A weather radar metal ball calibration precision improvement method

Abstract

The application discloses a weather radar metal ball calibration precision improving method, and belongs to the technical field of metal ball calibration. The method comprises the following steps: quantifying the background clutter error of the metal ball calibration position, and removing the background clutter error in the calibration process; constructing a metal ball stability model, quantifying and compensating the dynamic error caused by the swing of the metal ball; quantifying and compensating the metal ball deviation from the beam center error after removing the background clutter error and the metal ball stability satisfying certain conditions; and simultaneously identifying, adjusting and compensating the metal ball cross-warehouse phenomenon. The weather radar metal ball calibration precision is improved through the method.

Description

Technical Field

[0001] This invention relates to the field of metal ball calibration technology, and more specifically to a method for improving the calibration accuracy of metal balls for weather radar. Background Technology

[0002] Recent published research indicates that metal sphere calibration practices are affected by inter-pulse fluctuations, refractive index disturbances, background clutter, and pointing errors, leading to increased uncertainty in reflectivity calibration and challenging the stringent accuracy (0.1–0.2 dB) of polarization quantities such as ZDR. Existing literature largely focuses on experimental procedures or overall corrections, lacking a calculable and quantifiable "detection + compensation" closed loop addressing each type of attributable error. This is mainly reflected in the following aspects:

[0003] (1) Background clutter bias is not systematically quantified and subtracted. Conventional methods often use a simplified "minimum / median value of background" for rough estimation, which is difficult to stably obtain background bias under conditions of complex ground object echoes, low elevation angle and close distance, or scattered meteorological echoes, resulting in the target signal estimate being too high or too low.

[0004] (2) There is a lack of a "stability model" for suspending metal balls for UAVs. The horizontal and vertical swaying of the metal ball caused by wind / maneuvering will introduce transient angle deviation, velocity Doppler broadening and distance sway, which will reduce the effective echo amplitude and increase the estimation variance. The existing process mostly relies on experience to remove "outliers" according to time windows, and lacks quantifiable thresholds and weights.

[0005] (3) The systematic deviation of the metal sphere when it deviates from the beam center lacks quantitative correction. Existing works mostly approximate the main lobe center by "centering sampling" or "two-dimensional surface fitting to find the peak value", but rarely provide calculable amplitude correction factors for the case that has deviated, especially under the conditions of elliptical main lobe, azimuth / elevation asymmetric beam, or synthetic beam (DBF).

[0006] (4) Insufficient identification and compensation for cross-range gate (cross-range gate) effects. When the metal sphere is located at the range gate boundary or when the energy distribution across the gate caused by the pulse-window function is significant, the energy of the central reservoir is "diluted," resulting in a lower reflectivity estimate and a larger deviation in pulse width / beamwidth inversion. Existing technologies mostly use the "maximum reservoir" approach and have not proposed specific and effective methods for cross-range gate identification or practical steps for "manual fine-tuning the position," making it impossible to quantitatively infer the cross-range gate amplitude and perform accurate compensation.

[0007] In summary, radar metal ball calibration typically aims for an accuracy of ±1 dB or better for targets with absolute reflectivity. Existing metal ball-based calibration methods struggle to achieve this consistently in complex scenarios. While external metal ball calibration procedures are available, practical, traceable, calculable, and feasible solutions for multi-source errors are lacking. Summary of the Invention

[0008] In view of this, the present invention provides a method for improving the accuracy of weather radar metal ball calibration, which takes the algorithm flow as the core and improves the accuracy, robustness and reproducibility of weather radar metal ball calibration results.

[0009] To achieve the above objectives, the present invention adopts the following technical solution: A method for improving the calibration accuracy of a weather radar metal sphere includes the following steps: The background clutter error at the calibration location of the metal sphere is quantified, and the background clutter error is removed during the calibration process: Construct a stability model for a metal sphere to quantify and compensate for dynamic errors caused by the oscillation of the metal sphere. After removing background clutter errors and ensuring that the stability of the metal sphere meets certain conditions, the error of the metal sphere deviating from the beam center is quantified and compensated. Simultaneously, the phenomenon of metal balls crossing the storage area is identified, adjusted, and compensated.

[0010] Furthermore, the background clutter error is quantified, specifically including: The radar was used to perform an air scan of the calibration position of the metal ball to obtain the echo power of the metal ball and the background clutter power at the same range. Based on the radar-received echo power of the metallic ball and the background clutter power at the same range, the signal-to-clutter ratio (SCR) is calculated as the quantized value of the background clutter error.

[0011] in, This indicates the power of the echo from the metal sphere received by the radar. This represents the background clutter power at the same distance.

[0012] Furthermore, background clutter errors are removed during the calibration process, specifically including: Based on the total echo power of the radar-scanned metal sphere during calibration and the quantized value (SCR) of the background clutter error, the echo power of the metal sphere after removing the background clutter error is obtained:

[0013] in, This represents the total echo power obtained by the radar during the actual scanning of the metal sphere during the calibration process.

[0014] Furthermore, a stability model for the metal sphere is constructed, specifically including: Using the position of the drone suspending the metal ball as the origin, calculate the horizontal displacement of the metal ball in the East-North-Zenith coordinate system under the current environment. The current environment includes the drone's current flight altitude and the length of the rope suspending the metal ball.

[0015] in, Represents eastward displacement Represents northward displacement, The radius of the Earth; Indicates the longitude coordinates of the drone, Representing the longitude coordinates of the metal sphere, Indicates the latitude coordinates of the drone, Represents the latitude coordinates of the metal sphere; The actual horizontal displacement of the metal sphere relative to the UAV was calculated based on the horizontal displacement of the metal sphere in the East-North-Zen coordinate system. :

[0016] Based on the length of the rope suspending the metal ball on the drone and the maximum swing angle of the metal ball The maximum horizontal displacement of the metal sphere is obtained. :

[0017] Based on the actual horizontal displacement of the metal ball relative to the drone and maximum horizontal displacement Establish a stability model for the metal sphere:

[0018] Where S represents the stability of the metal sphere under the current environment.

[0019] Furthermore, compensation for dynamic errors caused by the oscillation of the metal sphere specifically includes: Based on the stability S and maximum horizontal displacement of the metal sphere under the current environment The actual horizontal displacement of the metal sphere relative to the UAV was obtained. ; The actual horizontal displacement D can be used to deduce the true latitude and longitude coordinates of the metal sphere.

[0020] Furthermore, the error of the metal sphere deviating from the beam center is quantified, specifically including: Based on the scattering theory of metallic spheres and the detection performance of radar systems, the equivalent reflectivity factor of metallic sphere point targets calculated by weather radar based on the backscattering cross section is obtained. :

[0021] in, For radar wavelength, The radar cross section of a metal sphere The angle at which the metal sphere deviates from the vertical direction, The angle at which the metal sphere deviates from the horizontal direction, For radar pulse width, For the refractive index correction factor, Indicates the distance between the metal sphere and the radar. This is the antenna gain correction value for the Gaussian beam; Calculate the peak echo power of a single metal sphere target. :

[0022] in, For radar transmission power, Antenna gain; The formula for the radar constant C is:

[0023] in, This represents the total loss of the radar system feeder. Based on the equivalent reflectivity factor of the metallic spherical target Peak echo power of a single metal sphere target And the radar constant C, to obtain the equivalent reflectivity factor of a single metal sphere target observed by weather radar:

[0024] in, Atmospheric loss; Based on the equivalent reflectivity factor of a single metal sphere target observed by weather radar The average observed value of the equivalent reflectivity factor of the metal sphere within a set time period was obtained. ; Calculate the average gain loss caused by the oscillation of the metal ball within a set time period. And the average observed value of the equivalent reflectivity factor of the metal sphere. Compensation is performed to obtain the compensated equivalent reflectivity factor. :

[0025] in, This represents the average observed value of the equivalent reflectivity factor of the metal sphere. The absolute value of the average gain loss of the metal sphere. This represents the reflectivity factor after compensation.

[0026] Furthermore, the phenomenon of metal balls crossing the storage channel is identified, adjusted, and compensated for, specifically including: Cross-database identification: Based on radar scan data, extract the echo intensity data of the strongest echo point within the theoretical distance range of the metal ball, corresponding to distance n and the echo intensity data of the two preceding and following n+1 distances. Based on the echo intensity data, a two-dimensional planar polygonal graph is plotted within a specified distance range of the location of the metal ball, with the elevation angle of the radar relative to the metal ball as the abscissa and the echo intensity received by the radar from the metal ball as the ordinate. Cross-database identification based on two-dimensional planar polyline graphs; Cross-database adjustment: If there is cross-database, adjust the drone's nose position to the relative position value between the radar and the drone, and control the drone to fly away from or towards the radar. Cross-database compensation: When the metal ball is not in the center of the distance library, the distance library where the metal ball is located is compensated by the adjacent distance library with greater strength.

[0027] Furthermore, cross-database identification is performed based on the two-dimensional planar line graph, specifically including the following steps: When not crossing databases: In the two-dimensional planar polyline graph, the main echo in the distance n from the library exhibits a Gaussian curve shape; The echoes at distances n-1 and n+1 do not exhibit Gaussian graphic characteristics; instead, they show a fluctuating clutter pattern with a maximum amplitude not exceeding 25 dB. When crossing databases: In the two-dimensional planar polyline graph, the main echo n in the distance n from the library exhibits a Gaussian curve shape, and the echo amplitude is smaller than the theoretical value. In the n-1 or n+1 distance library, at least one echo exhibits Gaussian curve characteristics.

[0028] Furthermore, when the metal ball is not at the center of the distance pool, the distance pool containing the metal ball is compensated using an adjacent distance pool with greater strength. This specifically includes the following steps: The linear reflectivity factor is calculated based on the distance of the metal sphere minus the background clutter. And the linear reflectivity factor of adjacent distance pools with higher intensity after subtracting background clutter. Determine the power ratio ,

[0029] According to the power ratio Define the intermediate parameter x:

[0030] According to the power ratio The defined intermediate parameter x, and the linear reflectivity factor after subtracting background clutter from the distance of the metal sphere. The linear reflectivity factor of the distance from the compensated metal sphere to the target is obtained. :

[0031] in, This represents the mean reflectivity factor of the background clutter at the k-th distance in the FIX scan mode.

[0032] As can be seen from the above technical solution, compared with the prior art, the present invention discloses a method for improving the calibration accuracy of metal spheres in weather radar, which has the following beneficial effects: This invention improves the calibration accuracy of weather radar metal spheres by systematically addressing background clutter errors through: quantifying and subtracting background clutter errors; constructing a metal sphere stability model to quantify and compensate for dynamic errors caused by the metal sphere's oscillation; quantifying the error of the metal sphere deviating from the beam center; and identifying, adjusting, and compensating for the phenomenon of the metal sphere crossing the beam.

[0033] The cross-database identification method and steps proposed in this invention can achieve accurate calculation and precise compensation of the cross-database amplitude through quantitative back-calculation. Attached Figure Description

[0034] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0035] Figure 1 A schematic diagram of the metal sphere deviating from the beam center provided in this embodiment of the invention.

[0036] Figure 2 A schematic diagram of the echo of a metal ball crossing a reservoir, provided in an embodiment of the present invention.

[0037] Figure 3 A flowchart for improving the calibration accuracy of a weather radar metal sphere provided in this embodiment of the invention.

[0038] Figure 4 A schematic diagram illustrating the relationship between the received metal ball power deviation and the signal-to-noise ratio provided in this embodiment of the invention.

[0039] Figure 5 This invention provides a schematic diagram illustrating the real-time position capture and inversion of a metal sphere using a high-precision RTK drone system combined with a laser rangefinder camera.

[0040] Figure 6 The schematic diagram of the actual offset in each direction calculated by the metal sphere stability model provided in this embodiment of the invention.

[0041] Figure 7 A schematic diagram of power loss deviation of a simulated metal sphere deviating from the beam center provided in this embodiment of the invention.

[0042] Figure 8 The embodiment of the present invention provides a schematic diagram of the measured power loss deviation of a metal sphere from the beam center.

[0043] Figure 9 This is an example diagram of cross-library data from a metal sphere RHI scan provided in an embodiment of the present invention.

[0044] Figure 10 This is an example diagram of cross-library data obtained by PPI scanning of a metal ball, provided in an embodiment of the present invention.

[0045] Figure 11 This is an example diagram of a sphere-based FIX fixed-point scanning non-cross-database data provided in an embodiment of the present invention. Detailed Implementation

[0046] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] Figure 1 A schematic diagram showing the metal sphere deviating from the beam center. Figure 2 This diagram illustrates the echo signal when a metal sphere crosses a reservoir. For improving the calibration of metal spheres in weather radar, refer to... Figure 3 This invention discloses a method for improving the calibration accuracy of a weather radar metal sphere, comprising the following steps: The background clutter error at the calibration location of the metal sphere is quantified, and the background clutter error is removed during the calibration process: Construct a stability model for a metal sphere to quantify and compensate for dynamic errors caused by the oscillation of the metal sphere. After removing background clutter errors and ensuring that the stability of the metal sphere meets certain conditions, the error of the metal sphere deviating from the beam center is quantified and compensated. Simultaneously, the phenomenon of metal balls crossing the storage area is identified, adjusted, and compensated.

[0048] Specifically, this invention divides the metal sphere calibration and enhancement process into four collaborative steps: M1: background clutter deviation quantification and subtraction; M2: metal sphere stability model and weighted screening; M3: response correction for deviation from beam center; M4: cross-library identification and position / amplitude compensation.

[0049] In the specific implementation process, the echo voxel / radial data near the metal sphere is collected → M1 estimates and subtracts background bias → M2 estimates stability and weights / screens samples → M3 obtains eccentricity based on two-dimensional / three-dimensional beam fitting and performs amplitude correction → M4 identifies cross-pools, estimates the energy distribution ratio within the pool and performs position / amplitude compensation → outputs a standard response that is "equivalently centered, equivalent to no cross-pools, and equivalent to no background", which is used to invert Z, ZDR, beamwidth, gain, pulse width, etc.

[0050] In the process of calibrating a metal ball, taking into account the influence of background clutter, this invention proposes a quantitative calculation method for the deviation of the main signal of the metal ball caused by background clutter. According to different calibration scenarios and accuracy requirements, a calibration background field with a corresponding noise ratio is selected for calibration.

[0051] Metal ball calibration requires extracting pure point target power from a single distance library; therefore, it is necessary to reduce the impact of weather echoes and other clutter.

[0052] Meteorological clutter: The experiment was conducted on a day with no precipitation and low cloud cover. Ground humidity does not affect the far field, but weak echoes should still be avoided; Elevation angle ≥6–8°, avoid distant non-precipitating layers (stratus clouds, insects); Fly the drone radially, keeping the metal sphere at the same distance from the container; Avoid cutting through cloud bodies at an angle; when hovering, the vertical projection of the sphere-radar connection should not pass through obvious weather cells.

[0053] Other background clutter: Assume the power of the metallic sphere received by the radar is... The background clutter power of the metal sphere at the distance from the pool is The SCR is expressed by the following formula, with the unit being dB.

[0054]

[0055] Perform a dry scan at the theoretical location of the metal sphere to obtain the background clutter power at that distance when there is no metal sphere. When the metal sphere is present, obtain the total power. The SCR can be obtained by calculation, and the specific error value can be obtained by looking up a table.

[0056] During the calibration process of the metal sphere, the echo power of the metal sphere after removing background clutter error can be quickly obtained based on the total echo power of the radar scan of the metal sphere and the quantized value of the background clutter error, SCR.

[0057] in, This represents the total echo power obtained by the radar during the actual scanning of the metal sphere during the calibration process.

[0058] During metal sphere calibration, due to uneven atmospheric wind speeds, the sphere can hardly remain completely stationary, often exhibiting varying degrees of swaying, oscillation, or drift. However, the stability of the sphere directly affects the accuracy of echo measurements. To address the spatial deviations caused by metal sphere oscillations during calibration, this invention proposes a "metal sphere stability model" to quantify and compensate for dynamic errors caused by sphere oscillations, reducing positional shifts, antenna pointing errors, and scattering surface fluctuations resulting from sphere instability.

[0059] During calibration, the metal sphere is typically considered to be at the beam center and the standard radar equations can only be applied when there are very small angular and distance deviations in the air. However, in actual calibration, the metal sphere is carried to high altitudes by a drone, making it impossible to guarantee that its actual position is always sufficiently close to the assumed beam center or distance center. Therefore, the position of the metal sphere is first converted to an East-North-Zenith coordinate system, using the drone's position as the farthest point:

[0060] in, Represents eastward displacement Represents northward displacement, The radius of the Earth; Indicates the longitude coordinates of the drone, Representing the longitude coordinates of the metal sphere, Indicates the latitude coordinates of the drone, Represents the latitude coordinates of the metal sphere; in addition to the formulas mentioned above, for displacement in the zenith direction, and These represent the altitude of the drone and the altitude of the metal sphere, respectively.

[0061] in, This represents displacement in the zenith direction.

[0062] Then, the actual horizontal displacement of the metal sphere relative to the UAV is calculated based on the transformed East-North-Zen coordinate system of the metal sphere. :

[0063] The maximum displacement of the metal ball is obtained based on the rope length and the maximum swing angle:

[0064] At this point, the distance difference between the actual displacement and the maximum displacement is The stability of the metal sphere is:

[0065] A series of metal ball stability parameters can be obtained through parameterization (such as different flight altitudes, rope lengths, etc.). By matching the corresponding states, the swing amplitude can be predicted to change with time and position during flight, thereby establishing a metal ball stability model to complete the search and position centering of the metal ball echo signal in the shortest possible time.

[0066] In metal ball calibration, when there is wind in the calibration environment, the metal ball will shift relative to the drone or its theoretical position. Assume that in calm conditions, the metal ball will not shift, i.e., it will remain stationary directly below the drone; when there is wind, the metal ball will shift. Assume that the metal ball shifts in the direction perpendicular to the drone. According to the rope length The horizontal offset distance of the metal ball can be calculated. :

[0067] Assume the initial position of the metal sphere is at a distance from the radar. Then we can obtain the offset angle of the metal sphere relative to the horizontal direction of the radar (the angle of offset relative to the beam center):

[0068] Under normal circumstances, the effect of wind will cause the angle of deviation from the beam center to not exceed 0.3°. The gain loss caused by the degree of deviation from the beam center can be expressed by the following formula:

[0069] in, This refers to the horizontal beamwidth of the radar.

[0070] Therefore, to quantitatively assess and control the stability error of the metal sphere, a model of the sphere's motion is established, primarily considering the intermittent oscillation model and the drift caused by random disturbances. This model enables real-time monitoring of the sphere's position changes. Based on the above equations, and by studying sphere scattering theory and radar system detection performance, a weather radar equation for calculating the equivalent reflectivity factor of a metal sphere point target based on the backscattering cross section can be obtained:

[0071] in, For radar wavelength, The radar cross section of a metal sphere The angle at which the metal sphere deviates from the vertical direction, The angle at which the metal sphere deviates from the horizontal direction, For radar pulse width, For the refractive index correction factor, The distance between the metal sphere and the radar This is the antenna gain correction value for the Gaussian beam (1.4 dB). The measured value of the equivalent reflectivity factor of the metallic sphere observed by radar can be calculated according to the weather radar equation. The peak echo power of a single metallic sphere target is calculated as follows:

[0072] The formula for the radar constant C is:

[0073] Therefore, the formula for the equivalent reflectivity factor of a single metal sphere target observed by weather radar can be obtained:

[0074] in, The total power received by the antenna, For radar transmission power, For antenna gain, Atmospheric loss (0.011 for S-band, 0.019 for C-band, and 0.025 for X-band). The total loss of the system feeder (matching filter loss + transmit branch loss + receive branch loss + pulse compression loss (pulse compression radar)). The angle at which the metal sphere deviates from the beam center.

[0075] The above derivation exists in a theoretical scenario. In reality, the opposite is true. When the metal ball swings, it deviates from the beam center, resulting in a decrease in gain and a phase change. However, environmental clutter has little phase change, so the swing ultimately weakens the received signal strength. To compensate for this intensity loss, we combine real-time acquisition of the metal ball's position using a laser rangefinder camera and calculate the angle at which the metal ball deviates from the radar beam direction during swing. Finally, by calculating the effect of the angle deviation as described in the patent, we calculate the gain loss at all recorded positions of the metal ball.

[0076] Since the final calibration intensity is the average value of the fixed-point scans, we do not need to precisely determine the gain loss caused by the oscillation of the metal ball at every moment; we only need to compensate for it using incoherent averaging. We only need to calculate the average gain loss caused by the oscillation of the metal ball over a period of time and compensate for it on the average observation value of the metal ball:

[0077] in, This represents the average observed value of the metal sphere. The absolute value of the average gain loss of the metal sphere. This represents the compensated reflectivity factor. By parameterizing parameters (such as the gain loss caused by the metal sphere's deviation from the beam center, and the equivalent reflectivity factor of a single metal sphere target), the relationship between the angle change of the metal sphere's deviation from the beam center and the gain loss can be obtained. This can reduce the impact of metal sphere swaying and ground clutter during metal sphere calibration to a certain extent, thereby obtaining more accurate calibration results.

[0078] (When the radar is working, the detection range is divided into several equally spaced "range libraries," each corresponding to a certain range, used to record the target echo signal. As a point target, the echo of the metal sphere should theoretically fall within a specific range library, facilitating comparison and calibration with theoretical values.) 1.4, M4: Metal ball cross-library identification, adjustment and compensation Ideally, during weather radar calibration, the metal sphere should be positioned at the center of the radar beam's main lobe, ensuring its echo signal falls completely within the entire main range cell, thus maximizing the observed reflected power. However, in actual calibration, the metal sphere's position at the center of the range cell is accidental, leading to the echo signal being received by multiple range cells, potentially underestimating the total echo energy. This invention proposes a method for identifying metal spheres across range cells and outlines specific adjustment steps to address this issue. This method effectively mitigates energy loss caused by the metal sphere crossing range cells, significantly improving the accuracy and stability of radar calibration.

[0079] When a metal sphere is used as a point target, if it is not located at the center of a certain distance reservoir, the echo energy will be distributed between adjacent reservoirs, resulting in a lower echo value observed in the main reservoir and a large difference in values ​​between adjacent reservoirs. The core reason for this is insufficient "pulse-window overlap" and energy splitting. This invention proposes a method for identifying, adjusting, and correcting whether a metal sphere crosses reservoirs, combining data extraction and image recognition. 1. Extract data.

[0080] Extract the range library (n) corresponding to the strongest echo point near the theoretical distance position of the metal ball from the RHI and PPI scan data, as well as the echo intensity data of the two libraries before (n-1) and after (n+1). For radar data with high radial range resolution (≤30m), the range library where the metal ball is located and the data of the four libraries before and after (a total of five libraries) can be extracted for analysis. 2. Data inversion.

[0081] Draw a two-dimensional planar polyline graph with the angle (horizontal axis) and echo intensity (vertical axis) near the location of the metal sphere as the key elements; 3. Cross-database identification.

[0082] The echo amplitude and graphic characteristics are analyzed. When there is no cross-database movement: the main echo n exhibits a Gaussian curve shape; the echoes at distances n-1 and n+1 from the database show virtually no Gaussian graphic characteristics, mainly exhibiting fluctuating clutter patterns, with a maximum amplitude typically not exceeding 25 dB. When there is cross-database movement: the main echo n exhibits a Gaussian curve shape, with an echo amplitude smaller than the theoretical value. One of the echoes at n-1 or n+1 also exhibits Gaussian curve characteristics. In cases of severe cross-database movement, the amplitude of one of the echoes at n-1 or n+1 is closer to the amplitude of the main echo n.

[0083] 4. Cross-database adjustment.

[0084] If the echo characteristic of the (n-1)th pool is Gaussian and the echo amplitude is close to that of the nth pool, it indicates that the metal ball is biased towards the (n-1)th pool (the previous pool). In this case, adjust the drone's heading to match the relative heading of the radar and the drone, and control the drone to fly away from the radar. Taking a 250m radar distance pool as an example, the distance the drone should fly away from the radar can be set based on the deviation distance between the (n-1)th and nth pools shown in the line graph in step 2. Adjustments can be made in 20m increments, and repeated testing is performed until the echo pattern shows the state of not crossing any pools as described in step 3. When the metal ball is biased towards the (n+1)th pool, the drone operation is similar; simply fly the drone closer to the radar.

[0085] 5. Cross-database compensation.

[0086] When the metal sphere is not located at the center of the distance reservoir, its energy will leak into adjacent distance reservoirs. The distance reservoir corresponding to the strongest echo point near the theoretical distance position of the metal sphere is (n). Therefore, the echo intensity data of the two reservoirs before (n-1) and after (n+1) will show a "two large and one small" situation. The distance reservoir with the smaller intensity can be approximated as originating from ground features. Therefore, it is necessary to use the adjacent distance reservoir with the larger intensity to compensate for the distance reservoir where the metal sphere is located. The specific compensation strategy is as follows: (1) Power Calculation. The reflectivity factors of the two distance pools with relatively large intensity are converted to linear, and then the intensity of the background clutter is subtracted to obtain the contribution of the metal sphere to the reflectivity factor. The specific formula is as follows:

[0087] Where Pk represents the reflectivity factor after conversion to linearity, Zk represents the mean reflectivity factor of the k-th range library in FIX scanning mode, and Zb,k represents the mean reflectivity factor of the background clutter in the k-th range library in FIX scanning mode.

[0088] (2) Offset estimation. ΔR is half the distance to the library resolution; therefore, the normalized offset is defined as:

[0089] Where Δr represents the distance the metal sphere deviates from the center of the pool. It can be assumed that the strength of adjacent pools is related to x. 2 The strength of the metal sphere at a distance from the sphere is directly proportional to (1-x). 2 Proportional, defined as the power ratio:

[0090] Where Pc is the linear reflectivity factor of the distance library where the metal sphere is located minus the background clutter, and Pl is the linear reflectivity factor of the adjacent distance library with greater intensity minus the background clutter.

[0091] (3) Reflectivity factor compensation. According to the definition of power ratio, it can be seen that:

[0092] Therefore, the linear reflectivity factor of the distance Ck where the compensated metal sphere is located is:

[0093] Finally, the result can be converted into dB form to obtain the compensated reflectivity factor.

[0094] To improve the calibration accuracy of weather radar and to address issues such as background clutter, placement deviation of metal balls, and metal balls crossing storage areas during the calibration process, this invention proposes a multi-factor comprehensive correction method to enhance calibration accuracy.

[0095] To verify the effectiveness of this invention, the technical solution of this invention is verified, and the specific steps are as follows: I. Test Conditions 1. Frequency band: S / C / X (any one); Pulse mode: narrowband / linear frequency modulation (including pulse compression parameters); PRF: 300–1200 Hz; Distance resolution: ≤250m; 2. Sphere: Hollow aluminum sphere, diameter 0.15–0.6 m (selected according to wavelength and RCS requirements), suspension rope / suspension point avoids the main lobe.

[0096] 3. Carrier: Multi-rotor UAV with static hovering capability and RTK positioning capability.

[0097] II. Data Acquisition Set the scanning mode (cross scan or narrow sector scan) for the target sphere, and record the H / V channel base data and IQ data (including at least: Z, V, ZDR, ρHV).

[0098] III. Background Clutter Deviation Quantization and Compensation According to the background clutter deviation quantization model, the relationship between the received metal ball echo power deviation and the signal-to-noise ratio shows a nonlinear trend. Figure 4During the calibration of the metal sphere, the presence or absence of the metal sphere can be determined by comparing the echo characteristics of the spheres at different distances, and the SCR value can be calculated accordingly.

[0099] When no metal ball is detected in the target range library, its received power can be regarded as pure background clutter power; When the metal ball enters the bin, the measured power is the superposition of the target echo and background clutter. As shown in the figure, when the SCR is generally greater than 32dB, the error can be controlled within 0.2dB. In metal ball calibration, the duration is usually short, and it can be assumed that other background clutter changes little, while the phase of the metal ball signal is constantly changing. When other background clutter causes the SCR to fall short of 32dB, at least 0.2dB of deviation loss should be compensated for in the calibration result (not considering the effect of the metal ball deviating from the beam center). The specific compensation value can be linearly compensated based on the actual SCR value.

[0100] IV. Application of Stability Model and Examples of Lateral Misalignment A fusion scheme combining an RTK drone and a laser rangefinder camera was used to acquire high-precision effective position and attitude information of a metal sphere. Figure 5 ).

[0101] Based on the above metal ball stability model, the real-time position data of the metal ball collected during the first metal ball calibration was analyzed. With the maximum allowable swing angle preset to 30°, the eastward and northward displacements of the metal ball relative to the UAV were calculated to be within ±2.2m, and the total horizontal displacement of the metal ball relative to the UAV was basically within 2m. Therefore, the stability index of the metal ball was calculated to be above 97%. Figure 6 Based on this, determine whether it is "stable" or "requires retesting", and select the optimal calibration position with the least sway.

[0102] V. Simulation and Measured Results of Metal Sphere Off-Beam Center like Figure 7 The figure shows the simulation results of the relationship between the deviation of the metal sphere from the beam center and the reflectivity factor loss under three different waveband conditions. The results show that, as... Figure 7 As shown, with the increase of the angle at which the metal sphere deviates from the beam center, the measured reflectivity factor also shows a significant increasing trend. When the radar beamwidth is about 1°, a measured azimuth deviation of 0.5dB will cause a deviation of about 7dB. Figure 8 As shown, the measured echo deviation occurs when the radar azimuth deviation angle is manually changed. It can be seen that when the azimuth deviation is 0.5°, the reflectivity factor deviation is also around 7dB. The measured results are consistent with the theoretical simulation results.

[0103] VI. Metal Ball Cross-Library Consistency Correction During the calibration process, the metal sphere is usually not in the center of the reservoir, which will result in the phenomenon that the echo energy is distributed between adjacent reservoirs. This leads to the observed echo being lower and close to the echo of the adjacent reservoir. The core reason is the insufficient "pulse-window overlap" and energy splitting. The echo power of multiple adjacent volume units is judged. If the cross-reservoir condition is met (for example, the power ratio of adjacent unit reservoirs exceeds the set threshold), the cross-reservoir condition is determined.

[0104] Workflow: First, the position of the metal ball is adjusted by suspending it with a drone, placing the metal ball at a distance from the center of the database; second, the calibration data can be processed by multi-database weighted fusion to correct cross-database deviations. Figure 9 and Figure 10 The study demonstrated that during RHI and PPI scans, the metal ball was biased towards the next coil, causing a spike in the echo energy of the (n+1)th coil, resulting in a "double peak" phenomenon. After suspending the metal ball by a drone and moving its position 50 meters radially backward along the radar, a subsequent FIX point scan showed that the echoes of the (n-1)th and (n+1)th coils were essentially aligned at the bottom, with only the main peak echo standing out. Figure 11 This shows that a cross-basin deviation correction of about 4dB was achieved by correctly adjusting the position of the metal ball.

[0105] The feasibility of this method was verified using actual data in the embodiments of the present invention. The results are shown in the table below.

[0106]

[0107] In the RHI scan mode when crossing range objects, the reflectivity factors of the two range objects with higher intensity are 45.26 dBZ and 40.34 dBZ, respectively, according to the order of range object intensity. The theoretical value is 51.02 dBZ. Since the intensity of the ground objects is relatively low, this can be ignored here. In the PPI scan mode when crossing range objects, the reflectivity factors of the two range objects with higher intensity are 46.9 dBZ and 40.95 dBZ, respectively. After compensation using the method described above, the value in the RHI scan mode is 49.16 dBZ, and the value in the PPI scan mode is 50.45 dBZ. Because the cross-scan method performs RHI first and then PPI, the radar may not be aligned with the metal sphere during RHI; therefore, the value of the compensated PPI scan mode is used here. It can be seen that the compensated value is 50.45 dB, which is 0.57 dB lower than the theoretical value.

[0108] The distance to the metal sphere was adjusted to ensure it did not cross the filter. The theoretical value after adjustment was 51.32 dB, and the average value in FIX mode was 51.09 dB. Based on the data without crossing the filter, the detected value was 0.23 dB lower than the theoretical value. Comparing the results of cross-filter compensation and the results without crossing the filter, the difference between the compensated result and the theoretical value was 0.34 dB higher than the difference in the results without crossing the filter. Considering that data from only the PPI scanning mode may have some fluctuations, and that the compensated result meets the standard of reflectivity factor error less than 1 dB, it is generally believed that this compensation method can effectively reduce the loss caused by crossing the filter.

[0109] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section.

[0110] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for improving the calibration accuracy of a metal sphere for weather radar, characterized in that, Includes the following steps: The background clutter error at the calibration location of the metal sphere is quantified, and the background clutter error is removed during the calibration process: Construct a stability model for a metal sphere to quantify and compensate for dynamic errors caused by the oscillation of the metal sphere. Constructing a stability model for the metal sphere specifically includes: Using the position of the drone suspending the metal ball as the origin, calculate the horizontal displacement of the metal ball in the East-North-Zenith coordinate system under the current environment. The current environment includes the drone's current flight altitude and the length of the rope suspending the metal ball. in, Represents eastward displacement Represents northward displacement, The radius of the Earth; Indicates the longitude coordinates of the drone, Representing the longitude coordinates of the metal sphere, Indicates the latitude coordinates of the drone, Represents the latitude coordinates of the metal sphere; The actual horizontal displacement of the metal sphere relative to the UAV was calculated based on the horizontal displacement of the metal sphere in the East-North-Zen coordinate system. : Based on the length of the rope suspending the metal ball on the drone and the maximum swing angle of the metal ball The maximum horizontal displacement of the metal sphere is obtained. : Based on the actual horizontal displacement of the metal ball relative to the drone and maximum horizontal displacement Establish a stability model for the metal sphere: Where S represents the stability of the metal sphere under the current environment; After removing background clutter errors and ensuring that the stability of the metal sphere meets certain conditions, the error of the metal sphere deviating from the beam center is quantified and compensated. Simultaneously, the phenomenon of metal balls crossing the storage channel is identified, adjusted, and compensated for, specifically including: Cross-database identification: Based on radar scan data, extract the echo intensity data of the strongest echo point within the theoretical distance range of the metal ball, corresponding to distance n and the echo intensity data of the two preceding and following n+1 distances. Based on the echo intensity data, a two-dimensional planar polygonal graph is plotted within a specified distance range of the location of the metal ball, with the elevation angle of the radar relative to the metal ball as the abscissa and the echo intensity received by the radar from the metal ball as the ordinate. Cross-database identification based on two-dimensional planar polyline graphs; Cross-database adjustment: If there is cross-database, adjust the drone's nose position to the relative position value between the radar and the drone, and control the drone to fly away from or towards the radar. Cross-database compensation: When the metal ball is not at the center of the distance pool, the distance pool containing the metal ball is compensated using an adjacent distance pool with greater strength. This includes the following steps: The linear reflectivity factor is calculated based on the distance of the metal sphere minus the background clutter. And the linear reflectivity factor of adjacent distance pools with higher intensity after subtracting background clutter. Determine the power ratio , According to the power ratio Define the intermediate parameter x: According to the power ratio The defined intermediate parameter x, and the linear reflectivity factor after subtracting background clutter from the distance of the metal sphere. The linear reflectivity factor of the distance from the compensated metal sphere to the target is obtained. : in, This represents the mean reflectivity factor of the background clutter at the k-th distance in the FIX scan mode.

2. The method for improving the calibration accuracy of a weather radar metal sphere according to claim 1, characterized in that, Quantization of background clutter error includes: The radar was used to perform an air scan of the calibration position of the metal ball to obtain the echo power of the metal ball and the background clutter power at the same range. Based on the radar-received echo power of the metallic ball and the background clutter power at the same range, the signal-to-clutter ratio (SCR) is calculated as the quantized value of the background clutter error. in, This indicates the power of the echo from the metal sphere received by the radar. This represents the background clutter power at the same distance.

3. The method for improving the calibration accuracy of a weather radar metal sphere according to claim 2, characterized in that, Removing background clutter errors during the calibration process specifically includes: Based on the total echo power of the radar-scanned metal sphere during the calibration process, and the quantized value of the background clutter error (SCR), the echo power of the metal sphere after removing the background clutter error is obtained: in, This represents the total echo power obtained by the radar during the actual scanning of the metal sphere during the calibration process.

4. The method for improving the calibration accuracy of a weather radar metal sphere according to claim 1, characterized in that, Compensation for dynamic errors caused by the oscillation of the metal sphere specifically includes: Based on the stability S and maximum horizontal displacement of the metal sphere under the current environment The actual horizontal displacement of the metal sphere relative to the UAV was obtained. ; The actual horizontal displacement D can be used to deduce the true latitude and longitude coordinates of the metal sphere.

5. The method for improving the calibration accuracy of a weather radar metal sphere according to claim 1, characterized in that, The quantification of the error in the metal sphere's deviation from the beam center includes: Based on the scattering theory of metallic spheres and the detection performance of radar systems, the equivalent reflectivity factor of metallic sphere point targets calculated by weather radar based on the backscattering cross section is obtained. : in, For radar wavelength, The radar cross section of a metal sphere The angle at which the metal sphere deviates from the vertical direction, The angle at which the metal sphere deviates from the horizontal direction, For radar pulse width, For the refractive index correction factor, Indicates the distance between the metal sphere and the radar. This is the antenna gain correction value for the Gaussian beam; Calculate the peak echo power of a single metal sphere target. : in, For radar transmission power, Antenna gain; The formula for the radar constant C is: in, This represents the total loss of the radar system feeder. Based on the equivalent reflectivity factor of the metallic spherical target Peak echo power of a single metal sphere target And the radar constant C, to obtain the equivalent reflectivity factor of a single metal sphere target observed by weather radar: in, Atmospheric loss; Based on the equivalent reflectivity factor of a single metal sphere target observed by weather radar The average observed value of the equivalent reflectivity factor of the metal sphere within a set time period was obtained. ; Calculate the average gain loss caused by the oscillation of the metal ball within a set time period. And the average observed value of the equivalent reflectivity factor of the metal sphere. Compensation is performed to obtain the compensated equivalent reflectivity factor. : in, This represents the average observed value of the equivalent reflectivity factor of the metal sphere. The absolute value of the average gain loss of the metal sphere. This represents the reflectivity factor after compensation.

6. The method for improving the calibration accuracy of a weather radar metal sphere according to claim 1, characterized in that, Cross-database identification based on a two-dimensional planar polyline chart includes the following steps: When not crossing databases: In the two-dimensional planar polyline graph, the main echo in the distance n from the library exhibits a Gaussian curve shape; The echoes at distances n-1 and n+1 do not exhibit Gaussian graphic characteristics; instead, they show a fluctuating clutter pattern with a maximum amplitude not exceeding 25 dB. When crossing databases: In the two-dimensional planar polyline graph, the main echo n in the distance n from the library exhibits a Gaussian curve shape, and the echo amplitude is smaller than the theoretical value. In the n-1 or n+1 distance library, at least one echo exhibits Gaussian curve characteristics.

Images