A comparative evaluation method for offshore wind energy based on ocean cyclone removal

Abstract

The present invention discloses a comparative evaluation method for offshore wind energy based on the removal of ocean cyclones. The method utilizes reanalysis data and a cyclone tracking algorithm to extract ocean cyclone-related parameters; establishes an impact buffer zone according to the cyclone parameters; obtains the process quantity of sea level pressure during the cyclone removal process; obtains the sea level pressure after the cyclone is removed; constructs a surface wind field model for a coastal zone; obtains the surface wind field for a coastal zone without the influence of a cyclone; comparatively evaluates offshore wind energy based on the actual wind field for the coastal zone and the simulated wind field under the assumption of no cyclone; the present invention can conduct comparative evaluation on the influence of ocean cyclone influence on offshore wind energy, and has a reference role in the planning and construction of offshore wind power projects.

Description

Technical Field

[0001] The invention belongs to the technical field of offshore wind energy research, and in particular relates to an offshore wind energy comparative evaluation method based on ocean cyclone removal. Background Art

[0002] As a clean energy, offshore wind energy plays an important role in low-carbon environmental protection, energy conservation and emission reduction, and is vigorously promoted; the planning and construction of offshore wind power projects are closely related to offshore wind farms, and the frequent marine cyclones in recent years have had a great impact on the surface wind farms in the near-coastal zone; in the prior art, there are currently only related methods for tropical cyclone path analysis in wind farm areas, such as the invention patent with patent application number "202010840353.8" and the name "A path analysis method for tropical cyclones affecting offshore wind farms", which discloses a path analysis method for tropical cyclones affecting offshore wind farms. A tropical cyclone statistical area with a radius of 100 to 300 km is set with the center of the offshore wind farm site as the center of the circle. domain, and take this statistical area as the range where wind farms may be affected by tropical cyclones, and count all historical tropical cyclones that have passed through this area; calculate the tropical cyclone path deflection parameters, and count the path deflection of all tropical cyclones entering the statistical area according to whether they are deflected or not; statistically analyze the path, frequency, interannual and intermonthly changes, and intensity change characteristics of tropical cyclones that affect offshore wind farms; this method can effectively calculate the path of tropical cyclones entering the region, so as to feed back to the selection and design of wind farm units to minimize the impact of cyclones; but this method cannot effectively simulate and eliminate marine cyclones, so as to intuitively compare data and analyze the precise numerical value of the impact of marine cyclones on wind farms in the selected area.

[0003] In order to evaluate the impact of marine cyclones on offshore wind energy and further optimize the layout planning of offshore wind power projects, this study generated the sea level pressure field under the assumption of no cyclone based on relevant cyclone parameters and reanalysis data, and constructed a fitting model of the surface wind field in the near-coastal zone. The surface wind field in the near-coastal zone with and without cyclones was analyzed, and the impact of marine cyclones on offshore wind energy was compared and evaluated, providing a reference for the planning of offshore wind power projects. Summary of the invention

[0004] In order to achieve the above technical effects, the technical solution adopted by the present invention is:

[0005] A comparative evaluation method for offshore wind energy based on ocean cyclone removal comprises the following steps:

[0006] S1, extracting ocean cyclone related parameters: using reanalysis data combined with cyclone tracking algorithm, extracting relevant parameters including cyclone time period T, center position and trajectory P, cyclone radius R, original sea level pressure SLP_ori, and classifying and storing each parameter and recording it in the geographic information grid;

[0007] S2, establish a buffer zone for the cyclone's impact range: set up two buffer zone coefficients based on the cyclone's center position and trajectory P, as well as the cyclone's radius R, and establish two buffer zones Buffer_1 and Buffer_2 corresponding to the cyclone; the specific method is as follows:

[0008] S201. Establish the first buffer zone Buffer_1 within the cyclone impact range:

[0009] With the position trajectory P of the cyclone center as the center and C_1*cyclone radius R as the side length, a square buffer zone Buffer_1 corresponding to the cyclone is generated. C_1 is the buffer zone coefficient 1. According to the experiment, the value C_1=1.75;

[0010] S202, establish the first buffer zone Buffer_2 within the cyclone impact range:

[0011] With the position trajectory P of the cyclone center as the center and C_2*cyclone radius R as the side length, a square buffer zone Buffer_2 corresponding to the cyclone is generated. C_2 is the buffer zone coefficient 2. According to the experiment, the value C_2=2;

[0012] S3, obtain the sea level pressure process quantity during the cyclone removal process:

[0013] S301, select the research time and space range, obtain the cyclone time period T and buffer zone Buffer_1;

[0014] S302, calculate the monthly climatic mean sea level pressure SLP_clim in the corresponding time and space range, and use SLP_clim to replace SLP_ori, recorded as SLP_fix;

[0015] S303, calculating the sea level pressure process quantity SLP_Temp according to SLP_ori and SLP_fix;

[0016] S304, reprocessing SLP_Temp according to SLP_ori, SLP_clim, Buffer_1, and Buffer_2 to calculate SLP_Temp_s1; processing SLP_Temp_s1 using Gaussian filtering, and recording the processed SLP_Temp_s2;

[0017] S4, get the sea level pressure SLP_revomeCyc after the cyclone is removed:

[0018] S401, calculate the difference between SLP_Temp_s2 and SLP_ori, recorded as SLP_dis;

[0019] S402, for the area where SLP_dis<0, reassign the value to SLP_ori, and obtain the sea level pressure SLP_revomeCyc after the cyclone is removed;

[0020] S5, construct the surface wind field model near the coast:

[0021] Taking the original sea level pressure SLP_ori as input and the actual wind field near the coast as output, a surface wind field model near the coast is constructed using deep learning.

[0022] S6, obtain the surface wind field near the coast without the influence of cyclone:

[0023] Based on the coastal surface wind field model constructed in S5, SLP_revomeCyc is used as input to obtain the coastal wind field conditions under the assumption of no cyclone, including wind speed and wind direction;

[0024] S7, Comparative assessment of offshore wind energy:

[0025] Based on the actual wind field conditions in the coastal zone and the simulated wind field conditions in the coastal zone under the assumption of no cyclone, a comparative evaluation was conducted on the impact of cyclones on the coastal wind field. Combined with the future changing trend of the offshore pressure field, further analysis was conducted on offshore wind energy resources and the planning and construction of wind power projects.

[0026] Furthermore, in step S1, the reanalysis data include ERA5 and MERRA-2 reanalysis sea level pressure data, and the cyclone tracking algorithm uses the existing publicly available cyclone tracking algorithm; cyclone-related parameters are extracted, mainly including cyclone time period T, center position and trajectory P, cyclone radius R, original sea level pressure SLP_ori, cyclone center pressure, moving speed, cyclone intensity and cyclone coverage area.

[0027] Preferably, in the above-mentioned offshore wind energy comparative evaluation method based on ocean cyclone removal, the calculation method of SLP_Temp in step S303 is as follows:

[0028] SLP_Temp=0.2×SLP_ori+0.8×SLP_fix;

[0029] Among them, SLP_ori is the original sea level pressure; SLP_fix is ​​the replaced sea level pressure;

[0030] Preferably, in the above-mentioned offshore wind energy comparative evaluation method based on ocean cyclone removal, in step S304, SLP_Temp_s1 is calculated as follows: SLP_Temp is processed for the buffer marked by Buffer_2 and the buffer not marked by Buffer_1, and the processing method is to multiply SLP_ori and SLP_clim by a coefficient of 0.5 respectively, and add the two, and the calculation formula is as follows:

[0031] SLP_Temp_s1(SLP_Temp(Buffer_2==1&Buffer_1≠1))=

[0032] 0.5×SLP_ori(Buffer_2==1&Buffer_1≠1)+0.5×SLP_clim(Buffer_2==1&Buffer_1≠1);

[0033] Among them, SLP_ori is the original sea level pressure; SLP_clim is the climatic mean sea level pressure of the month corresponding to the cyclone time period T; Buffer_1 is the first buffer zone; Buffer_2 is the second buffer zone.

[0034] Preferably, in the above-mentioned offshore wind energy comparative evaluation method based on ocean cyclone removal, in step S5, the method of constructing a neural network is as follows:

[0035] S501. According to the ERA5 and MERRA-2 reanalysis wind field data, the actual wind field of the coastal zone corresponding to the cyclone time period T is obtained, including wind speed and wind direction;

[0036] S502, construct a multi-layer neural network model with 3 hidden layers. According to the experiment, the data distribution method is set to random, the training algorithm is set to Levenberg-Marquardt, the number of iterations is set to 5000, the performance index is expressed as mean square error, the performance target is set to 0.1, and the number of consecutive confirmation failures is set to 10;

[0037] S503, using the original sea level pressure SLP_ori as input and the actual wind field in the coastal zone as output, inputting the input and output into a multi-layer neural network model to construct a fitting model for the surface wind field in the coastal zone.

[0038] The beneficial effects of the present invention are:

[0039] (1) The present invention provides a set of actual ocean cyclone parameter sets and original sea level pressure fields, and develops a set of sea level pressure fields under the assumption of no cyclone based on the actual ocean cyclone parameters using the proposed ocean cyclone removal method;

[0040] (2) The present invention constructs a deep learning model based on the actual sea level pressure and the actual wind field in the coastal zone, and simulates and obtains the surface wind field in the coastal zone under the influence of a cyclone based on the developed sea level pressure field under the cyclone-free assumption;

[0041] (3) By comparing the surface wind field in the coastal zone with and without a cyclone, the present invention can evaluate the impact of the cyclone on offshore wind energy resources, thereby providing a reference for the planning and construction of offshore wind power projects. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 Construct a flow chart for fitting the surface wind field model in the near-coastal zone;

[0043] Figure 2 This is an example of cyclone tracking near the coast;

[0044] Figure 3 Example graph of model run for surface wind field fitting near the coast;

[0045] Figure 4 A flow chart for comparative assessment of offshore wind energy in coastal areas with and without cyclone influence;

[0046] Figure 5 Figure A is an example of the comparison of sea level pressure before and after the cyclone is removed;

[0047] Figure 6 Figure B is an example of the comparison of sea level pressure before and after the cyclone is removed;

[0048] Figure 7 This is an example of the comparison of offshore wind energy near the coast before and after the cyclone is removed;

[0049] Figure 8 It is a process logic frame of an embodiment of the present invention;

[0050] in, Figure 2 In the figure, black dots represent the actual monitoring paths of the cyclone, and colored dots represent the algorithm-tracked paths based on different reanalysis data; Figure 5 and Figure 6 In the figure, the left side shows the schematic diagram of sea level pressure before the cyclone is removed, and the right side shows the schematic diagram of sea level pressure after the cyclone is removed; Figure 7 In the figure, a and b represent the comparison example diagrams of offshore wind energy in the coastal zone before the cyclone is removed, and c and d represent the comparison example diagrams of offshore wind energy in the coastal zone after the cyclone is removed. DETAILED DESCRIPTION

[0051] like Figures 1 to 8 As shown, a comparative evaluation method for offshore wind energy based on ocean cyclone removal comprises the following steps:

[0052] S1, extracting ocean cyclone related parameters: using reanalysis data combined with cyclone tracking algorithm, extracting relevant parameters including cyclone time period T, center position and trajectory P, cyclone radius R, original sea level pressure SLP_ori, and classifying and storing each parameter and recording it in the geographic information grid;

[0053] S2, establish a buffer zone for the cyclone's impact range: set up two buffer zone coefficients based on the cyclone's center position and trajectory P, as well as the cyclone's radius R, and establish two buffer zones Buffer_1 and Buffer_2 corresponding to the cyclone; the specific method is as follows:

[0054] S201. Establish the first buffer zone Buffer_1 within the cyclone impact range:

[0055] With the position trajectory P of the cyclone center as the center and C_1*cyclone radius R as the side length, a square buffer zone Buffer_1 corresponding to the cyclone is generated. C_1 is the buffer zone coefficient 1. According to the experiment, the value C_1=1.75;

[0056] S202, establish the first buffer zone Buffer_2 within the cyclone impact range:

[0057] With the position trajectory P of the cyclone center as the center and C_2*cyclone radius R as the side length, a square buffer zone Buffer_2 corresponding to the cyclone is generated. C_2 is the buffer zone coefficient 2. According to the experiment, the value C_2=2;

[0058] S3, obtain the sea level pressure process quantity during the cyclone removal process:

[0059] S301, select the research time and space range, obtain the cyclone time period T and buffer zone Buffer_1;

[0060] S302, calculate the monthly climatic mean sea level pressure SLP_clim in the corresponding time and space range, and use SLP_clim to replace SLP_ori, recorded as SLP_fix;

[0061] S303, calculating the sea level pressure process quantity SLP_Temp according to SLP_ori and SLP_fix;

[0062] S304, reprocessing SLP_Temp according to SLP_ori, SLP_clim, Buffer_1, and Buffer_2 to calculate SLP_Temp_s1; processing SLP_Temp_s1 using Gaussian filtering, and recording the processed SLP_Temp_s2;

[0063] S4, get the sea level pressure SLP_revomeCyc after the cyclone is removed:

[0064] S401, calculate the difference between SLP_Temp_s2 and SLP_ori, recorded as SLP_dis;

[0065] S402, for the area where SLP_dis<0, reassign the value to SLP_ori, and obtain the sea level pressure SLP_revomeCyc after the cyclone is removed;

[0066] S5, construct the surface wind field model near the coast:

[0067] Taking the original sea level pressure SLP_ori as input and the actual wind field near the coast as output, a surface wind field model near the coast is constructed using deep learning.

[0068] S6, obtain the surface wind field near the coast without the influence of cyclone:

[0069] Based on the coastal surface wind field model constructed in S5, SLP_revomeCyc is used as input to obtain the coastal wind field conditions under the assumption of no cyclone, including wind speed and wind direction;

[0070] S7, Comparative assessment of offshore wind energy:

[0071] Based on the actual wind field conditions in the coastal zone and the simulated wind field conditions in the coastal zone under the assumption of no cyclone, a comparative evaluation was conducted on the impact of cyclones on the coastal wind field. Combined with the future changing trend of the offshore pressure field, further analysis was conducted on offshore wind energy resources and the planning and construction of wind power projects.

[0072] Furthermore, in step S1, the reanalysis data include ERA5 and MERRA-2 reanalysis sea level pressure data, and the cyclone tracking algorithm uses the existing publicly available cyclone tracking algorithm; cyclone-related parameters are extracted, mainly including cyclone time period T, center position and trajectory P, cyclone radius R, original sea level pressure SLP_ori, cyclone center pressure, moving speed, cyclone intensity and cyclone coverage area.

[0073] Preferably, in the above-mentioned offshore wind energy comparative evaluation method based on ocean cyclone removal, the calculation method of SLP_Temp in step S303 is as follows:

[0074] SLP_Temp=0.2×SLP_ori+0.8×SLP_fix;

[0075] Among them, SLP_ori is the original sea level pressure; SLP_fix is ​​the replaced sea level pressure;

[0076] Preferably, in the above-mentioned offshore wind energy comparative evaluation method based on ocean cyclone removal, in step S304, SLP_Temp_s1 is calculated as follows: SLP_Temp is processed for the buffer marked by Buffer_2 and the buffer not marked by Buffer_1, and the processing method is to multiply SLP_ori and SLP_clim by a coefficient of 0.5 respectively, and add the two, and the calculation formula is as follows:

[0077] SLP_Temp_s1(SLP_Temp(Buffer_2==1&Buffer_1≠1))=

[0078] 0.5×SLP_ori(Buffer_2==1&Buffer_1≠1)+0.5×SLP_clim(Buffer_2==1&Buffer_1≠1);

[0079] Among them, SLP_ori is the original sea level pressure; SLP_clim is the climatic mean sea level pressure of the month corresponding to the cyclone time period T; Buffer_1 is the first buffer zone; Buffer_2 is the second buffer zone.

[0080] Preferably, in the above-mentioned offshore wind energy comparative evaluation method based on ocean cyclone removal, in step S5, the method of constructing a neural network is as follows:

[0081] S501. According to the ERA5 and MERRA-2 reanalysis wind field data, the actual wind field of the coastal zone corresponding to the cyclone time period T is obtained, including wind speed and wind direction;

[0082] S502, construct a multi-layer neural network model with 3 hidden layers. According to the experiment, the data distribution method is set to random, the training algorithm is set to Levenberg-Marquardt, the number of iterations is set to 5000, the performance index is expressed as mean square error, the performance target is set to 0.1, and the number of consecutive confirmation failures is set to 10;

[0083] S503, using the original sea level pressure SLP_ori as input and the actual wind field in the coastal zone as output, inputting the input and output into a multi-layer neural network model to construct a fitting model for the surface wind field in the coastal zone.

[0084] The above-mentioned embodiments are only preferred technical solutions of the present invention and should not be regarded as limitations of the present invention. The embodiments and features in the embodiments of the present application can be arbitrarily combined with each other without conflict. The protection scope of the present invention shall be the technical solutions recorded in the claims, including the equivalent replacement solutions of the technical features in the technical solutions recorded in the claims. That is, equivalent replacement improvements within this scope are also within the protection scope of the present invention.

Claims

1. A comparative evaluation method for offshore wind energy based on ocean cyclone removal, characterized in that: The steps include: S1, extracting ocean cyclone related parameters: using reanalysis data combined with cyclone tracking algorithm, extracting relevant parameters including cyclone time period T, center position and trajectory P, cyclone radius R, original sea level pressure SLP_ori, and classifying and storing each parameter and recording it in the geographic information grid; S2, establish a buffer zone for the cyclone's impact range: set up two buffer zone coefficients based on the cyclone's center position and trajectory P, as well as the cyclone's radius R, and establish two buffer zones Buffer_1 and Buffer_2 corresponding to the cyclone; the specific method is as follows: S201. Establish the first buffer zone Buffer_1 within the cyclone impact range: With the position trajectory P of the cyclone center as the center and C_1*cyclone radius R as the side length, a square buffer zone Buffer_1 corresponding to the cyclone is generated. C_1 is the buffer zone coefficient 1, which is determined based on experiments. S202, establishing the second buffer zone Buffer_2 within the cyclone impact range: With the position trajectory P of the cyclone center as the center and C_2*cyclone radius R as the side length, a square buffer zone Buffer_2 corresponding to the cyclone is generated. C_2 is the buffer zone coefficient 2, which is determined based on experiments. S3, obtain the sea level pressure process quantity during the cyclone removal process: S301, select the research time and space range, obtain the cyclone time period T and buffer zone Buffer_1; S302, calculate the monthly climatic mean sea level pressure SLP_clim in the corresponding time and space range, and use SLP_clim to replace SLP_ori, recorded as SLP_fix; S303, calculating the sea level pressure process quantity SLP_Temp according to SLP_ori and SLP_fix; S304, reprocessing SLP_Temp according to SLP_ori, SLP_clim, Buffer_1, and Buffer_2 to calculate SLP_Temp_s1; processing SLP_Temp_s1 using Gaussian filtering, and recording the processed SLP_Temp_s2; S4, get the sea level pressure SLP_revomeCyc after the cyclone is removed: S401, calculate the difference between SLP_Temp_s2 and SLP_ori, recorded as SLP_dis; S402, for the area where SLP_dis<0, reassign the value to SLP_ori, and obtain the sea level pressure SLP_revomeCyc after the cyclone is removed; S5, construct the surface wind field model near the coast: Taking the original sea level pressure SLP_ori as input and the actual wind field in the coastal zone as output, a surface wind field model in the coastal zone is constructed using deep learning. The method for constructing the neural network model is as follows: S501. Based on the ERA5 and MERRA-2 reanalysis wind field data, obtain the actual wind field in the coastal zone corresponding to the cyclone time period T, including wind speed and wind direction; ERA-5 is the fifth generation global atmospheric reanalysis product released by the European Centre for Medium-Range Weather Forecasts; MERRA-2 is a global atmospheric reanalysis dataset released by NASA; S502, build a multi-layer neural network model with 3 hidden layers. According to the experiment, the data distribution method is set to random, the training algorithm is set to Levenberg-Marquardt, the number of iterations is set to 5000, the performance index is expressed as mean square error, and the performance target is set to 0.1, the number of consecutive confirmation failures is set to 10; S503, using the original sea level pressure SLP_ori as input and the actual wind field in the coastal zone as output, inputting the input and output into a multi-layer neural network model to construct a fitting model for the surface wind field in the coastal zone; S6, obtain the surface wind field near the coast without the influence of cyclone: Based on the coastal surface wind field model constructed in S5, SLP_revomeCyc is used as input to obtain the coastal wind field conditions under the assumption of no cyclone, including wind speed and wind direction; S7, Comparative assessment of offshore wind energy: Based on the actual wind field conditions in the coastal zone and the simulated wind field conditions in the coastal zone under the assumption of no cyclone, a comparative assessment is made on the impact of cyclones on the coastal wind field. Combined with the future changing trend of the offshore pressure field, further analysis is conducted on the planning and construction of offshore wind energy resources and wind power projects.

2. The offshore wind energy comparative evaluation method based on ocean cyclone removal according to claim 1, characterized in that: The SLP_Temp calculation method in step S303 is as follows: ； Among them, SLP_ori is the original sea level pressure; SLP_fix is ​​the replaced sea level pressure.

3. The offshore wind energy comparative evaluation method based on ocean cyclone removal according to claim 1 is characterized in that: In step S304, SLP_Temp_s1 is calculated as follows: SLP_Temp is processed for the buffer marked by Buffer_2 and the buffer not marked by Buffer_1 by multiplying SLP_ori and SLP_clim by a coefficient of 0.5 respectively, and the two are added together. The calculation formula is as follows: ； Among them, SLP_ori is the original sea level pressure; SLP_clim is the climatic mean sea level pressure of the month corresponding to the cyclone time period T; Buffer_1 is the first buffer zone; Buffer_2 is the second buffer zone.

Images