Ocean wind power plant planning system and method based on wind speed prediction

Abstract

The invention relates to the technical field of data prediction optimization, in particular to an ocean wind power plant planning system and method based on wind speed prediction, and the system comprises a wind energy flow monitoring module, a wind speed abrupt change prediction module, a fan wake flow influence analysis module, a fan group intelligent arrangement module and a wind power output optimization module. According to the method, the judgment precision of the wind energy distribution trend is improved by screening the abrupt change points; the wind speed change rate and the turbulence intensity change rate are combined with data of the wind speed monitoring buoy and the turbulence sensor, so that the wind speed sudden change prediction capability can be enhanced; the wind speed loss ratio, the angle difference and the wind energy diffusion model are used for wake flow area identification, fan layout is dynamically adjusted, and wake flow loss is reduced; the fan configuration can be optimized by combining the fan basic arrangement diagram, the wind energy capture rate, the arrangement distance and the steering angle; optimal parameters can be screened by combining power output interval calculation with blade attack angle adjustment data, torque load and fan rated power, fan power dynamic adjustment is achieved, and equipment operation stability is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data prediction optimization, and particularly to an ocean wind farm planning system and method based on wind speed prediction. Background Art

[0002] The technical field of data prediction optimization includes technical methods for analyzing, modeling, and calculating data to predict future trends and optimize resource allocation. The core content of this technical field is based on historical data and real-time information, using technical means such as mathematical modeling, statistical analysis, and machine learning to speculate on possible future changes and provide optimization solutions to improve decision-making efficiency. In the application of data prediction optimization technology, common methods include time series analysis, regression analysis, and neural network technologies. Through the mining and analysis of data, valuable prediction results are formed. This technical field is widely used in multiple industries such as energy management, financial markets, and logistics scheduling to support accurate decision-making and optimal resource allocation.

[0003] Among them, the ocean wind farm planning system refers to a system that designs and adjusts the site selection, layout, and operation plan of an offshore wind farm based on wind speed prediction data. It covers the collection and processing of wind resource data, the analysis of ocean environmental conditions, and the generation of an optimized layout plan for the wind farm. Specifically, this system conducts wind speed prediction through historical wind speed data and real-time meteorological data, evaluates suitable areas for wind turbine layout in combination with ocean hydrological data, and uses geographic information systems to optimize the arrangement of wind turbines to improve wind energy utilization efficiency and equipment stability. In addition, this system also adjusts the operation strategy of wind turbines according to the time-varying characteristics of wind energy resources, making the overall planning of the wind farm more in line with long-term power supply requirements.

[0004] The prior art has insufficient recognition of wind speed gradient anomalies and mutation points in wind energy resource assessment. The wind speed prediction method does not fully combine the change of turbulence intensity, making it difficult to accurately predict the local wind speed change in the wind farm. The wake effect analysis lacks wind energy diffusion modeling and relies on fixed spacing settings, resulting in high wake losses and reducing the overall power generation efficiency of wind turbines; the optimization of wind turbine layout is insufficient, and the spacing and steering angle cannot be dynamically adjusted, limiting the wind energy capture ability; the power output adjustment depends on static parameters and lacks real-time control, making the power response of wind turbines lag, affecting the long-term power supply stability of the wind farm. These problems will lead to a reduction in wind energy utilization efficiency and make it difficult for the wind farm to adapt to complex wind condition changes. Summary of the Invention

[0005] The purpose of the present invention is to solve the deficiencies existing in the prior art, and to propose an ocean wind farm planning system and method based on wind speed prediction.

[0006] To achieve the above purpose, the present invention adopts the following technical solution: An ocean wind farm planning system based on wind speed prediction includes: Wind energy flow monitoring module, which acquires the wind speed sequence, wind direction variability, and wind energy density gradient within an offshore wind farm, calculates the wind energy flux distribution value, determines abnormal wind speed gradients, screens mutation points, and generates a wind energy distribution trend value; Wind speed mutation prediction module, which, based on the wind energy distribution trend value, acquires data from wind speed monitoring buoys and offshore turbulence sensors, calls the wind speed change rate and turbulence intensity change rate, and determines the probability of wind speed mutation; Wind turbine wake influence analysis module, which, based on the wind speed mutation probability, acquires data from wind turbine wake monitoring radars and sea surface wind shear, calls the wind speed loss ratio and angle difference, calls a wind energy diffusion model, calculates the wake influence area, screens areas exceeding the threshold, and generates a wake influence coefficient; Intelligent layout module for wind turbine clusters, which, based on the wake influence coefficient, acquires the wind turbine foundation layout diagram and wind energy capture rate, calls the layout spacing and steering angle, calculates the layout optimization coefficient, and generates a wind turbine optimized layout table; Wind farm power output optimization module, which, based on the wind turbine optimized layout table, acquires the rated power of the wind turbine, blade pitch adjustment data, and torque load, calculates the power output interval, screens the optimal parameters within the interval, and generates wind turbine power adjustment parameters.

[0007] As a further solution of the present invention, the wind energy distribution trend value includes a wind speed sequence, wind direction variability, wind energy density gradient, wind energy flux distribution value, abnormal wind speed gradient, and mutation point; the wind speed mutation probability includes data from wind speed monitoring buoys, offshore turbulence sensor data, wind speed change rate, and turbulence intensity change rate; the wake influence coefficient includes data from wind turbine wake monitoring radars, sea surface wind shear data, wind speed loss ratio, angle difference, wind energy diffusion model, wake influence area, and area exceeding the threshold; the wind turbine optimized layout table includes a wind turbine foundation layout diagram, wind energy capture rate, layout spacing, steering angle, and layout optimization coefficient; the wind turbine power adjustment parameters include the rated power of the wind turbine, blade pitch adjustment data, torque load, power output interval, and optimal parameters.

[0008] As a further solution of the present invention, the wind energy flow monitoring module includes: Wind speed sequence extraction sub-module, which acquires wind speed data at multiple monitoring points within an offshore wind farm, arranges the data in a time series, calculates the wind speed mean, variance, and coefficient of variation, screens and records abnormal wind speed points, and simultaneously calls wind direction data to analyze the spatial characteristics of wind speed changes and generates a wind speed change sequence; Wind energy density calculation sub-module, which, based on the wind speed change sequence, calculates the probability distribution within the wind speed interval, combines with air density parameters, calculates the wind energy density in multiple intervals, and uses the formula: ; Performs operations to obtain the wind energy density distribution at multiple monitoring points and obtains the wind energy density gradient; Among them, represents the wind energy density, is the air density, is the representative wind speed within the wind speed range, is the probability of the wind speed range, is the change amount of adjacent wind speeds, represents at the wind speed and time the wind speed gain coefficient under the condition of, is the wind speed observation height, is the observation period, represents the number of wind speed ranges; The wind energy flux change analysis sub-module calls the wind energy density gradient, calculates the change of wind energy in the differentiated area of the offshore wind farm, analyzes the wind speed gradient mutation points, screens the wind energy density mutation areas, calculates the wind energy flux change rate, and generates the wind energy flux distribution value; The wind energy trend identification sub-module calls the wind energy flux distribution value, calculates the wind energy flux change trend of the differentiated time interval, identifies the long-term trend of wind energy change, calculates the mean square deviation of wind energy fluctuation through multi-period data, screens the mutation points and generates the trend curve, and obtains the wind energy distribution trend value.

[0009] As a further solution of the present invention, the wind speed mutation prediction module includes: The wind speed monitoring data acquisition sub-module, based on the wind energy distribution trend value, acquires the wind speed data of the wind speed monitoring buoy and the turbulence data of the offshore turbulence sensor, and at the same time calls the time series data stored in the sensor, extracts the wind speed time series, calculates the wind speed difference between adjacent time points and performs normalization processing, constructs the wind speed change trend sequence, and obtains the wind speed change rate sequence; The turbulence intensity calculation sub-module, based on the wind speed change rate sequence, calculates the turbulence intensity within the corresponding time window, using the formula: ; Performs operations to obtain the turbulence intensity value sequence, screens the change of turbulence intensity within the target wind speed change interval, and obtains the turbulence intensity change rate sequence; Among them, represents the turbulence intensity, represents the wind speed at the moment of, represents the average wind speed within the time window, represents the number of wind speed sampling points within the time window, represents the average air temperature within the time window, represents the turbulence intensity value of the th turbulence sensor, represents the number of turbulence sensors; The wind speed mutation assessment submodule calls the wind speed change rate sequence and the turbulence intensity change rate sequence, calculates the joint probability density distribution, screens the time periods with more drastic wind speed changes, normalizes the wind speed change rate distribution, calculates the mutation risk value in combination with the turbulence intensity change trend, classifies according to the threshold, and obtains the wind speed mutation probability value.

[0010] As a further solution of the present invention, the wind turbine wake impact analysis module includes: The wake monitoring data processing submodule obtains the wind turbine wake monitoring radar data and the sea surface wind shear data based on the wind speed mutation probability, extracts the wind speed mutation probability in the radar data, matches the wind shear data, identifies the wind shear characteristics of the wind speed mutation area, calculates the wind shear change rate, and extracts the wind speed mutation characteristic value based on the matching relationship between the wind speed mutation probability and the wind shear change rate to obtain the wind speed mutation characteristic value set; The wind speed loss calculation submodule, based on the wind speed mutation characteristic value set, calls the wind speed loss ratio and the angle difference, calculates the wind speed loss rate of the wind turbine wake, establishes the wind speed loss rate matrix, and normalizes it, using the formula: ; Calculate and obtain the wind speed loss ratio, and establish a normalized wind speed loss matrix; in, represents the wind speed loss ratio, represents the probability corresponding to the wind shear change rate, Represents the sudden change in wind speed. Represents the angle between the wind shear direction and the wind speed direction, Represents the number of data points in the wind speed mutation feature value set; The wake area assessment submodule calls the normalized wind speed loss matrix, combines the wind energy diffusion model, calculates the wake influence area, and simultaneously screens the wind speed loss ratio within the wake influence area, extracts the area where the wind speed loss exceeds the threshold, calculates the wake influence range, and generates the wake influence coefficient.

[0011] As a further solution of the present invention, the wind turbine group intelligent arrangement module includes: The wake influence coefficient calculation submodule obtains the basic layout diagram of the wind turbine based on the wake influence coefficient, calls the wind turbine spacing, wind speed and wind direction data, calculates the wake effect between the wind turbines, calculates the wake influence coefficient through the wind speed attenuation model, and corrects it in combination with the wind turbine height and terrain parameters to obtain the wake influence coefficient matrix; The wind energy capture rate analysis submodule calculates the effective wind speed of the wind turbine after being affected by the wake based on the wake influence coefficient matrix, calls the wind turbine power curve, calculates the wind energy conversion efficiency, and calculates the wind energy capture rate in combination with the wind turbine rated power and meteorological data to obtain the wind energy capture rate matrix; Based on the wind energy capture rate matrix, the fan optimal layout sub-module calls the layout spacing and steering angle parameters and uses the formula: ; Calculate the fan optimal layout parameters, adjust the fan arrangement, and obtain the fan optimal layout table; Among them, represents the fan optimal layout parameters, represents the rated power of the th fan, represents the wind energy capture rate of the th fan, represents the distance between the th fan and its adjacent fan, represents the th wake influence coefficient of the fan, represents the wake diffusion radius of the th fan, is the total number of fans,

[0012] As a further solution of the present invention, the wind farm power output optimization module includes: The fan power data acquisition sub-module acquires the rated power of the fans in the fan optimal layout table, calls the fan blade angle of attack adjustment data, extracts the torque load information, screens the fan data that meets the rated output conditions, calculates the influence of the change in the blade angle of attack on the power, adjusts the fan torque load influence factor, and establishes a fan power parameter set; The power output interval calculation sub-module is based on the fan power parameter set and uses the formula: ; Calculate through operation to obtain the fan power output interval, screen the optimal parameters within the interval, and obtain the optimal power output interval; Among them, represents the adjusted power output interval, represents the th weight coefficient of the fan, represents the th rated power of the fan, represents the th blade angle of attack of the fan, represents the reference blade angle of attack, represents the th torque load of the fan, represents the number of fans, represents the number of torque load data points; The fan power adjustment parameter generation sub-module calls the optimal power output range, adjusts the fan operation parameters, calculates the relationship coefficient between the blade angle of attack and power of the fan, combines the torque load to dynamically adjust the parameters, and obtains the fan power adjustment parameters.

[0013] A method for planning an offshore wind farm based on wind speed prediction, which is executed according to the offshore wind farm planning system based on wind speed prediction described in any one of the above, includes the following steps: S1: Obtain the wind speed sequence, wind direction variability, and wind energy density gradient, calculate the wind energy flux distribution value, judge the wind speed gradient abnormality according to the wind energy flux distribution value, screen the mutation points and calculate and generate the wind energy distribution trend value; S2: Based on the wind energy distribution trend value, obtain the wind speed change rate of the wind speed monitoring buoy and the turbulence intensity change rate of the offshore turbulence sensor, call the wind speed change rate and the turbulence intensity change rate for matching analysis, judge the wind speed mutation probability threshold, screen the area where the wind speed change rate exceeds the threshold as the risk area, and calculate and generate the wind speed mutation probability in combination with the turbulence intensity change rate; S3: Based on the wind speed mutation probability, obtain the wind speed loss ratio of the fan wake monitoring radar and the angle difference of the sea surface wind shear, call the wind speed loss ratio and the angle difference for comparison analysis, calculate the wake influence area according to the wind energy diffusion model, screen the wake influence area, and generate the wake influence coefficient; S4: Based on the wake influence coefficient, obtain the layout spacing of the fan foundation layout diagram and the steering angle of the wind energy capture rate, call the layout spacing and the steering angle to calculate the layout optimization coefficient, adjust the fan layout according to the layout optimization coefficient, and generate the fan optimization layout table; S5: Based on the fan optimization layout table, obtain the blade angle of attack adjustment data and torque load of the fan rated power, call the blade angle of attack adjustment data and torque load to calculate the power output range, screen the parameters within the power output range, and generate the fan power adjustment parameters.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In the present invention, by analyzing the wind speed sequence, wind direction variability, and wind energy density gradient, the wind energy flux distribution value is calculated, and mutation points are screened, making the judgment of the wind energy distribution trend more accurate and improving the reliability of wind resource assessment. The combination of the wind speed change rate and the turbulence intensity change rate with the data of the wind speed monitoring buoy and the offshore turbulence sensor can enhance the wind speed mutation prediction ability and optimize the operation control of the wind farm. The wind speed loss ratio, angle difference, and wind energy diffusion model are applied to the wake area identification, enabling the dynamic adjustment of the wind turbine layout and reducing the wake loss. The wind turbine foundation layout diagram and the wind energy capture rate, combined with the layout spacing and steering angle, can improve the wind energy capture ability and optimize the wind turbine spacing configuration. The calculation of the power output interval, combined with the data of the blade pitch angle adjustment, torque load, and the rated power of the wind turbine, can accurately screen the optimal parameters, realize the dynamic adjustment of the wind turbine power, improve the operation stability of the equipment, and optimize the long-term power generation strategy of the wind farm. Description of the Drawings

[0015] Figure 1 is the system flowchart of the present invention; Figure 2 is the flowchart of the wind energy flow monitoring module in the present invention; Figure 3 is the flowchart of the wind speed mutation prediction module in the present invention; Figure 4 is the flowchart of the wind turbine wake influence analysis module in the present invention; Figure 5 is the flowchart of the intelligent layout module of the wind turbine group in the present invention; Figure 6 is the flowchart of the wind farm power output optimization module in the present invention. Detailed Embodiments

[0016] In order to make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0017] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as limiting the present invention. In addition, in the description of the present invention, the meaning of "a plurality of" is two or more unless otherwise specifically defined.

[0018] Embodiment 1 Please refer toFigure 1 , the present invention provides an ocean wind farm planning system based on wind speed prediction, including: A wind energy flow monitoring module, which acquires the wind speed sequence, wind direction variability, and wind energy density gradient within the ocean wind farm, calculates the wind energy flux distribution value, determines the abnormal wind speed gradient, screens the mutation points, and generates the wind energy distribution trend value; A wind speed mutation prediction module, which based on the wind energy distribution trend value, acquires the data of wind speed monitoring buoys and offshore turbulence sensors, calls the wind speed change rate and the turbulence intensity change rate, and determines the wind speed mutation probability; A wind turbine wake influence analysis module, which based on the wind speed mutation probability, acquires the data of wind turbine wake monitoring radars and sea surface wind shear, calls the wind speed loss ratio and the angle difference, calls the wind energy diffusion model, calculates the wake influence area, screens the areas exceeding the threshold, and generates the wake influence coefficient; A wind turbine group intelligent layout module, which based on the wake influence coefficient, acquires the wind turbine foundation layout diagram and the wind energy capture rate, calls the layout spacing and the steering angle, calculates the layout optimization coefficient, and generates the wind turbine optimized layout table; A wind farm power output optimization module, which based on the wind turbine optimized layout table, acquires the rated power of the wind turbine, the blade pitch angle adjustment data, and the torque load, calculates the power output interval, screens the optimal parameters within the interval, and generates the wind turbine power adjustment parameters.

[0019] The wind energy distribution trend value includes the wind speed sequence, wind direction variability, wind energy density gradient, wind energy flux distribution value, abnormal wind speed gradient, and mutation points. The wind speed mutation probability includes the data of wind speed monitoring buoys, offshore turbulence sensors, wind speed change rate, and turbulence intensity change rate. The wake influence coefficient includes the data of wind turbine wake monitoring radars, sea surface wind shear, wind speed loss ratio, angle difference, wind energy diffusion model, wake influence area, and areas exceeding the threshold. The wind turbine optimized layout table includes the wind turbine foundation layout diagram, wind energy capture rate, layout spacing, steering angle, and layout optimization coefficient. The wind turbine power adjustment parameters include the rated power of the wind turbine, blade pitch angle adjustment data, torque load, power output interval, and optimal parameters.

[0020] Please refer to Figure 2 , the wind energy flow monitoring module includes: A wind speed sequence extraction sub-module, which acquires the wind speed data of multiple monitoring points within the ocean wind farm, arranges the data according to the time sequence, calculates the wind speed mean, variance, and coefficient of variation, screens and records the abnormal wind speed points, and at the same time calls the wind direction data to analyze the spatial characteristics of the wind speed change and generates the wind speed change sequence; Wind speed data at multiple monitoring points in an offshore wind farm are collected by multiple wind speed sensors. These wind speed sensors are installed at different heights (such as 50m, 100m, 150m) and record wind speed data at fixed time intervals (such as 1min, 5min). The initially obtained data set contains timestamps, wind speed values, and monitoring point numbers. When sorting the wind speed data in time series, first arrange all wind speed values in ascending order according to the timestamps to ensure the integrity of the data sequence and compliance with the chronological relationship, and then calculate the average wind speed , and its calculation method is: ; Among them, is the number of monitoring points, is the wind speed value of each monitoring point. For example, a wind farm has 5 monitoring points, and the wind speeds at a certain time point are 5.2m / s, 6.3m / s, 5.8m / s, 6.0m / s, 5.5m / s respectively, then: ; Next, calculate the wind speed variance : ; Similarly, substitute the above data: ; Subsequently, calculate the coefficient of variation : ; Substitute the calculated variance value: ; This value is used to judge the degree of dispersion of wind speed changes. If it is determined that the wind speed fluctuates greatly, and abnormal values need to be further screened. In order to detect abnormal wind speed points, the principle is adopted, that is, set the threshold range: ; Substitute the data: ; If a measured wind speed value exceeds this range, it is marked as an abnormal value. For example, if the wind speed at monitoring point 1 at a certain moment is 8.2m / s, it is determined as an abnormal wind speed point and recorded. On this basis, call the wind direction data and calculate the wind speed change trend of different monitoring points, that is, for the wind speed data of different monitoring points, calculate the first-order difference in the time series: ; For example, the wind speeds of a certain monitoring point at time are 5.2, 5.5, 6.1; ; The result shows that the variation of wind speed at different time points can be reflected by the wind speed difference sequence. Subsequent calculation of wind energy density needs to be based on this data for zonal statistics. The construction of the wind speed change sequence provides basic data for wind energy calculation.

[0021] The wind energy density calculation sub-module calculates the probability distribution within the wind speed interval based on the wind speed change sequence, and combines with the air density parameter to calculate the wind energy density in multiple intervals, using the formula: ; Perform operations to obtain the wind energy density distribution at multiple monitoring points and get the wind energy density gradient; Among them, represents the wind energy density, is the air density, is the representative wind speed within the wind speed interval, is the probability of the wind speed interval, is the change amount of adjacent wind speeds, represents the wind speed gain coefficient under the wind speed and time , is the wind speed observation height, is the observation period, represents the number of wind speed intervals; First, determine the wind speed interval. For example, set the wind speed interval as etc., and count the occurrence probability of the wind speed within each wind speed interval . Assume that a total of 2000 wind speed records are made during the observation period, and among them, the number of data falling within the interval is 350 times, then: ; Combine with the air density to calculate the wind energy density : ; Assume that the representative wind speed of a certain wind speed interval, then: ; If the wind speed gain coefficient , , the observation height , the observation period : ; Therefore, the contribution of the wind energy density in this interval is: ; The result shows that in the wind speed range of 5.5 m / s, the corresponding wind energy density is , and this result is closely related to factors such as the wind speed range probability and air density. This data can be used to calculate the wind energy density gradient of the overall wind farm.

[0022] The wind energy flux change analysis sub-module calls the wind energy density gradient to calculate the change of wind energy in the differentiated areas of the offshore wind farm, analyzes the sudden change points of the wind speed gradient, screens the areas with sudden changes in wind energy density, calculates the wind energy flux change rate, and generates the wind energy flux distribution value; First, calculate the wind energy flux : ; Assume that the wind energy density in a certain area is , the wind speed is 6 m / s, and the swept area of the fan blade is , then: ; Calculate the wind energy flux change rate in different areas: ; If the flux in area 1 is 45000 W and the flux in area 2 is 54000 W: ; The result shows that the change rate of wind energy flux between the two areas reaches 20%, which belongs to the area with sudden change in wind energy density. The subsequent wind energy trend analysis will calculate the long-term trend based on this flux change.

[0023] The wind energy trend identification sub-module calls the wind energy flux distribution value to calculate the wind energy flux change trend at different time intervals, identifies the long-term trend of wind energy change, calculates the mean square deviation of wind energy fluctuation through multi-period data, screens the sudden change points and generates the trend curve, and obtains the wind energy distribution trend value.

[0024] First, set the analysis time period. For example, set daily, weekly, and monthly as different time intervals, and count the change values of wind energy flux within these time intervals , and calculate the mean square deviation of wind energy fluctuation at different time intervals: ; Among them, is the wind energy flux at the th time point, is the mean value of wind energy flux within this time interval. Assume that the wind energy fluxes of a wind farm within 5 days are 52000 W, 53500 W, 51000 W, 49500 W, and 52500 W respectively, then calculate the mean value: ; Calculate the mean square deviation of wind energy fluctuation: ; ; ; ; The calculated mean square error indicates that there is a certain volatility in the wind energy flux during this period. If it exceeds a certain threshold (e.g., 1500 W), it means that the wind energy changes violently during this period and may be affected by environmental factors such as the passage of a cold front or monsoon changes. If the mean square error of the fluctuation is less than 500 W, it can be determined that the wind energy changes relatively stably.

[0025] Next, identify the mutation points of the wind energy flux and set the mutation threshold. For example, if the change in the wind energy flux at a certain time point exceeds the range of the mean value, it is marked as a mutation point: ; Substitute the calculated value: ; If the wind energy flux at a certain moment exceeds this range, such as reaching 56000 W, then this time point is identified as a mutation point and recorded.

[0026] Finally, generate the wind energy trend curve and calculate the wind energy distribution trend value: ; Among them, is the wind energy flux at the latest time point, is the wind energy flux at the starting time point, is the time interval. For example, if the wind energy flux rises from 50000 W to 55000 W within a month: ; This result shows that within a month, the wind energy flux increases at a rate of 6.94 W per hour, indicating that the wind resource shows an overall upward trend during this period. Combining with the long-term trend, it can be used to optimize the wind energy utilization strategy of the wind farm.

[0027] Table 1 Wind Energy Flux Change Trend Calculation Table

[0028] As shown in Table 1, record the changes in the wind energy flux at different time points, which can be used later to calculate the wind energy trend and identify mutation points.

[0029] Please refer to Figure 3 , the wind speed mutation prediction module includes: The wind speed monitoring data acquisition sub-module obtains the wind speed data of the wind speed monitoring buoy and the turbulence data of the offshore turbulence sensor based on the wind energy distribution trend value. At the same time, it calls the time series data stored in the sensor, extracts the wind speed time series, calculates the wind speed difference between adjacent time points and performs normalization processing, constructs the wind speed change trend sequence, and obtains the wind speed change rate sequence; First, make an initial determination according to the wind energy distribution trend value, select the buoy equipment that meets the wind speed monitoring requirements, and read the real-time wind speed data from the monitoring buoy. The data reading frequency is set to once every 10 seconds, and the data recorded each time includes the wind speed value and the corresponding timestamp. For example, the wind speed recorded by a certain wind speed monitoring buoy at 10:00:00 is 5.2 m / s, and the wind speed recorded at 10:00:10 is 5.5 m / s. At the same time, obtain the turbulence data in the same time period through the offshore turbulence sensor, and the data format is the same as the wind speed data. The synchronization time accuracy of the wind speed monitoring buoy and the offshore turbulence sensor is controlled within 1 second to ensure the consistency of the data time series. Then, call the stored time series data, extract the wind speed values in the past 5 minutes, and construct the wind speed time series. The number of wind speed data points in this time series is jointly determined by the sampling interval and the time window. For example, under a 5-minute time window and a sampling interval of 10 seconds, a total of 5x60 / 10 = 30 data points are obtained. After sorting this sequence by timestamp, calculate the wind speed difference between adjacent time points. For example, if the wind speed value at 10:00:00 is 5.2 m / s and the wind speed value at 10:00:10 is 5.5 m / s, the wind speed difference at this moment is calculated as 5.5 - 5.2 = 0.3 m / s. After calculating the wind speed differences of all adjacent data points, perform normalization processing. The normalization method uses min-max normalization, that is, for the wind speed difference sequence , calculate its maximum value and minimum value . The normalized wind speed change sequence The calculation formula is: ; Among them, if the maximum wind speed change value = 1.2 m / s and the minimum value = -0.8 m / s, for = 0.3 m / s, the normalized value is: ; After normalization, the wind speed change trend sequence is obtained. This sequence can characterize the change pattern of the wind speed over time. Subsequently, further calculate the wind speed change rate sequence. For each wind speed change trend data point, calculate the change rate of its adjacent time point. The wind speed change rate is defined as: ; Among them Let ; Finally, a wind speed change rate sequence is obtained. This result indicates that the variation characteristics of the current wind speed over time have been quantified and can be used to further evaluate the turbulence intensity and wind speed mutation situation.

[0030] The turbulence intensity calculation sub-module calculates the turbulence intensity within the corresponding time window based on the wind speed change rate sequence, using the formula: ; Operate to obtain a sequence of turbulence intensity values, screen the change of turbulence intensity within the target wind speed change interval, and obtain a sequence of turbulence intensity change rates; Among them, represents the turbulence intensity, represents the wind speed at the th moment, represents the average wind speed within the time window, represents the number of wind speed sampling points within the time window, represents the average air temperature within the time window, represents the turbulence intensity value of the th turbulence sensor, represents the average turbulence intensity value of all turbulence sensors, represents the number of turbulence sensors; First, extract the wind speed change rate sequence and set a fixed time window. For example, set the time window to 60 seconds. Within this time window, calculate the mean wind speed , such as the wind speed data within a certain time window is {5.2, 5.4, 5.8, 6.0, 5.7, 5.5} m / s, then the mean value is calculated as: ; Next, calculate the turbulence intensity within this time window, and the formula is as follows: ; Among them: set (the number of sampling points within the time window); - calculate the variance term: ; Take the average air temperature = 15 °C; Assume the number of offshore turbulence sensors = 3, and the turbulence intensity values are {0.8, 0.6, 0.7}, calculate their mean value: ; Calculate the turbulence intensity: ; ; ; The calculated turbulence intensity value is 0.271. Then, the target wind speed change interval is screened. For example, a mutation threshold of 0.3 m / s is set 2 , then the time window data that meet the conditions are screened, and the turbulence intensity change rate is calculated, defined as: ; For example, the turbulence intensity at 10:00:00 is 0.271, and the turbulence intensity calculated at 10:00:10 is 0.310. Then: ; Finally, a sequence of turbulence intensity change rates is obtained. This result shows that the fluctuation trend of the current turbulence intensity has been calculated and can be used in the wind speed mutation assessment stage to determine whether there is a significant change in the wind speed.

[0031] The wind speed mutation assessment sub-module calls the wind speed change rate sequence and the turbulence intensity change rate sequence, calculates the joint probability density distribution, screens the time periods with relatively large wind speed changes, normalizes the wind speed change rate distribution at the same time, calculates the mutation risk value in combination with the turbulence intensity change trend, and classifies according to the threshold to obtain the wind speed mutation probability value.

[0032] First, read the wind speed change rate sequence calculated previously and the turbulence intensity change rate sequence , calculate the joint probability density distribution of the wind speed change rate and the turbulence intensity change rate within the specified time window. The kernel density estimation method is used for the calculation. The wind speed change rate and the turbulence intensity change rate within each time window are used as sample points to calculate their density values in the joint distribution. Suppose 5 groups of data are sampled within a certain time window: Table 2 Example data of wind speed change rate and turbulence intensity change rate

[0033] As shown in Table 2, the joint probability density calculation is performed between data points using the bivariate kernel density estimation method. The corresponding joint density estimation formula is: ; Set = 0.1, = 0.002, substitute the data to calculate the joint density value. After obtaining the density distribution, set the threshold , such as = 0.02, screen the time windows with density values greater than the threshold. For example, at 10:00:30 and 10:00:40, the corresponding wind speed change rates of the screened data points are relatively high, and the change rate of turbulence intensity increases significantly.

[0034] Next, normalize the distribution of the wind speed change rate, and set the normalization formula: ; Set the maximum value of the wind speed change rate = 0.45 m / s2, the minimum value = 0.05 m / s2. For the data point at 10:00:30, r = 0.30 m / s2: ; Finally, calculate the mutation risk value in combination with the change trend of turbulence intensity. The risk value is calculated using the weighted method: ; Set the weights = 0.6, = 0.4, and calculate for the moment of 10:00:30: ; ; Set the mutation risk classification threshold: is a high risk; is a medium risk; is a low risk.

[0035] The mutation risk value at 10:00:30 is 0.6181, which falls into the medium risk interval. Finally, obtain the wind speed mutation probability value, classify and count the risk values at all moments, and obtain the time period of wind speed mutation.

[0036] Please refer to Figure 4 , the fan wake influence analysis module includes: The wake monitoring data processing sub-module, based on the wind speed mutation probability, obtains the fan wake monitoring radar data and the sea surface wind shear data, extracts the wind speed mutation probability from the radar data, matches the wind shear data, identifies the wind shear characteristics of the wind speed mutation area, calculates the wind shear change rate, and at the same time, based on the matching relationship between the wind speed mutation probability and the wind shear change rate, extracts the wind speed mutation characteristic value to obtain the wind speed mutation characteristic value set; First, call the fan wake monitoring radar data. The radar data contains the wind speed measurement values at different times. By comparing the wind speed changes at consecutive time points, it is judged whether the wind speed change in a certain time period reaches the set mutation threshold. The threshold can be set to the situation where the wind speed change rate exceeds 2.5 m / s2, and the data set that meets the wind speed mutation probability calculation standard is screened out.

[0037] Meanwhile, extract the wind shear values corresponding to the time period from the sea surface wind shear data. The calculation method of the wind shear value can adopt the horizontal wind shear calculation formula .

[0038] Among them, is the height difference, is the wind speed change amount at . Assume that the wind speeds measured at different heights at a certain moment are , and the height interval is , then . Then calculate the wind shear change rate, compare the wind shear values at different time points, and obtain its time change rate . For example, the wind shear values at two consecutive moments are and , and the time interval is , then .

[0039] Finally, match the wind speed mutation probability with the wind shear change rate, filter out the data of the time period with the wind speed mutation probability greater than 50%, and extract the wind speed mutation characteristic values. For example, within a certain time period, the wind speed mutation probability is 60%, and the wind speed change amount is 3.2 m / s, and the corresponding wind shear change rate is . Then record this data point and store it in the wind speed mutation characteristic value set. Finally, obtain the complete wind speed mutation characteristic value data set. This result shows that there is an obvious correlation between the time period with a high wind speed mutation probability and the wind shear change rate, which can be used to further calculate the wind speed loss and the wake influence area.

[0040] Based on the wind speed mutation characteristic value set, the wind speed loss calculation sub-module calls the wind speed loss ratio and the angle difference to calculate the wind turbine wake wind speed loss rate, establish a wind speed loss rate matrix, and perform normalization. The formula is: ; Calculate the wind speed loss ratio through the operation and establish a normalized wind speed loss matrix; Among them, represents the wind speed loss ratio, represents the probability corresponding to the wind shear change rate, represents the wind speed mutation value, represents the angle between the wind shear direction and the wind speed direction, represents the number of data points in the wind speed mutation characteristic value set; First, extract the wind speed change amount and the corresponding wind shear direction from the wind speed mutation characteristic value set, and set the reference parameters for calculating the wind speed loss ratio. Among them The calculation method is the angle between the wind shear direction and the wind speed direction, which is obtained by the method of decomposing the wind speed vector. Assuming the wind speed vector components = 8 m / s, = 6 m / s, then , and then calculate the wind speed loss rate value, using the formula: ; In the calculation, it is assumed that the wind speed mutation eigenvalue set contains 5 data points, as shown in the following table: Table 3 Wind speed loss calculation parameter table

[0041] According to the data in Table 3, calculate the wind speed loss ratio: ; The calculation result is: ; Then perform normalization processing. Assuming the maximum wind speed loss ratio is 3.5, the normalized wind speed loss ratio is: ; This result indicates that the wind speed loss ratio in the current wind turbine wake area is at a medium level. The ratio to the maximum loss ratio indicates that the wake effect of this wind turbine has not reached the extreme impact level. This data can be used to further evaluate the wake impact area and formulate wake optimization strategies.

[0042] The wake area evaluation sub-module calls the normalized wind speed loss matrix, combines with the wind energy diffusion model, calculates the wake impact area, simultaneously screens the wind speed loss ratio within the wake impact area, extracts the area where the wind speed loss exceeds the threshold, and calculates the wake impact range to generate the wake impact coefficient.

[0043] First, map the data points in the wind speed loss matrix to space according to the coordinate positions, calculate the wake impact area, set the parameters of the wind energy diffusion model. Assuming the influence radius of the wind speed loss is 300 meters, then screen out all areas where the wind speed loss ratio exceeds 0.5 within the radius range, count the data points that meet the standard, calculate the wake impact coefficient, and set the wake impact coefficient calculation formula: ; Assuming there are 4 data points that meet the conditions, and the normalized wind speed loss ratios are 0.609, 0.572, 0.645, and 0.688 respectively, then calculate the wake impact coefficient: ; The result shows that the average wind speed loss ratio in the affected area of the wind turbine wake is 0.6285, indicating that the wind speed in this area decreases significantly, but does not reach the extreme loss level. This value can be used to further evaluate the optimization requirements of the wind turbine cluster layout on the wake effect and provide a basis for subsequent wind speed loss correction.

[0044] Please refer to Figure 5 , the intelligent layout module of the wind turbine cluster includes: The wake effect coefficient calculation sub-module, based on the wake effect coefficient, obtains the basic layout diagram of the wind turbines, calls the data of the wind turbine spacing, wind speed and wind direction, calculates the wake effect between the wind turbines, calculates the wake effect coefficient through the wind speed attenuation model, and corrects it in combination with the wind turbine height and terrain parameters to obtain the wake effect coefficient matrix; First, it is necessary to clarify the initial layout method of the wind turbines. Assume that the wind turbines are arranged in a square grid, the spacing between each wind turbine is set to 600 meters, the wind speed value range is from 4 m / s to 12 m / s, and the wind direction angle range is from 0° to 360°. For the wake effect between the wind turbines, it is necessary to call the attenuation data of the wind speed downstream of the wind turbines and calculate using the wind speed attenuation model. Set the front-back row spacing of the wind turbines d1 = 600 m, the lateral spacing d2 = 500 m. Assume that the upstream wind speed of wind turbine A is V1 = 10 m / s, and the wake speed attenuation ΔV generated by wind turbine A is 2 m / s. Then the affected wind speed V2 of wind turbine B is V2 = V1 - ΔV = 8 m / s. Based on this, the wake effect coefficient is calculated Use the wake diffusion model to calculate the wake influence radius D of the wind turbine. If the wake diffusion parameter is set to 0.1, the calculation method of the wake diffusion radius D is as follows: ; Substitute d1 = 600 m and d2 = 500 m to get: ; Then calculate the wake effect coefficient , assuming that the energy loss ratio affected by the wind turbine wake is 10%, then: ; After the calculation is completed, construct the wind turbine wake effect coefficient matrix as follows: Table 4 Wake effect coefficient matrix

[0045] As shown in Table 4, the calculation of the wind turbine wake effect coefficient is completed. Combine the wind turbine height and terrain parameters for correction. If the terrain influence factor is set to 0.05, the wake effect coefficient of wind turbine A after correction is adjusted as follows: ; Calculate the wake effect coefficients of all wind turbines in turn, and finally obtain the complete wake effect coefficient matrix.

[0046] The wind energy capture rate analysis sub-module, based on the wake influence coefficient matrix, calculates the effective wind speed of the wind turbine affected by the wake, calls the wind turbine power curve, calculates the wind energy conversion efficiency, combines the rated power of the wind turbine and meteorological data, calculates the wind energy capture rate, and obtains the wind energy capture rate matrix; Set the rated power of the wind turbine P = 2000kW. According to the wind turbine power curve, if the output power of the wind turbine is P = 1600kW when the wind speed is 8m / s and the output power is P = 1200kW when the wind speed is 6.5m / s, the wind energy conversion efficiency is calculated according to the formula: ; When the wind speed is 8m / s: ; When the wind speed is 6.5m / s: ; Furthermore, calculate the wind energy capture rate matrix. Assuming that the average wind energy density given by the meteorological data is 400W / m2, the wind energy capture rate of wind turbine A is calculated as follows: ; Similarly, the wind energy capture rate matrices of all wind turbines can be calculated as follows: Table 5 Wind energy capture rate matrix

[0047] As shown in Table 5, the wind energy capture rates of all wind turbines have been calculated.

[0048] The wind turbine optimal layout sub-module, based on the wind energy capture rate matrix, calls the layout spacing and steering angle parameters, and uses the formula: ; Calculate the wind turbine optimal layout parameters, adjust the wind turbine layout, and obtain the wind turbine optimal layout table; Among them, represents the wind turbine optimal layout parameter, represents the rated power of the th wind turbine, represents the wind energy capture rate of the th wind turbine, represents the spacing between the th wind turbine and its adjacent wind turbine, represents the th wind turbine wake influence coefficient, represents the wake diffusion radius of the th wind turbine, is the total number of wind turbines, is the number of pairs of wind turbines affected by the wake.

[0049] Set the minimum distance between wind turbines to 600 m and the maximum distance to 1000 m. During the optimization process, use formulas to calculate the optimized layout parameters of the wind turbines. First, calculate the influence factors of the optimized layout of each wind turbine: ; Substitute the data for calculation. Assume the parameters of wind turbines A, B, and C are as follows: ; ; ; ; ; Calculate the optimized parameters of the wind turbines: ; ; ; ; The result shows that the optimized layout parameter of the current wind turbine arrangement is 3.2033. Combining with the optimized layout table, adjust the wind turbine arrangement to the configuration with the maximum wind energy capture rate to obtain the final optimized layout table of the wind turbines.

[0050] Please refer to Figure 6 , the wind farm power output optimization module includes: The wind turbine power data acquisition sub-module acquires the rated power of the wind turbines in the optimized layout table of the wind turbines, calls the wind turbine blade pitch angle adjustment data, extracts the torque load information, screens the wind turbine data that meets the rated output conditions, calculates the influence of the blade pitch angle change on the power, adjusts the wind turbine torque load influence factor, and establishes the wind turbine power parameter set; First, according to the optimized layout table of the wind turbines, acquire the rated power data of each wind turbine one by one. The specific process is to call the wind turbine database and query the factory-rated power parameters of each wind turbine. For example, if the rated power of a wind turbine is set to 3.0 MW, then the reference power data of this wind turbine is 3000 kW. Then, call the wind turbine blade pitch angle adjustment data to analyze the deviation between the current blade pitch angle value of the wind turbine and its rated pitch angle value. During the specific implementation process, the current blade pitch angle of the wind turbine is obtained in real time through the wind turbine monitoring system , and compared with the reference pitch angle set at the factory. For example, if , , then the deviation calculation result is . Then, extract the torque load information of the wind turbine. This information comes from the motor torque sensor monitored by the wind turbine main control system. If the torque load If it is = 3500 N·m, then store this data in the dataset to be screened. Then, screen the fan data that meets the rated output conditions. This operation uses an interval judgment method, and a rated output power deviation threshold is set. For example is set to , then screen the fan data that meets . For a certain fan, if the currently measured output power is 2900 kW and its deviation meets the screening conditions, so retain the fan data. Subsequently, calculate the impact of the blade angle of attack change on power. This calculation is approximately deduced based on the fan power curve model. For example, if the power decreases by 50 kW for every 1° increase in the angle of attack, then for a fan with a deviation of 2°, the power correction amount is calculated as -100 kW. Finally, adjust the fan torque load impact factor. This impact factor is set to the impact coefficient of torque load on power. For example, for a certain fan = 0.01, then the corrected power impact is = 35 kW. Integrate the above calculation results to form a fan power parameter set, including information such as the rated power, blade angle of attack, angle-of-attack corrected power, and torque-load corrected power of each fan. Finally, construct a dataset for subsequent calculations.

[0051] Based on the fan power parameter set, the power output interval calculation sub-module uses the formula: ; Perform operations to obtain the fan power output interval, screen the optimal parameters within the interval, and obtain the optimal power output interval; Among them, represents the adjusted power output interval, represents the weight coefficient of the th fan, represents the rated power of the th fan, represents the blade angle of attack of the th fan, represents the reference blade angle of attack, represents the torque load of the th fan, represents the number of fans, represents the number of torque load data points; First, call the power parameter data of all fans, including the rated power , the angle-of-attack corrected power and the torque-load corrected power , then calculate the power adjustment interval using the formula: ; Among them, the assignments of each parameter are as follows: the number of fans = 5, torque load data points = 5, rated power of the fan = [3000, 2900, 3100, 2950, 3050] kW, blade angle of attack = [6, 5, 7, 5.5, 6.2] °, reference angle of attack = 5 °, torque load = [3500, 3400, 3600, 3300, 3450] N·m, fan weight = [1, 0.9, 1.1, 1.0, 0.95].

[0052] First, calculate the first summation: ; ; Then, calculate the angle of attack deviation term: ; Then, calculate the correction term: ; Angle of attack correction amount: ; Final calculation: ; ; This result indicates that the current operating parameters of the fan result in a low power adjustment value, and it is necessary to further optimize the blade angle of attack or torque load parameters of the fan to obtain a better power output range.

[0053] The fan power adjustment parameter generation sub-module calls the optimal power output range, adjusts the fan operating parameters, calculates the relationship coefficient between the blade angle of attack and power of the fan, and dynamically adjusts the parameters in combination with the torque load to obtain the fan power adjustment parameters.

[0054] First, according to the calculated value, adjust the fan operating parameters, correct the blade angle of attack, and make it approach the reference angle of attack , for example, if the current blade angle of attack of a fan is 6 ° and the calculated power adjustment value is low, then the blade angle of attack needs to be adjusted to 5.5 ° to reduce power loss. At the same time, calculate the relationship coefficient between the blade angle of attack and power of the fan. Assuming that each degree change in the angle of attack results in a 50 kW power correction, then = 50. Subsequently, dynamically adjust the parameters in combination with the torque load. Assuming its value is 0.01, for a current torque load change = -100 N·m of a fan, then the power correction value = 0.01 x (-100) = -1 kW. Finally, the wind turbine power adjustment parameters are obtained, including the new blade angle of attack setting value, the corrected power output value, and the corresponding load adjustment parameters, to guide the optimized operation of the wind turbine.

[0055] A method for planning an offshore wind farm based on wind speed prediction, which is executed by an offshore wind farm planning system based on wind speed prediction, includes the following steps: S1: Obtain the wind speed sequence, wind direction variability, and wind energy density gradient, calculate the wind energy flux distribution value, judge the wind speed gradient abnormality according to the wind energy flux distribution value, screen the mutation points, and calculate and generate the wind energy distribution trend value; S2: Based on the wind energy distribution trend value, obtain the wind speed change rate of the wind speed monitoring buoy and the turbulence intensity change rate of the offshore turbulence sensor, call the wind speed change rate and the turbulence intensity change rate for matching analysis, judge the wind speed mutation probability threshold, screen the area where the wind speed change rate exceeds the threshold as the risk area, and calculate and generate the wind speed mutation probability in combination with the turbulence intensity change rate; S3: Based on the wind speed mutation probability, obtain the wind speed loss ratio of the wind turbine wake monitoring radar and the angular difference of the sea surface wind shear, call the wind speed loss ratio and the angular difference for comparison analysis, calculate the wake influence area according to the wind energy diffusion model, screen the wake influence area, and generate the wake influence coefficient; S4: Based on the wake influence coefficient, obtain the layout spacing of the wind turbine foundation layout drawing and the steering angle of the wind energy capture rate, call the layout spacing and the steering angle to calculate the layout optimization coefficient, adjust the wind turbine layout according to the layout optimization coefficient, and generate the wind turbine optimized layout table; S5: Based on the wind turbine optimized layout table, obtain the blade angle of attack adjustment data and torque load of the wind turbine rated power, call the blade angle of attack adjustment data and the torque load to calculate the power output range, screen the parameters within the power output range, and generate the wind turbine power adjustment parameters.

[0056] The above is only the preferred embodiment of the present invention, and does not limit the present invention in other forms. Any person skilled in the art may use the disclosed technical content to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical solution content of the present invention, any simple modification, equivalent change, and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. An offshore wind farm planning system based on wind speed prediction, characterized in that: The system comprises: The wind energy flow monitoring module obtains the wind speed sequence, wind direction variability, and wind energy density gradient in the offshore wind farm, calculates the wind energy flux distribution value, determines the wind speed gradient anomaly, screens the mutation point, and generates the wind energy distribution trend value; A wind speed mutation prediction module obtains data from wind speed monitoring buoys and offshore turbulence sensors based on the wind energy distribution trend value, calls the wind speed change rate and turbulence intensity change rate, and determines the probability of wind speed mutation; The wind turbine wake impact analysis module obtains the wind turbine wake monitoring radar and sea surface wind shear data based on the wind speed mutation probability, calls the wind speed loss ratio and angle difference, calls the wind energy diffusion model, calculates the wake impact area, screens the over-threshold area, and generates the wake impact coefficient; The wind turbine group intelligent arrangement module obtains the basic arrangement diagram of the wind turbine and the wind energy capture rate based on the wake influence coefficient, calls the arrangement spacing and the steering angle, calculates the arrangement optimization coefficient, and generates the wind turbine optimization arrangement table; The wind farm power output optimization module obtains the wind turbine rated power, blade angle adjustment data, and torque load based on the wind turbine optimization arrangement table, calculates the power output range, selects the optimal parameters within the range, and generates the wind turbine power adjustment parameters.

2. The offshore wind farm planning system based on wind speed prediction according to claim 1 is characterized in that: The wind energy distribution trend value includes wind speed sequence, wind direction variability, wind energy density gradient, wind energy flux distribution value, wind speed gradient anomaly, and mutation point. The wind speed mutation probability includes wind speed monitoring buoy data, offshore turbulence sensor data, wind speed change rate, and turbulence intensity change rate. The wake influence coefficient includes wind turbine wake monitoring radar data, sea surface wind shear data, wind speed loss ratio, angle difference, wind energy diffusion model, wake influence area, and super-threshold area. The wind turbine optimization layout table includes wind turbine basic layout diagram, wind energy capture rate, layout spacing, steering angle, and layout optimization coefficient. The wind turbine power adjustment parameters include wind turbine rated power, blade angle of attack adjustment data, torque load, power output range, and optimal parameters.

3. The offshore wind farm planning system based on wind speed prediction according to claim 2 is characterized in that: The wind energy flow monitoring module comprises: The wind speed sequence extraction submodule obtains wind speed data from multiple monitoring points in the offshore wind farm, arranges the data in time series, calculates the wind speed mean, variance and coefficient of variation, screens and records abnormal wind speed points, and calls wind direction data to analyze the spatial characteristics of wind speed changes to generate a wind speed change sequence; The wind energy density calculation submodule calculates the probability distribution within the wind speed interval based on the wind speed change sequence, and calculates the wind energy density in multiple intervals in combination with the air density parameter, using the formula: ； Calculate and obtain the wind energy density distribution at multiple monitoring points to obtain the wind energy density gradient; in, represents the wind energy density, is the air density, is the representative wind speed in the wind speed range, is the probability of the wind speed interval, is the change of adjacent wind speeds, Indicates wind speed and time The wind speed gain coefficient under the condition of is the wind speed observation height, is the observation period, Represents the number of wind speed intervals; The wind energy flux change analysis submodule calls the wind energy density gradient, calculates the change of wind energy in the differentiated areas of the offshore wind farm, analyzes the wind speed gradient mutation points, screens the wind energy density mutation areas, calculates the wind energy flux change rate, and generates the wind energy flux distribution value; The wind energy trend identification submodule calls the wind energy flux distribution value, calculates the wind energy flux change trend at differentiated time intervals, identifies the long-term trend of wind energy change, calculates the mean square error of wind energy fluctuations through multi-period data, screens mutation points and generates trend curves, and obtains wind energy distribution trend values.

4. The offshore wind farm planning system based on wind speed prediction according to claim 3 is characterized in that: The wind speed mutation prediction module includes: The wind speed monitoring data acquisition submodule obtains the wind speed data of the wind speed monitoring buoy and the turbulence data of the offshore turbulence sensor based on the wind energy distribution trend value, and at the same time calls the time series data stored in the offshore turbulence sensor to extract the wind speed time series, calculates the wind speed difference between adjacent time points and normalizes it, constructs a wind speed change trend sequence, and obtains a wind speed change rate sequence; The turbulence intensity calculation submodule calculates the turbulence intensity in the corresponding time window based on the wind speed change rate sequence, using the formula: ； Obtain a sequence of turbulence intensity values ​​by calculation, screen the turbulence intensity changes within the target wind speed change range, and obtain a sequence of turbulence intensity change rates; in, represents the turbulence intensity, Representative The wind speed at the moment, represents the mean wind speed in the time window, Represents the number of wind speed sampling points in the time window, represents the average temperature in the time window, Representative Turbulence intensity values ​​from offshore turbulence sensors, Represents the average turbulence intensity value of all offshore turbulence sensors, represents the number of offshore turbulence sensors; The wind speed mutation assessment submodule calls the wind speed change rate sequence and the turbulence intensity change rate sequence, calculates the joint probability density distribution, screens the time periods with more drastic wind speed changes, normalizes the wind speed change rate distribution, calculates the mutation risk value in combination with the turbulence intensity change trend, classifies according to the threshold, and obtains the wind speed mutation probability value.

5. The offshore wind farm planning system based on wind speed prediction according to claim 4 is characterized in that: The wind turbine wake impact analysis module includes: The wake monitoring data processing submodule obtains the wind turbine wake monitoring radar data and the sea surface wind shear data based on the wind speed mutation probability, extracts the wind speed mutation probability in the radar data, matches the wind shear data, identifies the wind shear characteristics of the wind speed mutation area, calculates the wind shear change rate, and extracts the wind speed mutation characteristic value based on the matching relationship between the wind speed mutation probability and the wind shear change rate to obtain the wind speed mutation characteristic value set; The wind speed loss calculation submodule, based on the wind speed mutation characteristic value set, calls the wind speed loss ratio and the angle difference, calculates the wind speed loss rate of the wind turbine wake, establishes the wind speed loss rate matrix, and normalizes it, using the formula: ； Calculate and obtain the wind speed loss ratio, and establish a normalized wind speed loss matrix; in, represents the wind speed loss ratio, represents the probability corresponding to the wind shear change rate, Represents the sudden change in wind speed. Represents the angle between the wind shear direction and the wind speed direction, Represents the number of data points in the wind speed mutation feature value set; The wake area assessment submodule calls the normalized wind speed loss matrix and combines it with the wind energy diffusion model to calculate the wake impact area. At the same time, it screens the wind speed loss ratio in the wake impact area, extracts the area where the wind speed loss exceeds the threshold, calculates the wake impact range, and generates the wake impact coefficient.

6. The offshore wind farm planning system based on wind speed prediction according to claim 5 is characterized in that: The fan group intelligent arrangement module includes: The wake influence coefficient calculation submodule obtains the basic layout diagram of the wind turbine based on the wake influence coefficient, calls the wind turbine spacing, wind speed and wind direction data, calculates the wake effect between the wind turbines, calculates the wake influence coefficient through the wind speed attenuation model, and corrects it in combination with the wind turbine height and terrain parameters to obtain the wake influence coefficient matrix; The wind energy capture rate analysis submodule calculates the effective wind speed of the wind turbine after being affected by the wake based on the wake influence coefficient matrix, calls the wind turbine power curve, calculates the wind energy conversion efficiency, and calculates the wind energy capture rate in combination with the wind turbine rated power and meteorological data to obtain the wind energy capture rate matrix; The wind turbine optimization layout submodule calls the layout spacing and steering angle parameters based on the wind energy capture rate matrix, using the formula: ； Calculate the fan optimization layout parameters, adjust the fan layout, and obtain the fan optimization layout table; in, represents the fan optimization layout parameters, Representative Rated power of the fan, Representative The wind energy capture rate of the typhoon turbine, Representative The distance between the typhoon and the adjacent typhoons, Representative The wind turbine wake influence coefficient is Representative The wake diffusion radius of a fan is is the total number of fans, is the number of wind turbines affected by the wake.

7. The offshore wind farm planning system based on wind speed prediction according to claim 6 is characterized in that: The wind farm power output optimization module comprises: The fan power data acquisition submodule obtains the fan rated power in the fan optimization arrangement table, calls the fan blade angle adjustment data, extracts the torque load information, selects the fan data that meets the rated output conditions, calculates the influence of the blade angle change on the power, adjusts the fan torque load influence factor, and establishes the fan power parameter set; The power output interval calculation submodule adopts the formula based on the wind turbine power parameter set: ； Calculate and obtain the power output range of the wind turbine, select the optimal parameters within the range, and obtain the optimal power output range; in, Represents the adjusted power output range, Representative The weight coefficient of the typhoon turbine, Representative Rated power of the fan, Representative The blade angle of attack of the typhoon turbine, represents the reference blade angle of attack, Representative Torque load of the fan, Represents the number of fans, Represents the number of torque load data points; The fan power adjustment parameter generation submodule calls the optimal power output range, adjusts the fan operation parameters, calculates the relationship coefficient between the fan blade angle of attack and power, and obtains the fan power adjustment parameters in combination with the torque load dynamic adjustment parameters.

8. A method for planning an offshore wind farm based on wind speed prediction, characterized in that: The offshore wind farm planning system based on wind speed prediction according to any one of claims 1 to 7 comprises the following steps: S1: Obtain wind speed sequence, wind direction variability, wind energy density gradient, calculate wind energy flux distribution value, judge wind speed gradient anomaly according to wind energy flux distribution value, screen mutation points and calculate and generate wind energy distribution trend value; S2: Based on the wind energy distribution trend value, the wind speed change rate of the wind speed monitoring buoy and the turbulence intensity change rate of the offshore turbulence sensor are obtained, the wind speed change rate and the turbulence intensity change rate are called for matching analysis, the wind speed mutation probability threshold is determined, and the area where the wind speed change rate exceeds the threshold is screened as a risk area, and the wind speed mutation probability is calculated in combination with the turbulence intensity change rate; S3: based on the probability of sudden change in wind speed, obtain the wind speed loss ratio and the angle difference of sea surface wind shear of the wind turbine wake monitoring radar, call the wind speed loss ratio and the angle difference for comparison and analysis, calculate the wake influence area according to the wind energy diffusion model, screen the wake influence area, and generate the wake influence coefficient; S4: Based on the wake influence coefficient, the arrangement spacing of the wind turbine basic arrangement diagram and the steering angle of the wind energy capture rate are obtained, the arrangement spacing and the steering angle are called to calculate the arrangement optimization coefficient, the wind turbine arrangement is adjusted according to the arrangement optimization coefficient, and the wind turbine optimization arrangement table is generated; S5: Based on the fan optimization arrangement table, blade angle adjustment data and torque load of the fan rated power are obtained, the blade angle adjustment data and torque load are called to calculate the power output range, parameters within the power output range are screened, and the fan power adjustment parameters are generated.

Images