategy, combined with a constructed circular periodic model, to accurately eliminate periodic mechanical errors caused by joint coaxiality errors and feeder polarization loss asymmetry. This spatially symmetrical sampling mechanism greatly improves the robustness of the calibration results and solves the problem of inconsistent observations at different azimuth angles in traditional methods.

[0090] Furthermore, this invention innovatively taps into the statistical potential of zenith scan data. By calculating the standard deviation profiles of each polarization basis data, a real-time mapping logic is established between meteorological echo fluctuation characteristics and the health status of the radar's underlying hardware. This online inspection method, similar to a medical CT scan, can achieve automatic early warning of dual-channel receiver link stability, transmitter phase-locked loop phase noise, and system coherence degradation without interrupting operations, providing scientific data support for the predictive maintenance of modern radar.

[0091] In summary, this invention features a scientific process, requires no additional field auxiliary equipment, and boasts a high degree of automation. It not only significantly improves the accuracy of quantitative precipitation estimation using dual-polarization radar but also greatly reduces the operation and maintenance costs of unattended radar sites. This method is highly suitable for large-scale application in operational weather radar networks and represents an efficient and intelligent technical means to enhance radar observation quality and hardware management.

[0092] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0093] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions.

[0094] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the claims should be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention. Clearly, those skilled in the art can make various modifications and variations to the invention without departing from its spirit and scope. Thus, if these modifications and variations of the invention fall within the scope of the claims and their equivalents, the invention should also include these modifications and variations.

Claims

1. A method for calibration and diagnosis of a zenith rotating dual-polarization radar, characterized in that, include: In response to the coordinated fulfillment of preset triggering conditions by the acquired environmental detection parameters and meteorological echo characteristics, the radar antenna is controlled to perform an operation of pointing vertically towards the zenith; In the vertical pointing state, the radar antenna is driven to perform azimuth rotation scanning, and polarization base data with azimuth full-circumference distribution characteristics and radial height level characteristics are collected based on the preset sampling configuration. Spatial domain constraint screening and physical parameter threshold cleaning were performed on the polarization basis data to filter out interference echoes and non-liquid precipitation echoes, and obtain benchmark analysis samples. Multidimensional modeling is performed on the benchmark analysis samples to separate the periodic fluctuation error caused by the asymmetry of the radar structure, obtain calibration characteristic quantities and use them for real-time compensation of polarization basis data, and map and identify the status of the radar hardware components based on the comparison results of the standard deviation profile of the benchmark analysis samples and the theoretical accuracy index to output diagnostic information.

2. The zenith rotating dual-polarization radar calibration and diagnostic method as described in claim 1, characterized in that, In response to the coordinated fulfillment of preset triggering conditions by the acquired environmental detection parameters and meteorological echo characteristics, the radar antenna is controlled to perform an operation of pointing vertically towards the zenith, including: Real-time acquisition of environmental detection parameters and meteorological echo characteristics, including ground wind speed and wind speed at a preset altitude, and meteorological echo characteristics including reflectivity factor distribution; The ground wind speed and the wind speed at a preset height are compared with preset wind speed thresholds, and the period of vertical fall of precipitation particles is determined when both the ground wind speed and the wind speed at a preset height are less than the preset wind speed threshold. During the vertical fall of precipitation particles, meteorological echoes with intensity within a preset intensity range are selected from the reflectivity factor distribution as candidate echoes. Identify the zero-degree layer height in the reflectivity factor distribution, and combine it with the preset antenna far-field region to delineate a pure liquid precipitation area located above the lower limit of the antenna far-field region and below the zero-degree layer height in the spatial distribution of candidate echoes. A calibration task window is constructed using a defined pure liquid precipitation area. In response to the opening of the calibration task window, the radar's current detection task is paused and the scanning parameters are saved, driving the radar antenna to adjust to a state of vertical pointing towards the zenith.

3. The zenith rotating dual-polarization radar calibration and diagnostic method as described in claim 2, characterized in that, In vertical pointing mode, the radar antenna is driven to perform azimuth rotation scanning, and polarization basis data with azimuth full-circle distribution characteristics and radial height level characteristics are collected based on a preset sampling configuration, including: Keep the radar antenna in a vertical pointing state and drive the radar antenna to perform a full or multiple rotational scan in the azimuth direction at a preset rotation speed; The radar is controlled to transmit detection waveforms at a preset pulse repetition frequency so that the number of independent samples in each azimuth step interval reaches a preset statistical significance threshold during the continuous rotation scanning of the radar antenna. During the continuous rotating scan and detection waveform transmission of the radar antenna, the original echo signals corresponding to the full azimuth angle and full detection range are acquired in real time. By utilizing the linear mapping relationship between the full detection range library and the vertical height under vertical pointing conditions, the original echo signal is converted into polarization-based data with azimuth full-circumference distribution characteristics and radial height-level characteristics; wherein, the polarization-based data includes: differential reflectivity factor, reflectivity factor, differential propagation phase shift and correlation coefficient.

4. The zenith rotating dual-polarization radar calibration and diagnostic method as described in claim 3, characterized in that, Spatial domain constraint screening and physical parameter threshold cleaning are performed on the polarization basis data to filter out interference echoes and non-liquid precipitation echoes, obtaining benchmark analysis samples, including: The polarization-based data is constrained within a pure liquid precipitation region defined in a vertically pointing state. The lower boundary limit of the pure liquid precipitation region is used to eliminate the blind zone and near-field interference introduced by the transient delay of the transceiver switch. The upper boundary limit of the pure liquid precipitation region is used to eliminate the echo interference of non-spherical ice phase particles above the zero-degree layer, thus obtaining the polarization-based data that has been spatially constrained and filtered. The signal-to-noise ratio, signal quality index, and correlation coefficient in the polarization basis data that have been filtered by spatial domain constraints are compared with a preset set of physical quantity thresholds to remove noise echoes and non-meteorological target interference, thus obtaining polarization basis data after physical parameter threshold cleaning. The total number of polarization basis data after physical parameter threshold cleaning is counted, and it is determined whether the total number of samples has reached the preset minimum calibration sample size threshold. In response to the total number of samples reaching the preset minimum calibration sample size threshold, the polarization basis data after physical parameter threshold cleaning is used as the benchmark analysis sample for performing calibration diagnosis.

5. The zenith rotating dual-polarization radar calibration and diagnostic method as described in claim 4, characterized in that, Multidimensional modeling is performed on the benchmark analysis samples to separate the periodic fluctuation error caused by the radar structural asymmetry, obtain calibration characteristic quantities, and use them for real-time compensation of polarization basis data, including: The differential reflectance factor values ​​in the benchmark analysis sample are denoised or outlier removed, and the estimated system bias is obtained through statistical calculation. The benchmark analysis samples are binned and statistically analyzed in the azimuth domain according to a preset angular resolution. The confidence evaluation index for each azimuth bin is constructed by combining the reflectivity factor, signal-to-noise ratio, correlation coefficient, signal quality index and sample number contained in the benchmark analysis samples. Based on the confidence evaluation index and the system bias estimate, the periodic error decoupling model is iteratively solved to achieve confidence constraints on bin samples with different azimuth angles during the parameter fitting process. After the periodic error decoupling model converges, the solution result of the DC bias component is obtained. The fitting objective of the periodic error decoupling model includes the DC bias component to be solved, as well as the multi-order harmonic disturbance component adaptively determined based on the convergence of the fitting residual, model complexity constraints, and the energy ratio of the periodic component. The solution results are used as the differential reflectivity system deviation after removing multi-period composite fluctuation errors. At the same time, the amplitude, phase and energy ratio of multi-order harmonic disturbance components in the periodic error decoupling model are extracted as auxiliary diagnostic information to characterize the state of radar hardware structure. The differential reflectivity system deviation is converted into a calibration compensation constant and fed back in real time to the compensation parameter table of the preset signal processor for online automatic calibration of polarization basis data.

6. The zenith rotating dual-polarization radar calibration and diagnostic method as described in claim 5, characterized in that, Based on the comparison between the standard deviation profile of the benchmark analysis sample and the theoretical accuracy index, the status of the radar's hardware components is mapped and identified to output diagnostic information, including: Calculate the measured standard deviation profiles of the distribution of each parameter with height in the benchmark analysis sample. The measured standard deviation profiles include the differential reflectivity factor standard deviation profile, the differential propagation phase shift standard deviation profile, the correlation coefficient standard deviation profile, and the reflectivity factor standard deviation profile. Each profile in the measured standard deviation profile is compared with the corresponding preset theoretical standard deviation profile, and the deviation characteristic value of the measured standard deviation profile relative to the corresponding theoretical standard deviation profile is analyzed. In response to the deviation of the eigenvalue characterizing the differential reflectivity factor standard deviation profile being higher than the corresponding theoretical standard deviation profile in the pure liquid precipitation region, the stability of the radar's dual-channel receiving link is determined to be reduced and the first diagnostic information is output. In response to abnormal fluctuations in the differential propagation phase shift standard deviation profile that deviates from the eigenvalue, which exceeds the preset fluctuation range relative to the corresponding theoretical standard deviation profile, the radar transmitter phase-locked loop phase noise is determined to be degraded or the radar coherence is reduced, and second diagnostic information is output. In response to the deviation of the eigenvalue characterization correlation coefficient standard deviation profile being higher than the corresponding theoretical standard deviation profile in the pure liquid precipitation region, or the dispersion abrupt change exceeding the preset jump threshold between adjacent height layers, the radar dual polarization channel consistency is determined to be reduced, the polarization isolation is deteriorated, or the feed and polarization separation network performance is abnormal, and third diagnostic information is output. In response to the deviation of the standard deviation profile of the reflectivity factor, which represents the deviation of the eigenvalue, from the corresponding theoretical standard deviation profile, the radar stability is determined to have changed and the fourth diagnostic information is output. By integrating differential reflectivity system deviation, first diagnostic information, second diagnostic information, third diagnostic information, fourth diagnostic information, and auxiliary diagnostic information, a radar health status inspection report is generated.

7. The zenith rotating dual-polarization radar calibration and diagnostic method as described in claim 6, characterized in that, Calculate the measured standard deviation profiles of each parameter's distribution with height in the baseline analysis sample, including: The pure liquid precipitation area is divided into multiple height-level sampling windows according to the preset height resolution, and the benchmark analysis samples mapped to each height-level sampling window are extracted. The correlation coefficients of the benchmark analysis samples within the sampling window at each altitude level are statistically analyzed, and the correlation time of the pulse sequence within the sampling window at each altitude level is calculated by combining the obtained radar antenna rotation speed, pulse repetition frequency and dwell time. The total number of samples within each height level sampling window is corrected using the relevant time, and the number of independent samples corresponding to the differential reflectivity factor and reflectivity factor is calculated. The equivalent number of samples corresponding to the differential propagation phase shift and correlation coefficient is also calculated. The correlation coefficient, number of independent samples, and number of equivalent samples within each height level sampling window are used as feature variables and substituted into the standard deviation analytical model constructed based on the parameter statistical coherence characteristics. By decoupling the influence of signal coherence time on sampling capacity, the measured standard deviation of each parameter corresponding to each height level sampling window is obtained in real time. The measured standard deviations of each parameter are processed continuously according to the vertical height order to generate a measured standard deviation profile that includes the differential reflectivity factor standard deviation profile, the differential propagation phase shift standard deviation profile, the correlation coefficient standard deviation profile, and the reflectivity factor standard deviation profile.

8. A zenith rotating dual-polarization radar calibration and diagnostic system, characterized in that, include: The zenith detection control module is used to control the radar antenna to perform vertical pointing towards the zenith in response to the coordinated satisfaction of the acquired environmental detection parameters and meteorological echo characteristics with preset triggering conditions. The scanning sampling execution module is used to drive the radar antenna to perform azimuth rotation scanning in a vertical pointing state, and to collect polarization base data with azimuth full-circumference distribution characteristics and radial height level characteristics based on a preset sampling configuration; The sample cleaning and quality control module is used to perform spatial domain constraint screening and physical parameter threshold cleaning on polarization basis data, filter out interference echoes and non-liquid precipitation echoes, and obtain benchmark analysis samples. The calibration and diagnostic integration module is used to perform multi-dimensional modeling of the benchmark analysis sample to separate the periodic fluctuation error caused by the asymmetry of the radar structure, obtain calibration characteristic quantities and use them for real-time compensation of polarization basis data, and map and identify the status of the radar hardware components based on the comparison results of the standard deviation profile of the benchmark analysis sample and the theoretical accuracy index to output diagnostic information.

9. A zenith rotating dual-polarization radar calibration and diagnostic device, characterized in that, include: At least one controller; and a memory that is communicatively connected to at least one controller; The memory stores instructions that can be executed by at least one controller, which enables the at least one controller to perform the zenith rotating dual polarization radar calibration and diagnostic method as described in any one of claims 1-7.

10. A computer-readable storage medium storing computer-executable instructions thereon, characterized in that, When the executable instructions are executed by the controller, they implement the zenith rotating dual-polarization radar calibration and diagnostic method as described in any one of claims 1-7.

Images