Method, system, medium and device for active suppression of abnormal loads on offshore wind turbine generators

Abstract

The present application relates to the field of offshore wind farm control, and discloses a kind of offshore wind turbine load abnormality active inhibition method, system, medium and equipment, comprising: the data of each unit of wind farm is collected to construct initial data set, the composite wind speed of each unit of whole field is calculated to determine the wake superposition intensity index of unit;According to the deviation degree of calculated damage equivalent load relative to the average damage equivalent load of whole field, the load abnormality risk index is determined;The mapping relationship between the yaw angle offset of upstream unit and the lateral deflection of wake center line is established to update the radial distance of downstream unit suffered from wake, and then the wake superposition intensity index and the comprehensive equivalent fatigue load are updated;With the minimization of whole field load abnormality risk index and the minimization of power loss as double target, the optimal yaw angle offset combination is solved with the yaw angle offset of controlled unit as control variable;The dynamic rolling update of yaw angle instruction is carried out, and the optimal yaw angle offset is output for safety check and then issued for execution.

Description

Technical Field

[0001] This invention relates to the field of offshore wind farm control technology, and in particular to a method, system, medium, and equipment for actively suppressing load anomalies in offshore wind turbines considering wake superposition effects. Background Technology

[0002] Offshore wind farms typically employ an array layout, with high turbine density and continuously increasing individual turbine capacity. When natural wind blows over upstream wind turbines, the rotors extract energy from the wind, leading to a decrease in downstream wind speed and an increase in turbulence intensity—a phenomenon known as the wake effect. In large offshore wind farms, the wakes of multiple turbines overlap, creating a wake superposition effect that can affect areas ranging from tens of kilometers to over 100 kilometers.

[0003] The main hazards caused by the wake superposition effect include: (1) a significant reduction in downwind wind speed, resulting in an overall loss of power generation in the wind farm, with power loss reaching 30% to 40%; (2) a significant increase in turbulence intensity, causing a significant increase in fatigue load on downstream wind turbine units; and (3) uneven distribution of fatigue load on downstream turbine units affected by the wake over a long period of time, leading to an increase in failure rate and maintenance costs. Studies have shown that wake-induced power loss ranges from 5% to 40%, while under conditions of high turbulence or yaw misalignment, the structural load increase can exceed 30%.

[0004] The existing wake management and load suppression methods have the following shortcomings: (1) Existing methods mostly adopt fixed layout optimization or post-maintenance strategies, which cannot actively adjust the unit's operating status when the wind conditions change to avoid the area with the most severe wake superposition. They mainly rely on passive response and lack the ability to actively predict. (2) Existing wake analysis methods mainly focus on the impact of wind speed attenuation on power generation. They use simplified analytical algorithms and are insufficient in characterizing the turbulence distribution and load characteristics after the wake superposition of multiple units. (3) Existing active wake control focuses on power enhancement and pays insufficient attention to the suppression of fatigue loads of downstream units. Studies have shown that although active yaw control can enhance power generation, it also changes the load distribution of wind turbines. (4) Existing methods mostly rely on offline model calculations and fail to make full use of real-time operating data such as SCADA to dynamically perceive and adaptively control the wake superposition state. Summary of the Invention

[0005] To address the aforementioned problems, the present invention aims to provide a method, system, medium, and equipment for actively suppressing abnormal loads on offshore wind turbines. By real-time sensing of the wake superposition state, quantifying the risk of abnormal loads, and dynamically optimizing the yaw angle of the upstream turbine, the invention achieves active deflection of the wake path and coordinated suppression of the load.

[0006] To achieve the above objectives, in a first aspect, the technical solution adopted by the present invention is as follows: a method for actively suppressing load anomalies in offshore wind turbines, comprising: a data acquisition and preprocessing step: acquiring SCADA operation data and status data of each turbine in the wind farm to construct an initial dataset; a wake superposition dynamic sensing step: based on the Gaussian wake model and the principle of energy conservation superposition, calculating the composite wind speed of each turbine in the entire field according to the initial dataset, and determining the wake superposition intensity index of each turbine according to the composite wind speed and the incoming wind speed; a load anomaly quantification and evaluation step: calculating the comprehensive equivalent fatigue load based on the load time series data of each turbine within the sliding time window, and determining the load anomaly risk index according to the deviation of the comprehensive equivalent fatigue load of each turbine from the average comprehensive equivalent fatigue load of the entire field; Yaw Angle-Load Mapping Steps: Establish a mapping relationship between the yaw angle offset of the upstream unit and the lateral deflection of the wake centerline. Update the radial distance of the downstream unit to the wake based on the mapping relationship, and then update the wake superposition intensity index and the comprehensive equivalent fatigue load. Multi-Objective Yaw Angle Optimization Steps: With minimizing the overall load anomaly risk index and minimizing power loss as dual objectives, and the yaw angle offset of the controlled unit as the control variable, use a multi-objective particle swarm optimization algorithm to solve for the optimal combination of yaw angle offsets. Rolling Optimization Update and Output Steps: Based on wind condition prediction data, cyclically execute the wake superposition dynamic sensing step to the multi-objective yaw angle optimization step to realize the dynamic rolling update of the yaw angle command. Output the optimal yaw angle offset of each controlled unit and issue it for execution after safety verification.

[0007] Furthermore, the specific implementation process of the wake superimposed dynamic sensing step includes: Establish a Gaussian wake field for a single unit. For any upstream unit, determine the wind speed loss rate at a set downstream position based on the real-time thrust coefficient, rotor diameter, and set wake attenuation constant of the upstream unit. For any downstream target unit, the composite wind speed at the hub of the downstream target unit is calculated by the wind speed loss rate at each corresponding location of all upstream units that have a wake effect on the target unit. The wake superposition intensity index of the downstream target unit is determined by the percentage of the difference between the incoming wind speed and the resultant wind speed relative to the incoming wind speed.

[0008] Furthermore, the specific implementation process of the load anomaly quantification assessment step includes: For each unit, the tower bottom bending moment load and blade root bending moment load of the unit are collected in real time. The real-time rainflow counting algorithm is used to count the load time series data composed of tower bottom bending moment load and blade root bending moment load within the sliding time window, obtain the amplitude of each load cycle and its corresponding cycle number, and calculate the equivalent fatigue load of the unit according to the Miner linear cumulative damage criterion. The arithmetic mean of the equivalent fatigue loads of all organic components in the entire field is taken as the average comprehensive equivalent fatigue load of the entire field. The load anomaly risk index of the unit is calculated based on the equivalent fatigue load of each unit and the average comprehensive equivalent fatigue load of the entire field. When the load anomaly risk index of a certain unit exceeds a preset threshold, active yaw control of that unit is triggered.

[0009] Furthermore, the mapping relationship between the upstream unit's yaw angle offset and the lateral deflection of the wake centerline is as follows: ; in, This is the lateral deflection of the wake centerline; This represents the yaw angle offset generated by upstream unit j. Downstream distance; The wind direction angle is denoted as . When the wake deflects, the radial distance of the wake to the downstream unit is updated as the wake deflects, thereby updating the wake superposition intensity index and the comprehensive equivalent fatigue load of the downstream unit.

[0010] Furthermore, the specific implementation process of the multi-objective yaw angle optimization steps includes: The yaw angle offset of the front and middle rows of the entire field is selected as the control variable, and the yaw angle offset of each controlled unit meets the set constraints. With minimizing the overall load anomaly risk index and minimizing power loss as the two objectives, a bi-objective function with a penalty term is constructed. When using the multi-objective particle swarm optimization algorithm, a population containing multiple particles is initialized. The position vector of each particle represents a combination of yaw angle offsets. The sum of squares of the global load anomaly risk index and the power loss rate corresponding to each particle are calculated as the fitness vector of that particle. The Pareto dominance principle is used to update the individual historical best position of each particle and the global best leader in the external archive. An adaptive grid method is used to maintain the diversity of the external archive. After a preset number of iterations, the optimal compromise solution is selected from the Pareto front solution set based on the fuzzy membership function as the optimal yaw angle offset combination.

[0011] Furthermore, the biobjective function with a penalty term is:

[0012] in, M represents the power loss caused by yaw of the upstream units, and M is the total number of units in the front and middle rows of the entire field. To control variables, For the j-th upstream unit, the yaw angle offset due to execution The resulting power loss per unit, This represents the yaw angle offset generated by upstream unit j. This represents the total power across the field without yaw. Take 0.7, A value of 0.3 is used to prioritize load balance. This represents the total number of wind turbines in the wind farm. This is the load anomaly risk index for unit i.

[0013] Furthermore, based on wind condition forecast data, the wake overlay dynamic sensing step is iteratively executed to a multi-target yaw angle optimization step, achieving dynamic rolling updates of yaw angle commands, including: Wind speed and direction prediction data for a preset time period are obtained from wind farm meteorological meteorological towers or numerical weather prediction. In the current control cycle, the wind speed and direction prediction sequence is input into the optimizer to solve for the optimal yaw angle sequence in the future time domain. Then, only the optimal yaw angle command at the current moment is executed and the yaw angle commands at subsequent moments are discarded. When the next control cycle arrives, based on the latest collected measured SCADA operation data and updated wind speed and direction prediction data, the wake superimposed dynamic sensing, load anomaly quantification assessment, yaw angle-load mapping and multi-objective yaw angle optimization are re-executed to solve for the new optimal yaw angle command for the current moment and issue it for execution. This process is repeated to achieve dynamic rolling update of the yaw angle command.

[0014] Secondly, the technical solution adopted by this invention is as follows: an active suppression system for load anomalies in offshore wind turbines, comprising: a data acquisition and preprocessing module: acquiring SCADA operation data and status data of each turbine in the wind farm to construct an initial dataset; a wake superposition dynamic sensing module: based on the Gaussian wake model and the principle of energy conservation superposition, calculating the composite wind speed of each turbine in the entire field according to the initial dataset, and determining the wake superposition intensity index of each turbine according to the composite wind speed and the incoming wind speed; a load anomaly quantification and evaluation module: calculating the comprehensive equivalent fatigue load based on the load time series data of each turbine within the sliding time window, and determining the load anomaly risk index according to the deviation of the comprehensive equivalent fatigue load of each turbine from the average comprehensive equivalent fatigue load of the entire field; yaw angle-load... Load mapping module: Establishes a mapping relationship between the yaw angle offset of the upstream unit and the lateral deflection of the wake centerline. Based on the mapping relationship, it updates the radial distance of the wake received by the downstream unit, and then updates the wake superposition intensity index and the comprehensive equivalent fatigue load. Multi-objective yaw angle optimization module: With minimizing the overall load anomaly risk index and minimizing power loss as dual objectives, and using the yaw angle offset of the controlled unit as the control variable, it uses a multi-objective particle swarm optimization algorithm to solve for the optimal combination of yaw angle offsets. Rolling optimization update and output module: Based on wind condition prediction data, it iteratively executes the wake superposition dynamic sensing step to the multi-objective yaw angle optimization step to realize the dynamic rolling update of yaw angle commands; and outputs the optimal yaw angle offset of each controlled unit and issues it for execution after safety verification.

[0015] Thirdly, the technical solution adopted by the present invention is: a computer-readable storage medium for storing one or more programs, wherein the one or more programs include instructions that, when executed by a computing device, cause the computing device to perform any of the above-described methods.

[0016] Fourthly, the technical solution adopted by the present invention is as follows: a computing device, comprising: a memory, one or more processors and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include instructions for performing any of the above methods.

[0017] The present invention has the following advantages due to the adoption of the above technical solutions: 1. This invention provides real-time sensing of wake superposition effects, resulting in strong early warning capabilities. By employing a Gaussian wake model and the principle of energy conservation superposition, combined with real-time SCADA data, a dynamic wake field sensing capability is constructed. This capability can quantify the wake superposition intensity index (WSIi) of each unit in real time, enabling the early identification of areas with severe wake superposition compared to traditional offline models.

[0018] 2. This invention establishes a correlation model between wake superposition intensity and load characteristics, and comprehensively assesses the load anomaly risk by integrating multiple dimensions such as load fluctuation amplitude and rate of change, thus overcoming the one-sidedness of traditional methods that rely solely on wind speed attenuation assessment.

[0019] 3. This invention takes full-field load equalization and power loss minimization as dual objectives. It solves the optimal yaw angle combination through multi-objective particle swarm optimization, thereby achieving coordinated optimization of wake deflection control and load suppression. The control objectives take into account both power and load. Attached Figure Description

[0020] Figure 1 This is a flowchart of the active suppression method for abnormal loads on offshore wind turbines in an embodiment of the present invention. Detailed Implementation

[0021] To address the technical problems existing in the prior art, this invention provides a method, system, medium, and equipment for actively suppressing load anomalies in offshore wind turbines. The method includes: constructing a dynamic perception model of wake superposition across the entire wind farm based on real-time SCADA data to quantify the wake superposition intensity and load anomaly risk experienced by each turbine; using the turbine yaw angle as the control variable and aiming at both load equalization and power loss minimization, solving for the optimal yaw angle combination through a multi-objective optimization algorithm; and employing rolling time-domain predictive control to achieve dynamic rolling optimization of the yaw angle. This invention achieves real-time perception of wake superposition effects, quantitative assessment of load anomaly risk, and active coordinated control of the yaw angle, effectively solving the problems of passive response, coarse assessment, and single control objective in existing methods.

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention are within the scope of protection of the present invention.

[0023] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0024] In one embodiment of the present invention, a method for actively suppressing abnormal loads on offshore wind turbines is provided. In this embodiment, as... Figure 1 As shown, the method includes the following steps: 1) Data acquisition and preprocessing steps: Collect SCADA (Supervisory Control and Data Acquisition) operation data and status data of each unit in the wind farm to build an initial dataset.

[0025] 2) Wake superposition dynamic sensing step: Based on the Gaussian wake model and the principle of energy conservation superposition, the composite wind speed of each unit in the field is calculated according to the initial dataset, and the wake superposition intensity index of each unit is determined according to the composite wind speed and the incoming wind speed.

[0026] 3) Load anomaly quantitative assessment steps: Calculate the comprehensive equivalent fatigue load based on the load time sequence data of each unit within the sliding time window, and determine the load anomaly risk index according to the degree of deviation of the comprehensive equivalent fatigue load of each unit from the overall average comprehensive equivalent fatigue load.

[0027] 4) Yaw angle-load mapping steps: Establish the mapping relationship between the yaw angle offset of the upstream unit and the lateral deflection of the wake centerline. Update the radial distance of the wake received by the downstream unit according to the mapping relationship, and then update the wake superposition intensity index and the comprehensive equivalent fatigue load.

[0028] 5) Multi-objective yaw angle optimization steps: With minimizing the overall load anomaly risk index and minimizing power loss as the dual objectives, and the yaw angle offset of the controlled unit as the control variable, the multi-objective particle swarm optimization algorithm is used to solve for the optimal combination of yaw angle offsets.

[0029] 6) Rolling optimization update and output steps: Based on wind condition prediction data, the wake superposition dynamic sensing step and multi-target yaw angle optimization step are executed cyclically to realize the dynamic rolling update of yaw angle commands; and the optimal yaw angle offset of each controlled unit is output and issued for execution after safety verification.

[0030] In step 1) above, the initial dataset constructed includes: the incoming wind speed at time t. Angle with incoming wind direction ; Measured wind speed at the hub of each unit at time t Output power Generator speed Yaw angle of each unit With pitch angle Front and rear directions of each unit's tower base Bending moment and lateral bending moment Blade root waving moment of each unit and oscillation moment .

[0031] Where i = 1, 2, …, N represents the unit number, and N is the total number of wind farm units. Median filtering is applied to the collected time-series data to eliminate communication transient interference and sensor spike noise.

[0032] In step 2) above, the specific implementation process of the wake superimposed dynamic sensing step includes the following steps: 2.1) Establish the Gaussian wake field of a single unit. For any upstream unit, determine the wind speed loss rate at a set downstream position based on the real-time thrust coefficient, rotor diameter, and set wake attenuation constant of the upstream unit.

[0033] Specifically, this embodiment uses the Bastankhah Gaussian wake model to describe the velocity distribution downstream of the upstream unit. For the upstream unit j, the wind speed deficit rate at its downstream distance x and radial distance r is... for:

[0034] in, is the real-time thrust coefficient of unit j (obtained by interpolation of the wind speed-thrust curve); The diameter of the wind turbine; The wake attenuation constant; The standard deviation of the wake cross section represents the lateral width of the wake. Let be the wake velocity at coordinates (x, r) downstream of the j-th unit. The distance along the direction of the incoming wind. The radial distance is perpendicular to the flow direction; For incoming wind speed.

[0035] 2.2) For any downstream target unit, the composite wind speed at the hub of the downstream target unit is calculated by the wind speed loss rate at each corresponding location of all upstream units that have a wake effect on the target unit.

[0036] Specifically, due to the dense arrangement of turbines within a wind farm, downstream turbines are typically affected by the wakes of multiple upstream turbines simultaneously. Therefore, this invention employs the energy conservation (Katic) superposition principle to calculate the composite wind speed at the hub of the i-th turbine. :

[0037] in, The set of all upstream units that have a wake effect on unit i; The distance in the flow direction from unit j to unit i; This is the lateral offset distance (determined by the wind direction and the relative position of the unit).

[0038] 2.3) Determine the wake superposition intensity index of the downstream target unit based on the percentage of the difference between the incoming wind speed and the resultant wind speed relative to the incoming wind speed.

[0039] Specifically, the wake superposition intensity index of unit i at the current moment is defined. :

[0040] This index directly reflects the degree to which unit i is affected by wake superposition. When the value is >15%, the unit is determined to be in the strong wake influence zone, and proceeds to step 3).

[0041] In step 3) above, to avoid the problem of unclear physical meaning of traditional time-domain statistical indicators (such as amplitude and rate of change), this embodiment introduces damage equivalent load (DEL) as the core evaluation indicator. Its physical meaning is a constant amplitude load that produces the same amount of fatigue damage at a specified number of cycles. Specifically, the load anomaly quantification evaluation step includes the following steps: 3.1) For each unit i, the tower base bending moment load of that unit is collected in real time. Leaf root bending moment load A real-time rainflow counting algorithm is used to cyclically count the load time series data consisting of tower base bending moment load and blade root bending moment load within the sliding time window T, and obtain the amplitude of each load cycle. and the corresponding number of loops The comprehensive equivalent fatigue load of the unit was calculated based on Miner's linear cumulative damage criterion.

[0042] Specifically, in this embodiment, the sliding time window T is set to 10 minutes. Equivalent fatigue load. The calculation formula is:

[0043] In the formula, m is the slope of the material's SN curve; This represents the equivalent cycle number. During the calculation, the equivalent load at the tower base was obtained. Equivalent load of leaf root The final comprehensive load index is the weighted sum of the two values ​​as the comprehensive equivalent fatigue load. : , in Take 0.6.

[0044] 3.2) The arithmetic mean of the comprehensive equivalent fatigue load of all organic components in the entire field is taken as the average comprehensive equivalent fatigue load of the entire field. .Right now: .

[0045] 3.3) The load anomaly risk index of the unit is calculated based on the comprehensive equivalent fatigue load of each unit and the average comprehensive equivalent fatigue load of the entire field.

[0046] Within the same wind farm, due to consistent wind conditions, the load fluctuations of healthy turbines should exhibit statistical similarity. This invention defines a load anomaly risk index. The load anomaly risk index is the degree of deviation of the current unit's DEL from the field average DEL. for:

[0047] When the load abnormality risk index of a certain unit Exceeding the preset threshold This indicates that the unit is experiencing fatigue loads significantly higher than the average level, triggering active yaw control of the unit to proceed to step 4).

[0048] The specific process of active yaw control is as follows: Step 4) establishes the yaw angle offset of the upstream unit. The quantitative mapping relationship of downstream unit loads is established; the yaw angle offset of the front and middle rows of units in the field is used as the control variable, and the optimal yaw angle combination is solved between load balancing and power efficiency through the multi-objective particle swarm optimization algorithm in step 5); then, according to the rolling time domain predictive control strategy in step 6), the yaw command is dynamically updated based on the wind speed and direction prediction sequence for the next 30 minutes; finally, after safety verification, it is issued for execution, realizing the closed-loop active suppression of "perception-evaluation-optimization-execution-re-perception".

[0049] In step 4) above, when the upstream unit j generates a yaw angle offset... At this time, its wake centerline will deflect laterally. According to the yaw wake steering model, the lateral deflection of the wake centerline is... Yaw angle offset from upstream unit The mapping relationship between them is as follows: ; in, This is the lateral deflection of the wake centerline; This represents the yaw angle offset generated by upstream unit j. Downstream distance; The wind direction angle; the radial distance of the wake experienced by the downstream unit after the wake deflection. The lateral offset distance is updated as the wake deflects, thereby updating the wake superposition intensity index of the downstream units. and comprehensive equivalent fatigue load .

[0050] In step 5) above, the specific implementation process of the multi-target yaw angle optimization step includes the following steps: 5.1) Select the yaw angle offset of the front and middle rows of all aircraft (M in total). As control variables, the aforementioned M units are the controlled units in this step. The yaw angle offset of each controlled unit satisfies the set constraints.

[0051] Specifically, due to the mechanical characteristics of the yaw system and the safety limitations of the unit, the constraints are as follows: .

[0052] 5.2) With minimizing the overall load anomaly risk index and minimizing power loss as the two objectives, a bi-objective function with a penalty term is constructed.

[0053] Specifically, in this embodiment, the biobjective function with a penalty term is:

[0054] in, M represents the power loss caused by yaw of the upstream units, and M is the total number of units in the front and middle rows of the entire field. To control variables, For the j-th upstream unit, the yaw angle offset due to execution The resulting power loss per unit, This represents the yaw angle offset generated by upstream unit j. This represents the total power across the field without yaw. Take 0.7, A value of 0.3 is used to prioritize load balance. This represents the total number of wind turbines in the entire wind farm. This is the load anomaly risk index for unit i.

[0055] 5.3) When using the multi-objective particle swarm optimization algorithm to solve the problem, initialize a population containing multiple particles, and the position vector of each particle... This represents a set of yaw angle offset combinations. The sum of squares of the global load anomaly risk index and the power loss rate for each particle are calculated as its fitness vector. The Pareto dominance principle is used to update the individual historical best position of each particle and the global best leader in the external archive. An adaptive grid method is used to maintain the diversity of the external archive to avoid getting trapped in local optima. After a preset number of iterations (e.g., 200), the optimal compromise solution is selected from the Pareto front solution set based on a fuzzy membership function compromise. This serves as the optimal combination of yaw angle offsets.

[0056] In step 6) above, given the significant random fluctuations in sea wind speed and direction, this invention introduces a rolling time-domain predictive control (RHPC) strategy. Specifically, based on wind condition prediction data, the wake superposition dynamic sensing step is iteratively executed to a multi-target yaw angle optimization step, realizing the dynamic rolling update of the yaw angle command, including the following steps: 6.1) Obtain wind speed and direction forecast data for a preset time period based on the wind farm's meteorological meteorological tower or numerical weather prediction (NWP). For example, the preset time period is... minute.

[0057] 6.2) In the current control cycle The wind speed and direction prediction sequence is input into the MOPSO optimizer to solve for the optimal yaw angle sequence in the future time domain. Then, only the optimal yaw angle command at the current time is executed and the yaw angle commands at subsequent time times are discarded.

[0058] 6.3) In the next control cycle Upon arrival ( and (At 5-minute intervals), based on the latest collected measured SCADA operation data and updated wind speed and direction prediction data, the wake superimposed dynamic sensing, load anomaly quantification assessment, yaw angle-load mapping, and multi-objective yaw angle optimization are re-executed to solve for the new optimal yaw angle command at the current moment and issue it for execution. This process is repeated to achieve dynamic rolling updates of the yaw angle command: sensing → mapping → optimization → execution → re-sensing.

[0059] In this embodiment, the optimal yaw angle offset combination of each controlled unit is output. The command is then sent to the unit's PLC for execution. A safety check is performed before sending the command: if... If the instantaneous increase in the tower base load exceeds 120% of the rated load, a limiting effect will be implemented. Retract to a safe range. Finally, continuously monitor the overall field load anomaly risk index. This ensures that the load distribution across the entire field converges to an equilibrium state.

[0060] In one embodiment of the present invention, an active suppression system for abnormal loads of offshore wind turbines is provided, comprising: Data acquisition and preprocessing module: Collects SCADA operation data and status data of each unit in the wind farm and constructs an initial dataset; Wake superposition dynamic sensing module: Based on the Gaussian wake model and the principle of energy conservation superposition, it calculates the composite wind speed of each unit in the field according to the initial dataset, and determines the wake superposition intensity index of each unit according to the composite wind speed and the incoming wind speed. Load anomaly quantitative assessment module: Calculates the comprehensive equivalent fatigue load based on the load time sequence data of each unit within the sliding time window, and determines the load anomaly risk index according to the degree of deviation of the comprehensive equivalent fatigue load of each unit from the overall average comprehensive equivalent fatigue load. Yaw angle-load mapping module: Establishes the mapping relationship between the yaw angle offset of the upstream unit and the lateral deflection of the wake centerline. Updates the radial distance of the wake received by the downstream unit according to the mapping relationship, and then updates the wake superposition intensity index and the comprehensive equivalent fatigue load. Multi-objective yaw angle optimization module: With minimizing the overall load anomaly risk index and minimizing power loss as dual objectives, and with the yaw angle offset of the controlled unit as the control variable, the multi-objective particle swarm optimization algorithm is used to solve for the optimal combination of yaw angle offsets. Rolling optimization update and output module: Based on wind condition prediction data, the module iteratively executes the wake superposition dynamic sensing step to the multi-target yaw angle optimization step to realize the dynamic rolling update of yaw angle commands; and outputs the optimal yaw angle offset of each controlled unit and issues it for execution after safety verification.

[0061] In the above embodiments, the specific implementation process of the wake superimposed dynamic sensing step includes: Establish a Gaussian wake field for a single unit. For any upstream unit, determine the wind speed loss rate at a set downstream position based on the real-time thrust coefficient, rotor diameter, and set wake attenuation constant of the upstream unit. For any downstream target unit, the composite wind speed at the hub of the downstream target unit is calculated by the wind speed loss rate at each corresponding location of all upstream units that have a wake effect on the target unit. The wake superposition intensity index of the downstream target unit is determined by the percentage of the difference between the incoming wind speed and the resultant wind speed relative to the incoming wind speed.

[0062] In the above embodiments, the specific implementation process of the load anomaly quantification assessment step includes: For each unit, the tower bottom bending moment load and blade root bending moment load of the unit are collected in real time. The real-time rainflow counting algorithm is used to count the load time series data composed of tower bottom bending moment load and blade root bending moment load within the sliding time window, obtain the amplitude of each load cycle and its corresponding cycle number, and calculate the equivalent fatigue load of the unit according to the Miner linear cumulative damage criterion. The arithmetic mean of the equivalent fatigue loads of all organic components in the entire field is taken as the average comprehensive equivalent fatigue load of the entire field. The load anomaly risk index of the unit is calculated based on the equivalent fatigue load of each unit and the average comprehensive equivalent fatigue load of the entire field. When the load anomaly risk index of a certain unit exceeds a preset threshold, active yaw control of that unit is triggered.

[0063] In the above embodiments, the mapping relationship between the upstream unit's yaw angle offset and the lateral deflection of the wake centerline is as follows: ; in, This is the lateral deflection of the wake centerline; This represents the yaw angle offset generated by upstream unit j. Downstream distance; The wind direction angle is denoted as . When the wake deflects, the radial distance of the wake to the downstream unit is updated as the wake deflects, thereby updating the wake superposition intensity index and the comprehensive equivalent fatigue load of the downstream unit.

[0064] In the above embodiments, the specific implementation process of the multi-target yaw angle optimization step includes: The yaw angle offset of the front and middle rows of the entire field is selected as the control variable, and the yaw angle offset of each controlled unit meets the set constraints. With minimizing the overall load anomaly risk index and minimizing power loss as the two objectives, a bi-objective function with a penalty term is constructed. When using the multi-objective particle swarm optimization algorithm, a population containing multiple particles is initialized. The position vector of each particle represents a combination of yaw angle offsets. The sum of squares of the global load anomaly risk index and the power loss rate corresponding to each particle are calculated as the fitness vector of that particle. The Pareto dominance principle is used to update the individual historical best position of each particle and the global best leader in the external archive. An adaptive grid method is used to maintain the diversity of the external archive. After a preset number of iterations, the optimal compromise solution is selected from the Pareto front solution set based on the fuzzy membership function as the optimal yaw angle offset combination.

[0065] In this embodiment, the biobjective function with a penalty term is:

[0066] in, M represents the power loss caused by yaw of the upstream units, and M is the total number of units in the front and middle rows of the entire field. To control variables, For the j-th upstream unit, the yaw angle offset due to execution The resulting power loss per unit, This represents the yaw angle offset generated by upstream unit j. This represents the total power across the field without yaw. Take 0.7, A value of 0.3 is used to prioritize load balance. This represents the total number of wind turbines in the wind farm. This is the load anomaly risk index for unit i.

[0067] In the above embodiments, the wake overlay dynamic sensing step and the multi-target yaw angle optimization step are executed iteratively based on wind condition prediction data to achieve dynamic rolling updates of the yaw angle command, including: Wind speed and direction prediction data for a preset time period are obtained from wind farm meteorological meteorological towers or numerical weather prediction. In the current control cycle, the wind speed and direction prediction sequence is input into the optimizer to solve for the optimal yaw angle sequence in the future time domain. Then, only the optimal yaw angle command at the current moment is executed and the yaw angle commands at subsequent moments are discarded. When the next control cycle arrives, based on the latest collected measured SCADA operation data and updated wind speed and direction prediction data, the wake superimposed dynamic sensing, load anomaly quantification assessment, yaw angle-load mapping and multi-objective yaw angle optimization are re-executed to solve for the new optimal yaw angle command for the current moment and issue it for execution. This process is repeated to achieve dynamic rolling update of the yaw angle command.

[0068] The system provided in this embodiment is used to execute the above-described method embodiments. For specific processes and details, please refer to the above embodiments, which will not be repeated here.

[0069] In one embodiment of the present invention, a computing device is provided, which may be a terminal and may include: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by one or more processors, and the one or more programs include instructions for performing any of the methods in the above embodiments.

[0070] Furthermore, when the aforementioned instructions can be implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0071] In one embodiment of the present invention, a computer program product is provided, the computer program product including a computer program stored on a non-transitory computer-readable storage medium, the computer program including program instructions, and when the program instructions are executed by a computer, the computer is able to perform the methods provided in the above-described method embodiments.

[0072] In one embodiment of the present invention, a non-transitory computer-readable storage medium is provided, which stores server instructions that cause a computer to perform the methods provided in the above embodiments.

[0073] The computer-readable storage medium provided in the above embodiments has a similar implementation principle and technical effect to the above method embodiments, and will not be described again here.

[0074] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0075] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0076] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0077] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for actively suppressing abnormal loads in offshore wind turbines, characterized in that, include: Data acquisition and preprocessing steps: Collect SCADA operation data and status data of each unit in the wind farm to construct an initial dataset; Wake superposition dynamic sensing steps: Based on the Gaussian wake model and the principle of energy conservation superposition, the composite wind speed of each unit in the field is calculated according to the initial dataset, and the wake superposition intensity index of each unit is determined according to the composite wind speed and the incoming wind speed. Load anomaly quantitative assessment steps: Calculate the comprehensive equivalent fatigue load based on the load time series data of each unit within the sliding time window, and determine the load anomaly risk index according to the degree of deviation of the comprehensive equivalent fatigue load of each unit from the overall average comprehensive equivalent fatigue load. Yaw angle-load mapping steps: Establish the mapping relationship between the yaw angle offset of the upstream unit and the lateral deflection of the wake centerline. Update the radial distance of the wake received by the downstream unit according to the mapping relationship, and then update the wake superposition intensity index and the comprehensive equivalent fatigue load. The multi-objective yaw angle optimization process involves two objectives: minimizing the overall load anomaly risk index and minimizing power loss. The yaw angle offsets of the controlled units are used as control variables. A multi-objective particle swarm optimization algorithm is employed to solve for the optimal combination of yaw angle offsets. Specifically, the multi-objective yaw angle optimization process includes: selecting the yaw angle offsets of a predetermined number of units in the front and middle rows as control variables, ensuring that the yaw angle offsets of each controlled unit satisfy set constraints; constructing a bi-objective function with a penalty term, using the minimization of the overall load anomaly risk index and power loss as the two objectives; and employing a multi-objective particle swarm optimization algorithm. When solving the swarm optimization algorithm, a population containing multiple particles is initialized. The position vector of each particle represents a combination of yaw angle offsets. The sum of squares of the global load anomaly risk index and the power loss rate corresponding to each particle are calculated as the fitness vector of that particle. The Pareto dominance principle is used to update the individual historical best position of each particle and the global best leader in the external archive. An adaptive grid method is used to maintain the diversity of the external archive. After a preset number of iterations, the optimal compromise solution is selected from the Pareto front solution set based on the fuzzy membership function as the optimal combination of yaw angle offsets. The biobjective function with a penalty term is: ； in, M represents the power loss caused by yaw of the upstream units, and M is the total number of units in the front and middle rows of the entire field. To control variables, For the j-th upstream unit, the yaw angle offset due to execution The resulting power loss per unit, This represents the yaw angle offset generated by upstream unit j. This represents the total power across the field without yaw. Take 0.7, A value of 0.3 is used to prioritize load balance. This represents the total number of wind turbines in the wind farm. The load anomaly risk index for unit i; Rolling optimization update and output steps: Based on wind condition prediction data, the wake superposition dynamic sensing step and multi-target yaw angle optimization step are executed cyclically to realize the dynamic rolling update of yaw angle commands; and the optimal yaw angle offset of each controlled unit is output and issued for execution after safety verification.

2. The active suppression method for abnormal loads of offshore wind turbines as described in claim 1, characterized in that, The specific implementation process of the wake superposition dynamic sensing step includes: Establish a Gaussian wake field for a single unit. For any upstream unit, determine the wind speed loss rate at a set downstream position based on the real-time thrust coefficient, rotor diameter, and set wake attenuation constant of the upstream unit. For any downstream target unit, the composite wind speed at the hub of the downstream target unit is calculated by the wind speed loss rate at each corresponding location of all upstream units that have a wake effect on the target unit. The wake superposition intensity index of the downstream target unit is determined by the percentage of the difference between the incoming wind speed and the resultant wind speed relative to the incoming wind speed.

3. The active suppression method for abnormal loads of offshore wind turbines as described in claim 1, characterized in that, The specific implementation process of the load anomaly quantification assessment steps includes: For each unit, the tower bottom bending moment load and blade root bending moment load of the unit are collected in real time. The real-time rainflow counting algorithm is used to count the load time series data composed of tower bottom bending moment load and blade root bending moment load within the sliding time window, obtain the amplitude of each load cycle and its corresponding cycle number, and calculate the equivalent fatigue load of the unit according to the Miner linear cumulative damage criterion. The arithmetic mean of the equivalent fatigue loads of all organic components in the entire field is taken as the average comprehensive equivalent fatigue load of the entire field. The load anomaly risk index of the unit is calculated based on the equivalent fatigue load of each unit and the average comprehensive equivalent fatigue load of the entire field. When the load anomaly risk index of a certain unit exceeds a preset threshold, active yaw control of that unit is triggered.

4. The active suppression method for abnormal loads of offshore wind turbines as described in claim 1, characterized in that, The mapping relationship between the upstream unit's yaw angle offset and the lateral deflection of the wake centerline is as follows: ； in, This is the lateral deflection of the wake centerline; This represents the yaw angle offset generated by upstream unit j. Downstream distance; The wind direction angle is denoted as . When the wake deflects, the radial distance of the wake to the downstream unit is updated as the wake deflects, thereby updating the wake superposition intensity index and the comprehensive equivalent fatigue load of the downstream unit.

5. The active suppression method for abnormal loads on offshore wind turbines as described in claim 1, characterized in that, Based on wind condition forecast data, the wake overlay dynamic sensing step and multi-target yaw angle optimization step are executed cyclically to achieve dynamic rolling updates of yaw angle commands, including: Wind speed and direction prediction data for a preset time period are obtained from wind farm meteorological meteorological towers or numerical weather prediction. In the current control cycle, the wind speed and direction prediction sequence is input into the optimizer to solve for the optimal yaw angle sequence in the future time domain. Then, only the optimal yaw angle command at the current moment is executed and the yaw angle commands at subsequent moments are discarded. When the next control cycle arrives, based on the latest collected measured SCADA operation data and updated wind speed and direction prediction data, the wake superimposed dynamic sensing, load anomaly quantification assessment, yaw angle-load mapping and multi-objective yaw angle optimization are re-executed to solve for the new optimal yaw angle command for the current moment and issue it for execution. This process is repeated to achieve dynamic rolling update of the yaw angle command.

6. An active suppression system for abnormal loads on offshore wind turbines, used to implement the active suppression method for abnormal loads on offshore wind turbines as described in claim 1, characterized in that, include: Data acquisition and preprocessing module: Collects SCADA operation data and status data of each unit in the wind farm and constructs an initial dataset; Wake superposition dynamic sensing module: Based on the Gaussian wake model and the principle of energy conservation superposition, it calculates the composite wind speed of each unit in the field according to the initial dataset, and determines the wake superposition intensity index of each unit according to the composite wind speed and the incoming wind speed. Load anomaly quantitative assessment module: Calculates the comprehensive equivalent fatigue load based on the load time sequence data of each unit within the sliding time window, and determines the load anomaly risk index according to the degree of deviation of the comprehensive equivalent fatigue load of each unit from the overall average comprehensive equivalent fatigue load. Yaw angle-load mapping module: Establishes the mapping relationship between the yaw angle offset of the upstream unit and the lateral deflection of the wake centerline. Updates the radial distance of the wake received by the downstream unit according to the mapping relationship, and then updates the wake superposition intensity index and the comprehensive equivalent fatigue load. Multi-objective yaw angle optimization module: With minimizing the overall load anomaly risk index and minimizing power loss as dual objectives, and with the yaw angle offset of the controlled unit as the control variable, the multi-objective particle swarm optimization algorithm is used to solve for the optimal combination of yaw angle offsets. Rolling optimization update and output module: Based on wind condition prediction data, the wake superimposed dynamic sensing step and multi-target yaw angle optimization step are executed cyclically to realize the dynamic rolling update of yaw angle commands; It outputs the optimal yaw angle offset for each controlled unit and performs a safety check before issuing the command for execution.

7. A computer-readable storage medium for storing one or more programs, characterized in that, The one or more programs include instructions that, when executed by a computing device, cause the computing device to perform any of the methods claimed in claims 1 to 5.

8. A computing device, characterized in that, include: A memory, one or more processors, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs including instructions for performing the method as claimed in any one of claims 1 to 5.

Images