Thunderstorm cloud spatial and temporal distribution identification method based on image processing

Abstract

The invention provides a thunderstorm cloud spatial and temporal distribution identification method based on image processing, belongs to the field of novel power system protection, and mainly utilizes the characteristic that thunderbolt continuously generated in the continuous development process of thunderstorm cloud still presents concentrated distribution characteristics in longitude and latitude distribution characteristics. After a thundercloud thunderbolt data set is subjected to time scale stripping, thundercloud activity ranges represented by high-density thunderbolt points are integrated by utilizing a morphological processing means to obtain development time and activity space of each thundercloud, and thundercloud activity indexes acquired by a thunderbolt positioning system are utilized to obtain thundercloud thunderbolt data. And carrying out real-time judgment on thundercloud risk capable of causing lightning stroke of the fan matrix of the new energy power plant. The invention aims to solve the defects that a high-dimensional data set cannot be identified and thundercloud has space stacking in different time periods and is difficult to extract in the current generally applied density clustering method, so that the accuracy of thundercloud development time and activity range extraction is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of new power system protection, and particularly to a method for identifying the spatio-temporal distribution of thunderstorm clouds based on image processing. Background Art

[0002] With the global climate anomaly, the occurrence ratios of extreme weather such as typhoons and thunderstorms are increasing year by year. With the advent of the era of high-proportion new energy development, active lightning protection methods for overhead lines and wind turbine photovoltaic arrays against thunderstorm activities have attracted more and more attention. Closely monitoring the development status and activity range of thunderstorm clouds can establish an emergency plan for lightning strike warning of power facilities, reduce the subsequent risks caused by lightning strike faults, conform to the concept of active lightning protection for transmission lines, and continuously escort the country's power development.

[0003] Currently, the method for determining the thundercloud range based on the lightning strike point generally applies density clustering. Considering that the lightning strike data is the data stacking at different times, density clustering cannot be applied to process high-dimensional data sets, and there are disadvantages in that it is difficult to separate the spatio-temporal stacked lightning strike data. Therefore, a method for identifying the spatio-temporal distribution of thunderstorm clouds based on image processing is proposed to solve the above problems. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a method for identifying the spatio-temporal distribution of thunderstorm clouds based on image processing to solve at least the above problems.

[0005] The technical solution adopted by the present invention is as follows.

[0006] A method for identifying the spatio-temporal distribution of thunderstorm clouds based on image processing, the method comprising the following steps:

[0007] S1. Perform data stripping of longitude and latitude in the time dimension for the continuous lightning strike distribution of thunderclouds in a three-dimensional spatio-temporal window;

[0008] S2. Perform edge detection on the distribution images of the lightning strike point longitude and latitude time series;

[0009] S3. Perform morphological processing on the lightning strike longitude image and the lightning strike latitude image after edge detection respectively;

[0010] S4. Integrate the pixel points of the processed lightning strike longitude image and the lightning strike latitude image with the time and longitude-latitude distribution of the original picture;

[0011] S5. Extract the number of discharges continuously generated by each thundercloud;

[0012] S6. Extract the average absolute value of the lightning discharge current of each thundercloud;

[0013] S7. Extract the distance between the centroid of each thundercloud and a certain wind turbine in the wind farm;

[0014] S8. Set risk weights for the number of continuous discharges, the average absolute value of the discharge lightning current, and the distance index between the centroid and a certain wind turbine in the wind farm.

[0015] S9. Calculate the disaster-causing risk levels of each thundercloud within the window.

[0016] Further, in step S1, the data stripping of longitude and latitude in the time dimension for the continuous lightning strike distribution of thunderclouds under the three-dimensional spatio-temporal window is specifically as follows:

[0017] For the continuous lightning strike distribution of thunderclouds under the three-dimensional spatio-temporal window, from the perspective of the time dimension and based on the time scale, separate the lightning strike longitude distribution and latitude distribution under the spatio-temporal distribution, and obtain and draw the longitude and latitude time distribution images.

[0018] Further, in step S2, it also includes sub-steps:

[0019] S21. Eliminate the image hue information by calculating the weighted sum of the R, G, and B components, and convert the color image of the longitude and latitude distribution in the time dimension into a grayscale image to complete the grayscale processing of the image. The operation method is as follows:

[0020] gray = 0.229R + 0.587G + 0.114B

[0021] S22. Use the Sobel operator for edge detection. The operator depends on two 3x3 convolution kernels, and completes the edge detection of the image by calculating the brightness differences in the horizontal and vertical directions of the image respectively. The Sobel operator expression is:

[0022]

[0023] where f(x, y) represents the grayscale information of the image.

[0024] Further, in step S3, it also includes sub-steps:

[0025] S31. Use the opening operation to smooth the image contour. The formula is as follows:

[0026]

[0027] The above formula means that the structuring element B first erodes the structuring element A, and then dilates the erosion result;

[0028] S32. Use the closing operation to bridge the holes in the image, and fill the gaps in the image contour while eliminating small holes. The formula is as follows:

[0029]

[0030] The above formula represents that structuring element B first dilates structuring element A and then erodes the dilation result;

[0031] S33. Facing the background region, all holes will be further filled with 1, and the formula is as follows

[0032] X k = (X k-1 ⊕ B) ∩ I c , k = 1, 2, 3,...

[0033] I represents the pre - image element, and I c is its complement, B represents the symmetric structuring element. The above formula includes all the filled holes and the boundaries of these holes.

[0034] A = X k ∪ I, k = 1, 2, 3,...

[0035] Furthermore, in step S4, it also includes sub - steps:

[0036] S41. Obtain the longitude activity range window of each thundercloud using the correspondence between the pixel range and the longitude - latitude range in the time series, as shown in the following formula:

[0037]

[0038] where [Lon min (i), Lon max (i)] is the true longitude coverage range of thundercloud i, [pix Lonmin (i), pix Lonmax (i)] is the pixel segment covered by the longitude of thundercloud i in the image after morphological processing, and [pix Lonmin , pix Lonmax is the total pixel segment of the longitude in the processed image;

[0039] S42. Obtain the latitude activity range window of each thundercloud using the correspondence between the pixel range and the longitude - latitude range in the time series, as shown in the following formula:

[0040]

[0041] where [Lat min (i), Lat max (i)] is the true latitude coverage range of thundercloud i, [pix Latmin (i), pix Latmax (i)] is the pixel segment covered by the latitude of thundercloud i in the image after morphological processing, and [pix Latmin , pix Latmax is the total pixel segment of the latitude in the processed image;

[0042] S43. Obtain the development time window of each thundercloud based on the relationship between the pixel range and the longitude and latitude range, as shown in the following formula:

[0043]

[0044] Among them, [T min (i), T max (i)] is the development time interval of thundercloud i, [T min , T max is the entire time window, [pix Tmin (i), pix Tmax (i)] is the pixel section of thundercloud i in the time dimension in the image after morphological processing, [pix Tmin , pix Tmax represents the pixel range of the entire time period.

[0045] Furthermore, in step S5, the specific method for extracting the number of discharges continuously generated by each thundercloud is as follows:

[0046] The number of discharges continuously generated by each thundercloud is represented by N i (j), which means that thundercloud i continuously generates j lightning strikes.

[0047] Furthermore, in step S6, the specific method for extracting the average absolute value of the discharge lightning current of each thundercloud is as follows:

[0048] The average absolute value of the lightning current of different cloud clusters is expressed as:

[0049]

[0050] Among them, i is the number of cloud clusters screened within the window, L(N i (j)) is the lightning current amplitude of the jth cloud-to-ground flash of thundercloud i within the window. 8. A method for identifying the spatio-temporal distribution of thunderstorm clouds based on image processing according to claim 7, characterized in that in step S7, it further includes sub-steps:

[0051] S71. Extract the position coordinates of each cloud cluster:

[0052]

[0053] Among them, (S(i), Q(i)) represents the position coordinates of thundercloud i, (S(N i (j)), Q(N i (j))) represents the position of the jth lightning strike of thundercloud i;

[0054] S72. Calculate the horizontal distance between each thundercloud and a certain wind turbine of the new energy wind farm:

[0055]

[0056] (S m ,Q m ) represents the position coordinates of a certain wind turbine, m is the number of the wind turbine, and d is the distance matrix between the position of thundercloud j and any wind turbine;

[0057] S73. Draw the distance matrix between thundercloud i and wind turbine j, and represent it with d:

[0058]

[0059] The closest distance between each thundercloud and the wind turbine is expressed as D = (mind(1) m , mind(2) m ,..., mind(i) m ).

[0060] Furthermore, in step S8, risk weights are specifically set for the number of discharges generated continuously, the average absolute value of the discharge lightning current, and the distance index between the centroid and a certain wind turbine in the wind farm:

[0061] For the number of discharges generated continuously, the average absolute value of the discharge lightning current, and the distances between the thundercloud and the centroid and a certain wind turbine in the wind farm, the weights are represented by ω1, ω2, and ω3 respectively. Considering that the horizontal distance between the cloud mass and the wind turbine largely determines the disaster-causing risk, and more lightning strikes continuously generated by the thundercloud will increase the risk probability of lightning striking the wind turbine, the weights of ω1, ω2, and ω3 are set to 0.2, 0.3, and 0.5 respectively.

[0062] Furthermore, in step S9, it also includes sub-steps:

[0063] S91. The risk coefficients of each index will be allocated according to the normalization results of the corresponding index data of different cloud masses:

[0064]

[0065] S92. Calculate the disaster-causing risks of each thundercloud i within the window:

[0066] T = ω1l(i) + ω2N i (j) + ω3D'.

[0067] Compared with the prior art, the beneficial effects of the present invention are

[0068] The present invention proposes a method for identifying and determining the spatio-temporal distribution of high-risk thunderstorm clouds based on image processing. The present invention can solve the drawbacks that the currently widely used density clustering method cannot identify high-dimensional data sets and it is difficult to extract the spatial stacking of thunderclouds in different time periods, thereby improving the accuracy of extracting the development time and activity range of thunderclouds. Description of the Drawings

[0069] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only the preferred embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0070] Figure 1 It is a schematic diagram of the overall process structure of a method for identifying the spatio-temporal distribution of thunderstorm clouds based on image processing according to an embodiment of the present invention. Detailed Embodiments

[0071] In order to make the objectives, technical solutions and advantages of the present invention more apparent, the following will describe exemplary embodiments according to the present invention in detail with reference to the drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments of the present invention. It should be understood that the present invention is not limited by the exemplary embodiments described herein. Based on the embodiments of the present invention described herein, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0072] In the following description, numerous specific details are given to provide a more thorough understanding of the present invention. However, it is obvious to those of ordinary skill in the art that the present invention can be implemented without one or more of these details. In other instances, in order to avoid confusion with the present invention, some well-known technical features are not described.

[0073] It should be understood that the present invention can be implemented in different forms and should not be construed as limited to the embodiments presented herein. On the contrary, providing these embodiments will make the disclosure thorough and complete, and will fully convey the scope of the present invention to those of ordinary skill in the art.

[0074] The purpose of the terms used herein is only to describe specific embodiments and is not a limitation of the present invention. When used herein, the singular forms "a", "an" and "the" are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the terms "comprising" and / or "including", when used in this specification, identify the presence of the stated features, integers, steps, operations, elements and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups. When used herein, the term "and / or" includes any and all combinations of the related listed items.

[0075] The principles and features of the present invention will be described below in conjunction with the accompanying drawings. The listed embodiments are only used to explain the present invention and are not intended to limit the scope of the present invention.

[0076] Referring to Figure 1 , the present invention provides a method for identifying the spatio-temporal distribution of thunderstorm clouds based on image processing. The method includes the following steps:

[0077] S1. Perform data stripping of longitude and latitude in the time dimension for the continuous lightning strike distribution of thunderclouds in a three-dimensional spatio-temporal window;

[0078] S2. Perform edge detection on the distribution images of the lightning strike points in the longitude and latitude time series;

[0079] S3. Perform morphological processing on the lightning strike longitude image and the lightning strike latitude image after edge detection respectively;

[0080] S4. Integrate the pixel points of the processed lightning strike longitude image and the lightning strike latitude image with the time and longitude-latitude distribution of the original picture;

[0081] S5. Extract the number of discharges continuously generated by each thundercloud;

[0082] S6. Extract the average absolute value of the discharge lightning current of each thundercloud;

[0083] S7. Extract the distance between the centroid of each thundercloud and a certain wind turbine in the wind farm;

[0084] S8. Set risk weights for the number of discharges continuously generated, the average absolute value of the discharge lightning current, and the distance between the centroid and a certain wind turbine in the wind farm;

[0085] S9. Calculate the disaster-causing risk level of each thundercloud in the window.

[0086] In step S1, specifically for the continuous lightning strike distribution of thunderclouds in a three-dimensional spatio-temporal window, data stripping of longitude and latitude in the time dimension is as follows:

[0087] For the continuous lightning strike distribution of thunderclouds in a three-dimensional spatio-temporal window, from the perspective of the time dimension and based on the time scale, separate the lightning strike longitude distribution and the latitude distribution in the spatio-temporal distribution, and obtain and draw the longitude and latitude time distribution images.

[0088] In step S2, it further includes sub-steps:

[0089] S21. By calculating the weighted sum of the R, G, and B components, eliminate the image hue information, and convert the color image of the longitude and latitude distribution in the time dimension into a grayscale image to complete the grayscale processing of the image. The operation method is as follows:

[0090] gray = 0.229R + 0.587G + 0.114B

[0091] S22. Edge detection is performed using the Sobel operator. The operator depends on two 3x3 convolution kernels and completes the edge detection of the image by calculating the brightness differences in the horizontal and vertical directions of the image respectively. The Sobel operator expression is as follows:

[0092]

[0093] Among them, f(x, y) represents the grayscale information of the image.

[0094] In step S3, it also includes sub-steps:

[0095] S31. Use opening operation to smooth the image contour. The formula is as follows:

[0096]

[0097] The above formula means that structuring element B first erodes structuring element A and then dilates the erosion result;

[0098] S32. Use closing operation to close the holes in the image, filling the gaps in the image contour while eliminating small holes. The formula is as follows:

[0099]

[0100] The above formula means that structuring element B first dilates structuring element A and then erodes the dilation result;

[0101] S33. For the background area, further fill all holes with 1. The formula is as follows

[0102] X k =(X k-1 ⊕B)∩I c , k = 1, 2, 3,...

[0103] I represents the pre-image element, and I c is its complement. B represents the symmetric structuring element. The above formula includes all filled holes and the boundaries of these holes.

[0104] A = X k ∪I, k = 1, 2, 3,...

[0105] In step S4, it also includes sub-steps:

[0106] S41. Obtain the longitude activity range window of each thundercloud using the corresponding relationship between the pixel range and the longitude and latitude range in the time series, as shown in the following formula:

[0107]

[0108] Among them, [Lon min (i), Lon max (i)] is the true longitude coverage range of thundercloud i, and [pix Lonmin (i), pix Lonmax (i)] is the pixel segment covered by the longitude in the image of thundercloud i after morphological processing. [pix Lonmin , pix Lonmax is the total pixel segment of the longitude in the processed image;

[0109] S42. Obtain the latitude activity range window of each thundercloud by using the corresponding relationship between the pixel range and the longitude and latitude range under the time series, as shown in the following formula:

[0110]

[0111] Among them, [Lat min (i), Lat max (i)] is the true latitude coverage range of thundercloud i, and [pix Latmin (i), pix Latmax (i)] is the pixel segment covered by the latitude in the image of thundercloud i after morphological processing. [pix Latmin , pix Latmax is the total pixel segment of the latitude in the processed image;

[0112] S43. Obtain the development time window of each thundercloud by using the relationship between the pixel range and the longitude and latitude range, as shown in the following formula:

[0113]

[0114] Among them, [T min (i), T max (i)] is the development time interval of thundercloud i, [T min , T max is the entire time window, [pix Tmin (i), pix Tmax (i)] is the pixel section of thundercloud i in the time dimension in the image after morphological processing, and [pix Tmin , pix Tmax represents the pixel range of the entire time period.

[0115] In step S5, the specific number of discharges continuously generated by each thundercloud is extracted as follows:

[0116] The number of discharges continuously generated by each thundercloud is represented by N i (j), which means that thundercloud i continuously generates j lightning strikes.

[0117] In step S6, the specific method for extracting the average absolute value of the discharge lightning current of each thundercloud is as follows:

[0118] The average absolute value of the lightning current of different cloud clusters is expressed as:

[0119]

[0120] where i is the number of cloud clusters screened within the window, and L(N i (j)) is the lightning current amplitude of the j-th ground flash of the i-th thundercloud within the window. 8. A method for identifying the spatio-temporal distribution of thunderstorm clouds based on image processing according to claim 7, characterized in that in step S7, it further includes sub-steps:

[0121] S71. Extract the position coordinates of each cloud cluster:

[0122]

[0123] where (S(i), Q(i)) represents the position coordinates of thundercloud i, and (S(N i (j)), Q(N i (j))) represents the lightning strike position of the j-th time of thundercloud i;

[0124] S72. Calculate the horizontal distance between each thundercloud and a certain wind turbine in the new energy wind farm:

[0125]

[0126] (S m , Q m ) represents the position coordinates of a certain wind turbine, m is the number of the wind turbine, and d is the distance matrix between the position of thundercloud j and any wind turbine;

[0127] S73. Draw the distance matrix between thundercloud i and wind turbine j, and represent it with d:

[0128]

[0129] The closest distance between each thundercloud and the wind turbine is expressed as D = (mind(1) m , mind(2) m ,..., mind(i) m ).

[0130] In step S8, risk weights are set for the continuously generated number of discharges, the average absolute value of the discharge lightning current, and the distance index between the centroid and a certain wind turbine in the wind farm, specifically:

[0131] For the continuously generated number of discharges, the weights of the average absolute value of the discharge lightning current and the distances between the thundercloud and the centroid and a certain wind turbine in the wind farm are represented by ω1, ω2, and ω3 respectively. Considering that the horizontal distance between the cloud cluster and the wind turbine largely determines the disaster-causing risk, and more lightning strikes continuously generated by the thundercloud will increase the risk probability of lightning striking the wind turbine, the weights of ω1, ω2, and ω3 are set to 0.2, 0.3, and 0.5 respectively.

[0132] In step S9, it further includes sub-steps:

[0133] S91. The risk coefficients of each index will be allocated according to the normalization results of the corresponding index data of different cloud clusters:

[0134]

[0135] S92. Calculate the disaster-causing risk of each thundercloud i within the window:

[0136] T = ω1l(i)+ω2N i (j)+ω3D'.

[0137] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for identifying the spatio-temporal distribution of thunderstorm clouds based on image processing, characterized in that, The method includes the following steps: S1. Perform data stripping of longitude and latitude in the time dimension for the continuous lightning strike distribution of thunderclouds under a three-dimensional spatio-temporal window; S2. Perform edge detection on the distribution images of the longitude and latitude time series of lightning strike points; S3. Perform morphological processing on the lightning strike longitude image and lightning strike latitude image after edge detection respectively; S4. Integrate the pixel points of the processed lightning strike longitude image and lightning strike latitude image with the time and longitude-latitude distribution of the original picture; S5. Extract the number of continuous discharges generated by each thundercloud; S6. Extract the average absolute value of the discharge lightning current of each thundercloud; S7. Extract the distance between the centroid of each thundercloud and a certain wind turbine in the wind farm; S8. Set risk weights for the number of continuous discharges generated, the average absolute value of the discharge lightning current, and the distance between the centroid and a certain wind turbine in the wind farm; S9. Calculate the disaster-causing risk level of each thundercloud within the window.

2. The method for identifying the spatio-temporal distribution of thunderstorm clouds based on image processing according to claim 1, characterized in that, In step S1, the specific operation of performing data stripping of longitude and latitude in the time dimension for the continuous lightning strike distribution of thunderclouds under a three-dimensional spatio-temporal window is as follows: For the continuous lightning strike distribution of thunderclouds under a three-dimensional spatio-temporal window, from the perspective of the time dimension and based on the time scale, separate the lightning strike longitude distribution and latitude distribution under the spatio-temporal distribution to obtain and draw the longitude and latitude time distribution images.

3. The method for identifying the spatio-temporal distribution of thunderstorm clouds based on image processing according to claim 2, characterized in that, In step S2, it also includes sub-steps: S21. Eliminate the image hue information by calculating the weighted sum of R, G, and B components, and convert the color image of the longitude and latitude distribution in the time dimension into a grayscale image to complete the grayscale processing of the image. The operation method is as follows: gray = 0.229R + 0.587G + 0.114B S22. Use the Sobel operator for edge detection. The operator depends on two 3x3 convolution kernels, and completes the detection of the image edge by calculating the brightness difference of the image in the horizontal and vertical directions respectively. The Sobel operator expression is: Among them, f(x, y) represents the grayscale information of the image.

4. The method for identifying the spatio-temporal distribution of thunderstorm clouds based on image processing according to claim 3, wherein In step S3, it also includes sub-steps: S31. Use the opening operation to smooth the image contour. The formula is as follows: The above formula means that the structuring element B first erodes the structuring element A, and then dilates the erosion result; S32. Use the closing operation to bridge the holes in the image, and fill the gaps in the image contour while eliminating small holes. The formula is as follows: The above formula means that the structuring element B first dilates the structuring element A, and then erodes the dilation result; S33. For the background area, further fill all holes with 1. The formula is as follows Let \(I\) denote the foreground image element, \(I\) c is its complement set, and \(B\) represents a symmetric structuring element. The above formula includes all filled holes and the boundaries of these holes. A = X k ∪I, k = 1, 2, 3,...。 5. A method for identifying the spatio-temporal distribution of thunderstorm clouds based on image processing according to claim 4, characterized in that, In step S4, it also includes sub-steps: S41. Obtain the longitude activity range window of each thundercloud by using the corresponding relationship between the pixel range and the longitude-latitude range in the time series, as shown in the following formula: Among them, [Lon min (i), Lon max (i)] is the true longitude coverage range of thundercloud i, and [pix Lonmin (i), pix Lonmax (i)] is the pixel segment covered by the longitude of thundercloud i in the image after morphological processing. [pix Lonmin , pix Lonmax is the total pixel segment of the longitude in the processed image; S42. Obtain the latitude activity range window of each thundercloud by using the corresponding relationship between the pixel range and the longitude-latitude range in the time series, as shown in the following formula: Among them, [Lat min (i), Lat max (i)] is the actual latitude coverage range of thundercloud i, and [pix Latmin (i), pix Latmax (i)] is the pixel segment covered by the latitude in the image of thundercloud i after morphological processing. [pix Latmin , pix Latmax is the total pixel segment of the latitude in the processed image; S43. Obtain the development time window of each thundercloud by using the relationship between the pixel range and the longitude-latitude range, as shown in the following formula: Among them, [T min (i), T max (i)] is the development time interval of thundercloud i, [T min , T max is the entire time window, [pix Tmin (i), pix Tmax (i)] is the pixel section of thundercloud i in the time dimension in the image after morphological processing, [pix Tmin , pix Tmax represents the pixel range of the entire time period.

6. The method for identifying the spatio-temporal distribution of thunderstorm clouds based on image processing according to claim 5, wherein, In step S5, the specific operation of extracting the number of continuous discharges generated by each thundercloud is as follows: Use N to extract the number of discharges continuously generated by each thundercloud i It is represented by (j), which means that thundercloud i continuously generates j lightning strikes.

7. A method for identifying the spatio-temporal distribution of thunderstorm clouds based on image processing according to claim 6, characterized in that, In step S6, the specific operation of extracting the average absolute value of the discharge lightning current of each thundercloud is as follows: The average absolute value of the lightning current of different cloud clusters is expressed as: where i is the number of cloud clusters selected within the window, and L(N i (j)) is the lightning current amplitude of the j-th ground flash of the i-th thundercloud within the window.

8. A method for identifying the spatio-temporal distribution of thunderstorm clouds based on image processing according to claim 7, characterized in that, In step S7, it further includes sub-steps: S71. Extract the position coordinates of each cloud cluster: Among them, (S(i), Q(i)) represents the position coordinates of thundercloud i, and (S(N i (j)), Q(N i (j))) represents the lightning strike position of thundercloud i at the j-th time; S72. Calculate the horizontal distance between each thundercloud and a certain wind turbine in the new energy wind farm: (S m ,Q m ) represents the position coordinates of a certain wind turbine, m is the number of the wind turbine, and d is the distance matrix between the position of thundercloud j and any wind turbine; S73. Draw the distance matrix between thundercloud i and wind turbine j, and denote it with d: The closest distance between each thundercloud and the wind turbine is expressed as D=(mind(1) m ,mind(2) m ,...,mind(i) m ).

9. A method for identifying the spatio-temporal distribution of thunderstorm clouds based on image processing according to claim 8, characterized in that, In step S8, risk weights are set for the continuously generated discharge times, the average absolute value of the discharge lightning current, and the distance index between the centroid and a certain wind turbine in the wind farm. Specifically: For the continuously generated discharge times, the average absolute value of the discharge lightning current, and the distances between the thundercloud and the centroid and a certain wind turbine in the wind farm, the weights are represented by ω1, ω2, and ω3 respectively. Considering that the horizontal distance between the cloud cluster and the wind turbine largely determines the disaster-causing risk, and more lightning strikes continuously generated by the thundercloud will increase the risk probability of lightning striking the wind turbine, the weights of ω1, ω2, and ω3 are set to 0.2, 0.3, and 0.5 respectively.

10. A method for identifying the spatio-temporal distribution of thunderstorm clouds based on image processing according to claim 1, characterized in that, In step S9, it further includes sub-steps: S91. The risk coefficients of each index will be allocated according to the normalization results of the corresponding index data of different cloud clusters: S92. Calculate the disaster-causing risk of each thundercloud i within the window: T = ω1l(i) + ω2N i (j) + ω3D'.

Images