PPI wind shear extraction method, device and storage medium based on laser wind radar

Abstract

The invention discloses a PPI wind shear extraction method, device and storage medium based on a laser wind measuring radar, comprising the following steps: S1, based on PPI scanning, obtaining radial data of N light beams within a cycle, and screening and marking the radial data; S2, inverting the horizontal wind field based on a light beam sliding window to obtain the horizontal wind speed and horizontal wind direction of the light beam; S3, gridding the horizontal wind field, and interpolating based on the detection distance; compared with the prior art, the invention is based on the radial data obtained by PPI scanning, adopts a sliding window method, substitutes the horizontal wind speed and horizontal wind direction calculated by the previous sliding window as the initial values ​​of the model parameters into the gradient descent solution model of the current light beam, thereby inverting the horizontal wind speed and horizontal wind direction of the current light beam, reducing the wind field inversion error when the wind direction is close to perpendicular to the light beam direction, and improving the inversion accuracy of the horizontal wind field.

Description

Technical Field

[0001] The present invention relates to the field of wind shear prediction, and in particular to a PPI wind shear extraction method, device and storage medium based on a laser wind radar. Background Art

[0002] At present, the PPI wind shear extraction method based on laser wind radar mainly uses the radial wind speed information directly obtained by PPI scanning to extract shear information. When the laser wind radar performs a PPI plane scan at a low elevation angle, the radial velocity can be approximately regarded as the radial projection of the atmospheric wind field on the horizontal plane. The radial shear value and the tangential shear value are calculated according to the radial wind speed characteristics, and the wind shear value of the horizontal wind field is obtained by combining them. However, in the actual process of wind shear information extraction, the information provided by the radial wind speed is not as complete as that of the two-dimensional horizontal wind field, resulting in differences between the shear value based on the radial wind speed and the wind shear distribution of the actual two-dimensional wind field, which cannot truly represent the wind shear information of the actual wind field, bringing certain difficulties to wind shear warning in the fields of aviation and meteorology.

[0003] In the process of PPI wind shear extraction by laser wind radar, there is a difference between the existing shear value calculated based on radial wind speed characteristics and the shear value of the actual horizontal wind field, which brings certain difficulties to the application of wind shear warning. Summary of the invention

[0004] The purpose of the present invention is to address the problem that during the PPI wind shear extraction process performed by the radar, the information provided by the radial wind speed is incomplete, resulting in a difference between the calculated shear value and the shear value of the actual horizontal wind field, which brings certain difficulties to the application of wind shear warning. To address this problem, a PPI wind shear extraction method, device and storage medium based on a laser wind measuring radar are proposed.

[0005] In order to achieve the above object, the present invention adopts the following technical solution:

[0006] The PPI wind shear extraction method based on laser wind radar includes the following steps:

[0007] S1. Based on PPI scanning, radial data of N light beams in one cycle are obtained, and the radial data are screened and marked. The radial data includes radial wind speed V. When the value of radial wind speed V of a light beam in the cycle is an invalid value Nan, the light beam is not inverted, that is, the horizontal wind speed of the light beam is an invalid value Nan;

[0008] S2, invert the horizontal wind field based on the beam sliding window to obtain the horizontal wind speed and horizontal wind direction of the beam;

[0009] S3. Divide the horizontal wind field into grids, and interpolate based on the detection distance to obtain the wind speed component in the X direction of the interpolation point. , Y direction wind speed component ;

[0010] S4. For the wind speed component of the interpolation point, linear fitting is performed based on the moving window method to obtain the wind speed along the X direction. Rate of change , along the Y direction Rate of change , along the X direction Rate of change , along the Y direction Rate of change ;in, is the wind speed component in the X direction The amount of change in the X direction; is the wind speed component in the Y direction The amount of change in the Y direction; is the change distance in the X direction; is the change distance in the Y direction;

[0011] S5. According to , the wind shear factor is calculated .

[0012] S11, when the radar is scanning based on PPI, the elevation angle remains unchanged and the azimuth angle is adjusted step by step;

[0013] S12, obtaining radial data of N light beams in one cycle; the radial data includes: radial distance Dist, azimuth angle Azi; N is an odd number;

[0014] S13, obtaining the maximum range Range of the scanning azimuth angle in the cycle and the azimuth angle deviation bias of a single light beam;

[0015] S14, taking the light beam at the (N+1) / 2 position as the central wave, and obtaining the azimuth angle Azi_cnt of the central wave;

[0016] S15. Filter out the satisfied The radial data within the central beam sector is obtained.

[0017] As a further preferred embodiment of the present invention, S2 comprises the following specific steps:

[0018] S21. Establish a dynamic model based on the VAD wind field inversion method based on gradient descent;

[0019] S22, calculating the horizontal wind speed and horizontal wind direction of the last sliding window, and using the horizontal wind speed and horizontal wind direction of the last sliding window as initial value parameters of the dynamic model; the last sliding window is the N light beams formed in the last cycle;

[0020] S23. Horizontal wind speed and horizontal wind direction of the inversion beam.

[0021] As a further preferred embodiment of the present invention, S3 comprises the following steps:

[0022] S31, confirming that the horizontal wind speed is not the maximum detection distance maxDist of Nan;

[0023] S32, divide the square grid with the maximum detection distance maxDist as the boundary; wherein the detection distance in the X direction is , Y direction detection distance ;

[0024] S33, reverse detection distance Location and detection distance Wind speed component at , , and insert it into the corresponding interpolation point (Xm, Yn) of the square grid;

[0025] S34, the maximum distance from the interpolation point (Xm, Yn) to the spatial adjacent point set (Xgm, Ygn) is MaxD;

[0026] in, ;

[0027] S35. Traverse all grid point position sets (Xg, Yg) whose distance to the interpolation point is less than the maximum distance MaxD, and the corresponding wind speed component is and ;

[0028] S36, according to

[0029] ,

[0030] ,

[0031] ,

[0032] in, is the weight corresponding to the interpolation point (Xm, Yn);

[0033] Calculate the weighted average of the wind speed components at the interpolation point and .

[0034] A PPI wind shear extraction device based on laser wind radar. The device includes one or more processors and a memory, wherein the memory is used to store one or more programs. When the one or more programs are executed by the processor, the processor is enabled to implement a PPI wind shear extraction method based on laser wind radar.

[0035] A computer storage medium stores at least one program instruction, and the at least one program instruction is used to be loaded and executed by a processor to implement a PPI wind shear extraction method based on a laser wind radar.

[0036] Compared with the prior art, the PPI wind shear extraction method, device and storage medium based on laser wind radar proposed in the present invention have the following beneficial effects:

[0037] 1. The present invention is based on the radial data obtained by PPI scanning, and adopts a sliding window method to substitute the horizontal wind speed and horizontal wind direction calculated by the previous sliding window as the initial values ​​of the model parameters into the gradient descent solution model of the current light beam, thereby inverting the horizontal wind speed and horizontal wind direction of the current light beam, reducing the wind field inversion error when the wind direction is close to perpendicular to the light beam direction, and improving the inversion accuracy of the horizontal wind field.

[0038] 2. The present invention starts from the horizontal wind field of the PPI scanning sector, constructs a horizontal wind field grid, and adopts a moving window method to extract wind shear information based on the rate of change of the X-direction velocity component and the Y-direction velocity component of the horizontal wind field. This solves the problem that the tangential shear value is too large at close range for the same radial wind speed difference and the radial shear values ​​are different for wind shears of the same intensity in different wind directions, thereby improving the accuracy of wind shear information extraction. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is based on the PPI scanning schematic diagram;

[0040] Figure 2 It is a schematic diagram of wind field inversion;

[0041] Figure 3 is a schematic diagram of the change of the central beam of the radial data set before sliding;

[0042] Figure 4 is a schematic diagram of the change of the central beam of the radial data set after sliding;

[0043] Figure 5 It is a schematic diagram of the change of horizontal wind direction after inversion of fixed model parameters;

[0044] Figure 6 It is a schematic diagram of the change of horizontal wind direction from the inversion of dynamic model parameters;

[0045] Figure 7 is an example of a moving window calculation. DETAILED DESCRIPTION

[0046] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0047] The purpose of the present invention is to provide a PPI wind shear extraction method, device and storage medium based on laser wind radar to solve the problem that the tangential shear value of the same radial wind speed difference is larger in the near distance and the radial shear value of wind shear of the same intensity is different in different wind directions.

[0048] Embodiment 1: A PPI wind shear extraction method based on a laser wind radar comprises the following steps:

[0049] S1. Based on PPI scanning, radial data of N light beams in one cycle are obtained, and the radial data are screened and marked. The radial data includes radial wind speed V. When the value of radial wind speed V of a light beam in the cycle is an invalid value Nan, the light beam will not be inverted, that is, the horizontal wind speed of the light beam is an invalid value Nan.

[0050] S11. When the radar scans based on PPI, the elevation angle remains unchanged and the azimuth angle is adjusted in steps.

[0051] S12, obtain radial data of N light beams in one cycle; the radial data includes: radial distance Dist, radial wind speed V, azimuth Azi; N is an odd number. Take N=9 as an example, the scanning schematic diagram is shown in Figure 1 .

[0052] S13, obtaining the maximum range Range of the scanning azimuth angle within the cycle and the azimuth angle deviation bias of a single beam; the azimuth angle deviation of a single beam is the scanning accuracy of the radar azimuth angle, that is, the deviation between the azimuth angle of the current beam set and the actual scanning azimuth angle of the radar.

[0053] S14. Take the light beam at the (N+1) / 2 position as the central wave, and obtain the azimuth angle Azi_cnt of the central wave.

[0054] S15. Filter out the satisfied The radial data within the central beam sector is obtained.

[0055] S2. Invert the horizontal wind field based on the light beam sliding window to obtain the horizontal wind speed and horizontal wind direction of the light beam.

[0056] Based on the assumption that the local wind field is uniform, the velocity azimuth display (VAD) method is currently a widely used method for retrieving the wind field based on a single laser wind radar, such as Figure 2 As shown, taking the positive direction of the x-axis as due north, the positive direction of the y-axis as due east, and the positive direction of the z-axis as due up, ignoring the influence of the vertical wind speed, the relationship between the radial wind speed V and the scanning azimuth angle measured by the coherent laser wind radar is shown in the following formula:

[0057] ,

[0058] in, is the scanning azimuth, is the elevation angle, u and v are the x and y component values ​​of the wind speed.

[0059] In order to solve the horizontal wind speed and horizontal wind direction, the VAD wind field inversion method based on gradient descent is used to define a learning model A corresponding relationship can be established for any given input x and output y. Under the assumption that the local wind field is uniform, the model and the corresponding cost function are It can be expressed as:

[0060] ,

[0061] .

[0062] Select the optimal model parameters by gradient descent and make Minimum, where n represents the number of radial data used for calculation, and the horizontal wind speed in the area is inverted , horizontal wind direction , wind speed in x direction , wind speed in the y direction .

[0063] in, ;

[0064] ;

[0065] .

[0066] In the application of the VAD wind field inversion method based on gradient descent, the initial values ​​of the model parameters, the number of iterations, and the scanning range of the azimuth angle will affect the accuracy of wind speed and wind direction inversion. Especially when the wind direction is close to perpendicular to the light beam direction, the selection of the initial value will bring about a large wind field inversion error. In order to solve this problem, the horizontal wind field is inverted by using a beamNum beam sliding window. The horizontal wind speed and horizontal wind direction calculated in the previous sliding window are used as the initial values ​​of the model parameters and substituted into the gradient descent solution model of the current beam, thereby inverting the horizontal wind speed and horizontal wind direction of the current beam.

[0067] Therefore, S2 includes the following specific steps:

[0068] S21. A dynamic model is established based on the VAD wind field inversion method based on gradient descent.

[0069] S22, calculating the horizontal wind speed and horizontal wind direction of the last sliding window, and using the horizontal wind speed and horizontal wind direction of the last sliding window as initial value parameters of the dynamic model, wherein the last sliding window is the N light beams formed in the last cycle.

[0070] S23. Horizontal wind speed and horizontal wind direction of the inversion beam.

[0071] In the present invention, N=9, combined with Figure 3-6 , Figure 3 and Figure 4 It shows the change of the center beam of the radial data set before and after sliding. Figure 5 and Figure 6 The changes in horizontal wind direction inverted using fixed model parameters and dynamic model parameters are shown. It can be seen from the figure that the horizontal wind direction in the sector inverted using fixed model parameters is quite different from the true wind direction, while the horizontal wind direction inverted using the horizontal wind speed and horizontal wind direction calculated by the previous sliding window as dynamic model parameters is the same as the true wind direction.

[0072] S3. Divide the horizontal wind field into grids, and interpolate based on the detection distance to obtain the wind speed component in the X direction of the interpolation point. , Y direction wind speed component .

[0073] The following steps are involved:

[0074] S31. Confirm that the horizontal wind speed is not the maximum detection distance maxDist of Nan.

[0075] S32. Divide a square grid with the maximum detection distance maxDist as the boundary; divide a square grid M×N with a grid resolution of dx, a grid size of (2×maxDist / dx+1)× (2×maxDist / dx+1), a horizontal distance X range of the grid is [-maxDist, maxDist], and a vertical distance Y range of [-maxDist, maxDist].

[0076] Among them, the detection distance in the X direction , Y direction detection distance They are:

[0077] ,

[0078] .

[0079] S33, reverse detection distance Location and detection distance Wind speed component at , , and insert it into the corresponding interpolation point (Xm, Yn) of the square grid.

[0080] S34, the maximum distance from the interpolation point (Xm, Yn) to the spatial adjacent point set (Xgm, Ygn) is MaxD;

[0081] in, .

[0082] Traverse all grid point positions (Xg, Yg) whose distance to the interpolation point is less than the maximum distance MaxD, and the corresponding wind speed component is and .

[0083] S36, according to

[0084] ,

[0085] ,

[0086] ,

[0087] in, is the weight corresponding to the interpolation point (Xm, Yn).

[0088] Calculate the weighted average of the wind speed components at the interpolation point and .

[0089] S4. For the wind speed component of the interpolation point, linear fitting is performed based on the moving window method to obtain the wind speed along the X direction. Rate of change , along the Y direction Rate of change , along the X direction Rate of change , along the Y direction Rate of change ;in, is the wind speed component in the X direction The amount of change in the X direction; is the wind speed component in the Y direction The amount of change in the Y direction; is the change distance in the X direction; is the change distance in the Y direction.

[0090] The size of the moving window is p×q, and in the example of the present invention, p=q=5. Figure 7 shown.

[0091] S5. According to , the wind shear factor is calculated .

[0092] Embodiment 2: A PPI wind shear extraction device based on laser wind radar, the device comprising one or more processors and a memory, the memory being used to store one or more programs; when the one or more programs are executed by the processor, the processor is enabled to implement a PPI wind shear extraction method based on laser wind radar.

[0093] Embodiment 3: A computer storage medium stores at least one program instruction, and the at least one program instruction is used to be loaded and executed by a processor to implement a PPI wind shear extraction method based on a laser wind radar.

[0094] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the above embodiments do not limit the present invention in any form, and any technical solution obtained by equivalent replacement or equivalent transformation falls within the protection scope of the present invention.

Claims

1. A PPI wind shear extraction method based on a laser wind radar, characterized in that: The following steps are involved: S1. Based on PPI scanning, radial data of N light beams in one cycle are obtained, and the radial data are screened and marked. The radial data includes radial wind speed V. When the value of radial wind speed V of a light beam in the cycle is an invalid value Nan, the light beam is not inverted, that is, the horizontal wind speed of the light beam is an invalid value Nan; S2, invert the horizontal wind field based on the beam sliding window to obtain the horizontal wind speed and horizontal wind direction of the beam; The specific steps of inverting the horizontal wind field include: S21. Establish a dynamic model based on the VAD wind field inversion method based on gradient descent; S22, calculating the horizontal wind speed and horizontal wind direction of the last sliding window, and using the horizontal wind speed and horizontal wind direction of the last sliding window as initial value parameters of the dynamic model; the last sliding window is the N light beams formed in the last cycle; S23, horizontal wind speed and horizontal wind direction of the inversion beam; S3. Divide the horizontal wind field into grids, and interpolate based on the detection distance to obtain the wind speed component in the X direction of the interpolation point. , Y direction wind speed component ; S4. For the wind speed component of the interpolation point, linear fitting is performed based on the moving window method to obtain the wind speed along the X direction. Rate of change , along the Y direction Rate of change , along the X direction Rate of change , along the Y direction Rate of change ;in, is the wind speed component in the X direction The amount of change in the X direction; is the wind speed component in the Y direction The amount of change in the Y direction; is the change distance in the X direction; is the change distance in the Y direction; S5. According to , the wind shear factor is calculated .

2. A PPI wind shear extraction method based on laser wind radar according to claim 1, characterized in that: The S1 comprises the following specific steps: S11, when the radar is scanning based on PPI, the elevation angle remains unchanged and the azimuth angle is adjusted step by step; S12, obtaining radial data of N light beams in one cycle; the radial data includes: radial distance Dist, azimuth angle Azi; N is an odd number; S13, obtaining the maximum range Range of the scanning azimuth angle in the cycle and the azimuth angle deviation bias of a single light beam; S14, taking the light beam at the (N+1) / 2 position as the central wave, and obtaining the azimuth angle Azi_cnt of the central wave; S15. Filter out the satisfied The radial data within the central beam sector is obtained.

3. The PPI wind shear extraction method based on laser wind radar according to claim 1 is characterized in that: The S3 comprises the following steps: S31, confirming that the horizontal wind speed is not the maximum detection distance maxDist of Nan; S32, divide the square grid with the maximum detection distance maxDist as the boundary; wherein the detection distance in the X direction is , Y direction detection distance ; ， , where Dist is the detection distance; S33, reverse detection distance Location and detection distance Wind speed component at , , and insert it into the corresponding interpolation point (Xm, Yn) of the square grid; S34, the maximum distance from the interpolation point (Xm, Yn) to the spatial adjacent point set (Xgm, Ygn) is MaxD; in, ; S35. Traverse all grid point position sets (Xg, Yg) whose distance to the interpolation point is less than the maximum distance MaxD, and the corresponding wind speed component is and ; S36, according to ， ， ， in, is the weight corresponding to the interpolation point (Xm, Yn); Calculate the weighted average of the wind speed components at the interpolation point and .

4. A PPI wind shear extraction device based on laser wind radar, characterized in that: The device includes one or more processors and a memory, wherein the memory is used to store one or more programs; when the one or more programs are executed by the processor, the processor is enabled to implement a PPI wind shear extraction method based on a laser wind radar as described in any one of claims 1 to 3.

5. A computer storage medium, characterized in that: The computer storage medium stores at least one program instruction, and the at least one program instruction is used to be loaded and executed by the processor to implement the PPI wind shear extraction method based on laser wind radar as described in any one of claims 1-3.

Images