An active wake yaw control method and device considering fatigue damage of wind turbine units

Abstract

This invention relates to the field of wind turbine technology and discloses an active wake yaw control method and device that considers wind turbine fatigue damage. The invention constructs a wake yaw optimization model by combining wind conditions, turbine coordinates, and operating parameters with an engineering wake model to ensure that the wake assessment closely matches actual operating conditions. Then, by calculating structural loads and target damage equivalent loads, the fatigue damage of the turbine is accurately quantified, and the yaw limit angle is determined accordingly to avoid excessive yaw exacerbating damage during control. A multi-objective yaw optimization method balancing power generation and fatigue damage is used to iteratively solve for the farm-level yaw control matrix, avoiding the use of old strategies after turbine fatigue has worsened, which would lead to further damage. Ultimately, this ensures the long-term safe and efficient operation of the wind farm, achieving the dual goals of long-term safe operation and efficient power generation, and effectively meeting the dual needs of wind farms for operational safety and economy throughout their entire life cycle.

Description

Technical Field

[0001] This invention relates to the field of wind turbine technology, specifically to an active wake yaw control method and device that takes into account fatigue damage of wind turbines. Background Technology

[0002] During wind farm operation, the wake effect significantly impacts the power generation efficiency of downstream turbines. Currently, its impact is primarily mitigated through two main methods: micro-site optimization and active wake control during operation. Active wake control during operation, as an effective method, can be further divided into power control methods based on altering the axial induction factor, and wake redirection methods that change the wake direction by adjusting the turbine's orientation. However, during long-term operation, factors such as yaw action and airflow fluctuations can cause fatigue damage to critical components like the yaw mechanism and blades. If this fatigue damage accumulates and is not controlled, it will not only shorten the turbine's lifespan but may also lead to equipment failure, affecting the overall operational safety and economy of the wind farm. Therefore, active wake control must also consider the issue of turbine fatigue damage.

[0003] Therefore, existing technologies consider incorporating fatigue load factors into the control strategy. However, the above methods do not take into account the differences in the health status of different units in actual operation to carry out targeted yaw control, which makes it difficult for existing control strategies to meet the needs of long-term safe and efficient operation of wind farms. Summary of the Invention

[0004] This invention provides an active wake yaw control method and device that takes into account the fatigue damage of wind turbine units, in order to solve the problem that the existing technology does not take into account the differences in the health status of different units in actual operation for targeted yaw control, which makes it difficult for the existing control strategy to meet the needs of long-term safe and efficient operation of wind farms.

[0005] In a first aspect, the present invention provides an active wake yaw control method considering fatigue damage of wind turbine units, the method comprising:

[0006] Using wind condition data, actual coordinate information, wind turbine operating parameters, and a preset engineering wake model, a wind farm wake yaw optimization model for the wind turbine is constructed.

[0007] Calculate the structural load of the wind turbine to obtain the load time series data of the wind turbine, and calculate the target damage equivalent load data of each load time series data.

[0008] Fatigue damage assessment is performed on the equivalent load data of each target damage, and the yaw limit angle of each wind turbine unit in the fatigue damage is determined based on the assessment results.

[0009] Using the yaw limit angle as a constraint and the combined optimization of power generation and fatigue damage of the wind farm as the objective, the wake yaw optimization model of the wind farm is iteratively solved to obtain the farm-level yaw control matrix of the wind turbine for active wake control.

[0010] This invention constructs a wake yaw optimization model by combining wind conditions, unit coordinates, and operating parameters with an engineering wake model to ensure that the wake assessment closely matches actual operating conditions. Furthermore, by calculating structural loads and target damage equivalent loads, it accurately quantifies unit fatigue damage and determines the yaw limit angle accordingly, avoiding excessive yaw during control that could exacerbate damage. Finally, it employs a multi-objective yaw optimization method that balances power generation and fatigue damage to iteratively solve for the farm-level yaw control matrix, balancing power generation and safety to ensure the long-term safe and efficient operation of the wind farm, achieving the dual objectives of long-term safe operation and efficient power generation.

[0011] In one optional implementation, the step of performing fatigue damage assessment on the equivalent load data of each of the target damages, and determining the yaw limit angle of each of the units in the fatigue damage range based on the assessment results, includes:

[0012] Based on the actual installed unit control strategy and unit safety design rules, a fatigue damage assessment model is constructed.

[0013] The target damage equivalent load data is input into the fatigue damage assessment model to perform fatigue damage assessment and generate assessment results.

[0014] Extract the yaw limit angle that takes into account the fatigue damage of the unit from the evaluation results, and input the yaw limit angle into the unit-level yaw damage analysis dataset for storage.

[0015] The yaw limit angle extracted from the evaluation results in this invention is a refined control of the yaw range based on fatigue damage analysis. It can effectively avoid excessive yaw angles that aggravate unit damage under specific wind conditions. At the same time, the angle is stored in the field-level health status dataset, providing a precise constraint basis for the subsequent wind farm wake yaw optimization model, and helping to achieve yaw control that balances power generation efficiency and unit safety.

[0016] In one optional implementation, the wind farm wake yaw optimization model is iteratively solved with the yaw limit angle as a constraint and the combined optimization of power generation and fatigue damage of the wind farm group as the objective, to obtain the wind turbine-level yaw control matrix for active wake control, including:

[0017] Based on the yaw limit angle, a first constraint condition is constructed;

[0018] Based on the pre-set unit-level yaw damage analysis dataset, a second constraint condition is constructed according to the maximum value of the target damage equivalent load of a single unit during the optimization process.

[0019] Based on the first and second constraints, the wind farm wake yaw optimization model is iteratively solved with the optimization objective of optimizing the combined power generation and fatigue damage of the wind farm cluster.

[0020] Based on the solution results, the field-level yaw control matrix for active wake control of the wind turbine is determined.

[0021] This invention iteratively solves the wake yaw optimization model under dual constraints with the objectives of maximizing wind farm power generation and minimizing damage loss. It breaks through the limitations of traditional single-objective models and ensures safety through constraints. The final determined farm-level yaw control matrix can achieve synergistic optimization of wind farm power generation efficiency and unit fatigue safety, providing a scientific and feasible execution scheme for active wake control.

[0022] In one alternative implementation, it further includes:

[0023] Initial operating condition data of key components of the wind turbine are collected, and the initial operating condition data is cleaned and preprocessed to generate target operating condition data.

[0024] The target operating condition data is divided into a training set and a test set;

[0025] The target working condition data corresponding to the training set is input into the preset initial working condition-load association model for training, and an updated working condition-load association model is generated.

[0026] The updated working condition-load association model is tested using the target working condition data corresponding to the test set to generate the target working condition-load association model;

[0027] Input the target working condition data of each of the key components into the target working condition-load correlation model, and output the dynamic load time series of each of the key components;

[0028] The load spectrum is obtained by counting the number of cycles corresponding to the dynamic load time series.

[0029] Based on the equivalent damage principle and the load spectrum, the equivalent damage life of each of the key components is calculated.

[0030] Based on the damage equivalent life, stress-life curve and linear damage accumulation principle of the key components, the remaining life prediction information of the key components is determined and input into the field-level health status operation assessment dataset.

[0031] When the predicted value of the remaining lifetime prediction information is less than the preset lifetime threshold, the field-level yaw control matrix is ​​updated.

[0032] This invention calculates the remaining life based on load spectrum and equivalent damage principle, and can monitor the health status of components in real time. When the predicted life value is lower than the threshold, the field-level yaw control matrix is ​​updated, which can avoid the damage deterioration of components caused by fatigue in a timely manner.

[0033] In one optional implementation, updating the field-level yaw control matrix when the predicted value of the remaining lifetime prediction information is less than a preset lifetime threshold includes:

[0034] The predicted value of the remaining lifespan prediction information is compared with a preset lifespan threshold.

[0035] When the predicted value of the remaining lifetime prediction information is greater than or equal to the preset lifetime threshold, the current field-level yaw control matrix is ​​maintained.

[0036] When the predicted value of the remaining life prediction information is less than the preset life threshold, the unit of the key component corresponding to the predicted value is marked.

[0037] The yaw angle limit of the marked unit is upgraded to the corresponding restriction level, and the first constraint, the second constraint and the optimization objective are updated. Then, the step of iteratively solving the yaw optimization model of the wind farm wake is executed until the target farm-level yaw control matrix is ​​obtained.

[0038] This invention identifies wind turbine units when critical components have insufficient lifespan, and strengthens protection for fatigued units by upgrading yaw limit levels, updating dual constraints, and optimizing objectives to prevent further damage. The iterative solution of the optimization model generates a target control matrix adapted to the current unit health status, dynamically adjusting and balancing overall power generation efficiency with the safety of high-risk units to ensure the long-term safe and efficient operation of the wind farm.

[0039] Secondly, the present invention provides an active wake yaw control device that takes into account fatigue damage of wind turbine units, the device comprising:

[0040] The construction module is used to construct the wind farm wake yaw optimization model of the wind turbine by using wind condition data, actual coordinate information, wind turbine operating parameters and preset engineering wake model.

[0041] The calculation module is used to calculate the structural load of the wind turbine, obtain the load time series data of the wind turbine, and calculate the target damage equivalent load data of each load time series data.

[0042] The evaluation module is used to perform fatigue damage assessment on the equivalent load data of each target damage, and determine the yaw limit angle of each wind turbine unit in the fatigue damage based on the evaluation results.

[0043] The solution module is used to iteratively solve the wind farm wake yaw optimization model with the yaw limit angle as a constraint and the combined optimization of power generation and fatigue damage of the wind farm group as the objective, so as to obtain the wind turbine's farm-level yaw control matrix for active wake control.

[0044] In one optional implementation, the evaluation module includes:

[0045] Based on the actual installed unit control strategy and unit safety design rules, a fatigue damage assessment model is constructed.

[0046] The evaluation unit is used to input the target damage equivalent load data into the fatigue damage evaluation model to perform fatigue damage evaluation and generate evaluation results.

[0047] The storage unit is used to extract the yaw limit angle considering the fatigue damage of the unit from the evaluation results, and input the yaw limit angle into the unit-level yaw damage analysis dataset for storage.

[0048] Thirdly, the present invention provides an electronic device, comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the active wake yaw control method considering wind turbine fatigue damage described in the first aspect or any corresponding embodiment.

[0049] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to execute the active wake yaw control method considering wind turbine fatigue damage described in the first aspect or any corresponding embodiment.

[0050] Fifthly, the present invention provides a computer program product, including computer instructions for causing a computer to execute the active wake yaw control method considering wind turbine fatigue damage as described in the first aspect or any corresponding embodiment. Attached Figure Description

[0051] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0052] Figure 1 This is a schematic flowchart of the first active wake yaw control method considering wind turbine fatigue damage according to an embodiment of the present invention.

[0053] Figure 2 This is a schematic diagram of the second process of the active wake yaw control method considering wind turbine fatigue damage according to an embodiment of the present invention.

[0054] Figure 3 This is a schematic diagram illustrating the simulation and database establishment performed by the simulation software according to an embodiment of the present invention;

[0055] Figure 4 This is a schematic diagram of the third process of the active wake yaw control method considering wind turbine fatigue damage according to an embodiment of the present invention.

[0056] Figure 5 This is a structural block diagram of an active wake yaw control device for wind turbine generator considering fatigue damage according to an embodiment of the present invention.

[0057] Figure 6 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0058] 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 embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0059] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0060] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0061] This invention provides an active wake yaw control method that considers wind turbine fatigue damage. It constructs a wake yaw optimization model by combining wind conditions, turbine coordinates, and operating parameters to ensure the model closely matches actual operating conditions. Then, by calculating structural loads and target damage equivalent loads, it accurately quantifies turbine fatigue damage and determines the yaw limit angle to prevent excessive yaw from exacerbating damage at the source. Subsequently, iteratively solving the farm-level yaw control matrix using this angle as a constraint balances power generation and safety. Furthermore, it monitors the remaining lifespan of components in real time and updates the matrix when a threshold is triggered, preventing the use of old strategies after turbine fatigue has worsened, thus ensuring the long-term safe and efficient operation of the wind farm.

[0062] This embodiment provides an active wake yaw control method that takes into account the fatigue damage of wind turbine units. Figure 1 This is a flowchart of an active wake yaw control method considering wind turbine fatigue damage according to an embodiment of the present invention, as shown below. Figure 1 As shown, the process includes the following steps:

[0063] Step S101: Using wind condition data, actual coordinate information, wind turbine operating parameters, and a preset engineering wake model, construct a wind farm wake yaw optimization model for the wind turbine.

[0064] It should be noted that wind condition data refers to a set of parameters describing the airflow state within a wind farm, mainly including basic wind condition parameters such as wind speed, wind direction, turbulence intensity, and wind shear index.

[0065] Actual coordinate information refers to the specific location coordinates of each wind turbine in the geographical space within the wind farm, which is used to determine the relative positional relationship between the turbines.

[0066] Wind turbine operating parameters refer to parameters that reflect the operating status and performance of wind turbines, including but not limited to thrust coefficient, power curve, yaw angle, etc.

[0067] The pre-defined engineering wake model refers to a mathematical model established through theoretical derivation, experimental data fitting, or numerical simulation to meet the engineering needs of actual wind farm design and operation optimization.

[0068] The wake yaw optimization model of a wind farm refers to a mathematical model used to optimize the yaw angle of each unit in a wind farm.

[0069] In this embodiment of the invention, a wind farm wake yaw optimization model for wind turbines is constructed by using wind condition data, actual coordinate information, wind turbine operating parameters, and engineering wake models. This model first clarifies the spatial relationship between each turbine based on actual coordinate information. Based on this, it combines the distribution characteristics of wind speed and direction in the wind condition data, as well as key indicators related to wake effects in the wind turbine operating parameters, such as thrust coefficient and power curve, to establish the wake influence correlation logic between upstream turbines and downstream turbines. Then, by integrating these data, correlation logic, and engineering wake models, a calculation framework is formed that can quantify the wake loss and power generation benefits of the entire field under different yaw angles.

[0070] Specifically, see Figure 2 As shown, the input information collected includes basic wind condition parameters such as wind direction, wind speed, turbulence intensity, and wind shear index; basic wind turbine parameters used for wake simulation calculations (including rotor diameter, hub height, thrust curve, power curve, etc.); and wind turbine operating parameters such as the actual installation location coordinates of the wind turbine.

[0071] By using the actual coordinates and operating parameters of each wind turbine, the upstream and downstream positions of the turbines under different windward directions are determined, and the impact of the wake on the downstream turbines on the wake of the upstream turbines is assessed, yielding the wake impact values. The actual coordinates and operating parameters are then input into the wake velocity loss model (i.e., the engineering wake model), and the wake velocity loss model, which considers wake deflection, is used to calculate the wake velocity loss of the downstream turbines. i All upstream units j To each i Wake loss .

[0072] The wake effect value reflects the downstream unit i Affected by upstream units j The basic impact range and trend of wake blast, and the wake loss value is for a single upstream unit. j For downstream units i The specific wind speed loss quantification data is then input into a preset wake superposition model (such as a linear superposition model or a sum of squares superposition model). The model will first filter out the wake impact values ​​that cover the downstream units. i All upstream units j And then these upstream units j The corresponding wake loss value is calculated according to the superposition rule, and the downstream unit is obtained by using the wake superposition model. i Wake velocity loss due to the wake effect of all upstream units .

[0073] Taking the two-dimensional BP model as an example, this model divides the wake deflection distance into two regions: linear deflection in the near-wake region and asymptotic deflection in the far-wake region. The position length and trajectory angle in the near-wake region are evaluated by vortex surface theory, and the formula is as follows:

[0074]

[0075]

[0076] In the formula, Environmental turbulence; , These are all empirical parameters; The wake characteristic coefficient; The angle of inclination of the wake.

[0077] when At that time, wake offset for:

[0078]

[0079] when At that time, wake offset for:

[0080]

[0081] In the formula, The position length of the near-wake region; The trajectory angle of the wake; The wake characteristic coefficient; The angle of inclination of the wake; , The characteristic width of the wake in the horizontal and vertical directions; D The characteristic size of the wake generation source.

[0082] For wind turbines i With any upstream wind turbine j The BP wake yaw model considering yaw in the windward direction is expressed as follows:

[0083]

[0084]

[0085] In the formula, The downstream wind speed is calculated by superimposing the wake. For incoming air velocity; The wake characteristic coefficient; The angle of inclination of the wake; This is the wake offset. Given the hub height, the calculation typically only addresses the wind speed at that hub height, where z = ; x Indicates the distance between upstream and downstream units along the wind direction; The position length of the near-wake region; y Indicates the distance perpendicular to the wind direction; , The characteristic width of the wake in the horizontal and vertical directions; , The coefficients of wake expansion are the horizontal and vertical directions. ; D The characteristic size of the wake generation source.

[0086] Finally, the superimposed wake loss is calculated using the sum of squares formula, which is:

[0087]

[0088] In the formula, The downstream wind speed is calculated by superimposing the wake. For incoming air velocity; N The number of units that generate wake turbulence; For the first j Wind speed affected by the wake of the turbine unit; For the first j Incoming air velocity of the unit.

[0089] Considering velocity loss, wake deflection, and wake superposition, a method for calculating wind farm wake loss with a certain degree of fidelity is established for evaluating downstream wind turbines. i Affected by upstream units j The velocity loss due to the wake superposition effect was investigated, and a preliminary single-objective optimization model for wind farm wake yaw was formed under the optimization framework of an improved genetic algorithm. The single-objective optimization model for wind farm wake yaw is optimized by taking into account the yaw angles of each turbine under different wind conditions. The initial yaw limit range needs to be set according to the turbine control strategy, generally set to a relative yaw angle (-30°, 30°). The optimization objective is to maximize the output power. The optimization algorithm is based on the Genetic Algorithm (GA). This algorithm was chosen because GA is the basic framework of the subsequently selected multi-objective algorithm NSGA-II (Non-dominated Sorting Genetic Algorithm II), which is adaptable and facilitates the development of multi-objective optimization programs. The output is the optimal yaw angle and calculated output power of each turbine under different wind conditions (for example, under the conditions of inflow wind direction = 270° and inflow wind speed = 6m / st, the total output power of the three turbines is maximized after the computer yaws 20° for turbine 1, 10° for turbine 2, and no yaw for turbine 3).

[0090] Step S102: Calculate the structural load of the wind turbine, obtain the load time series data of the wind turbine, and calculate the target damage equivalent load data of each load time series data.

[0091] It should be noted that structural load refers to the collective term for the forces, moments, bending moments, and other mechanical loads borne by the key components (such as blade roots, towers, yaw bearings, etc.) of a wind turbine during operation. These loads are mainly generated by factors such as wind load, the turbine's own weight, and operating inertia.

[0092] Load time series data refers to a continuous data sequence that records the changes in the structural load of a wind turbine over time, with time as the dimension. The data includes the magnitude and fluctuation pattern of the load at different times.

[0093] Target damage equivalent load data refers to load data that can quantify the equivalent fatigue damage level of structural components after fatigue damage analysis and data preprocessing of load time series data.

[0094] In this embodiment of the invention, based on wind condition data and actual operating parameters of the wind turbine (such as load simulation data), the loads such as forces and bending moments on key components of the turbine (such as blade roots, towers, and yaw mechanisms) under different operating conditions are simulated using flow field simulation software to generate load time series data that varies over time. Load cycle characteristic parameters are extracted from the load time series data, and the influence of static loads is eliminated by combining the stress-life curves of the corresponding components of the turbine with the mean correction model to obtain the equivalent alternating stress amplitude. Then, the total fatigue damage is calculated according to the linear damage accumulation principle. Finally, the total fatigue damage is equivalently converted into a constant load amplitude. After preprocessing such as outlier removal and data interpolation, the target damage equivalent load data is obtained.

[0095] Specifically, data such as aerodynamic data, elastic dynamics data, hydrodynamic data, servo system data, and structural mechanics data of wind turbine units are collected for load simulation calculations.

[0096] See Figure 3 As shown, the flow field simulation calculation was performed using the simulation software FAST.Farm. A simplified wake sway model (DWM) embedded in the software was used to save computational resources. Aerodynamic data, elastic dynamics data, hydrodynamic data, servo system data, and structural mechanics data of the wind turbine were all input into the dynamic wake sway model.

[0097] The aerodynamic data of wind turbines, including blade lift and drag distribution and airflow pressure, are used as the initial disturbance source for the wake. Combined with the vibration of the blades and tower reflecting the disturbance of the airflow, the data is combined with the data of elastic dynamics to reflect the disturbance of the airflow by the vibration of the blades and tower. For offshore units, the influence of waves and ocean currents on the flow field around the foundation from the hydrodynamic data is also incorporated. At the same time, the dynamic changes of the flow field caused by yaw and pitch actions are simulated through servo system data. Then, the structural mechanics data provides the parameters of the obstruction characteristics of the unit structure on the airflow. Based on these inputs, the model simulates the generation, diffusion, oscillation and interaction of the wake with the surrounding airflow. Finally, the output includes the flow field environment data of each unit, including parameters such as wind speed vector, turbulence intensity, and pressure distribution.

[0098] Before inputting the load simulation data into the dynamic wake meandering model for wake simulation, different wind-yaw coordinated working condition combinations with wind speed, wind direction, and yaw angle are pre-set for the input conditions.

[0099] Each wind condition-yaw coordinated operating condition combination calls upon the corresponding wind speed, turbulence intensity, and other parameters in the flow field environment. Combined with the actual coordinates of the unit and the yaw angle, the direction of the airflow and the wind turbine is determined. Then, the flow field parameters are converted into lift, drag, and torque on the blades through the aerodynamic load calculation module. At the same time, the component vibration load and the additional load generated by the servo system action in the elastic dynamics data are superimposed. For offshore units, hydrodynamic loads also need to be taken into account. After integration, the instantaneous load values ​​of the key components of the unit under each operating condition are obtained. The load data at different times are continuously recorded in time series to obtain the load time series data of each unit.

[0100] The load time series data of each unit were statistically analyzed using the rainflow counting method, and the load cycle characteristic parameters were extracted. The influence of static load in the load cycle characteristic parameters was eliminated by the mean correction model to obtain the equivalent alternating load amplitude.

[0101] Retrieve the material properties and stress-life curves (i.e., SN curves) of key components (such as yaw bearings and blade root connection structures), then convert the equivalent alternating load amplitude into the corresponding alternating stress amplitude, and find the number of cycles corresponding to the stress amplitude on the stress-life curve. This number of cycles is the fatigue life of the key component under the alternating load cycle.

[0102] The number of cycles already endured for each alternating load cycle and the fatigue life of the key components under that cycle are determined. Then, according to the principle of linear damage accumulation, the ratio of the number of cycles already endured to the fatigue life of each cycle is taken as the fatigue damage generated in that cycle. Subsequently, the fatigue damage values ​​of all alternating load cycles are accumulated, and the sum is the total fatigue damage of the key components under the current load sequence.

[0103] Based on the reference cycle number corresponding to the design life of the wind turbine, the total fatigue damage is equivalently converted into a constant load amplitude, and finally the initial damage equivalent load (DEL_simulation) is obtained to quantify the fatigue damage level of the load equivalent.

[0104] After processing the initial damage equivalent load data, the resulting target damage equivalent load data is stored in the unit-level yaw damage analysis dataset, which can be represented as a multi-dimensional array matrix of [wind speed, wind direction, turbulence intensity, yaw angle, DEL_simulation value].

[0105] Step S103: Perform fatigue damage assessment on the equivalent load data of each target damage, and determine the yaw limit angle of each wind turbine unit in the fatigue damage based on the assessment results.

[0106] It should be noted that fatigue damage assessment refers to the process of analyzing the current degree of fatigue damage of components and determining whether the damage exceeds the safe range based on the damage equivalent load data, material fatigue characteristics (such as stress-life curves), and safety design requirements of key components of each unit, using a specific assessment model.

[0107] The assessment results refer to the results obtained through fatigue damage assessment of the key components of each unit.

[0108] Yaw limit angle refers to the allowable range of yaw angles set based on fatigue damage assessment results in active wake control of wind turbine units to prevent the key components of the unit from suffering increased fatigue damage due to yaw action.

[0109] In this embodiment of the invention, a fatigue damage assessment model is constructed by combining the actual installed unit's control strategy (such as the original allowable yaw angle range) and the unit's safety design rules (such as the fatigue damage limit corresponding to the component's design life). The target damage equivalent load data obtained in step S102 is then input into the model. By comparing the damage equivalent load under different yaw angles with the preset safety damage threshold, different levels of fatigue risk intervals are divided. Finally, the yaw limit angle corresponding to each unit is determined based on the risk interval.

[0110] Step S104: With the yaw limit angle as a constraint and the combined optimization of power generation and fatigue damage of the wind farm as the objective, the wind farm wake yaw optimization model is iteratively solved to obtain the wind turbine-level yaw control matrix for active wake control.

[0111] It should be noted that the constraints refer to the yaw angle limits set during the solution of the wind farm wake yaw optimization model to ensure unit safety and avoid aggravated fatigue damage.

[0112] The power generation of a wind farm cluster refers to the sum of the total output power of all units within the wind farm.

[0113] The optimal overall fatigue damage refers to the state in which, in the optimization of wind farm wake yaw, the cumulative fatigue damage of all key components of wind turbines is kept within a reasonable range by adjusting the yaw angle of the turbines, while achieving a global optimal balance with the power generation of the wind farm cluster.

[0114] Iterative solution refers to the cyclical process of setting initial parameters, calculating objective function values, adjusting variables (such as yaw angle), and verifying constraint satisfaction for complex problems such as wind farm wake yaw optimization models.

[0115] The wind farm-level yaw control matrix refers to a structured parameter table used for overall active wake control of a wind farm, which includes key dimensions such as "unit number, wind speed, wind direction, turbulence intensity, and yaw angle".

[0116] In this embodiment of the invention, the yaw limit angle of each unit is used as the core constraint to clarify the boundary value of the yaw angle of each unit during the optimization process. Combined with the original wind condition data, unit coordinates and operating parameters in the wind farm wake yaw optimization model, the dual optimization objectives of "maximum power generation of the wind farm group" and "minimum fatigue damage of the wind farm group" are set. An algorithm adapted to multi-objective optimization is used to iteratively solve the model. In each iteration, it is verified whether the yaw angle of each unit meets the corresponding constraint requirements. If the constraint is exceeded, the parameters are adjusted and recalculated until the iteration result satisfies the balance of the dual objectives and there is no constraint violation. Finally, a field-level yaw control matrix with "unit number, wind speed, wind direction and yaw angle" as the core dimensions is output, providing a specific execution scheme for the active wake control of the wind farm.

[0117] This embodiment provides an active wake yaw control method that takes into account the fatigue damage of wind turbine units. Figure 4 This is a flowchart of an active wake yaw control method considering wind turbine fatigue damage according to an embodiment of the present invention, as shown below. Figure 4 As shown, the process includes the following steps:

[0118] Step S201: Using wind condition data, actual coordinate information, wind turbine operating parameters, and a preset engineering wake model, construct a wind farm wake yaw optimization model for the wind turbine.

[0119] Please see details Figure 1 Step S101 of the illustrated embodiment will not be described again here.

[0120] Step S202: Calculate the structural load of the wind turbine, obtain the load time series data of the wind turbine, and calculate the target damage equivalent load data of each load time series data.

[0121] Please see details Figure 1 Step S102 of the illustrated embodiment will not be described again here.

[0122] Step S203: Perform fatigue damage assessment on the equivalent load data of each target damage, and determine the yaw limit angle of each wind turbine unit in the fatigue damage based on the assessment results.

[0123] Specifically, step S203 includes:

[0124] Step S2031: Based on the actual installed unit control strategy and unit safety design rules, construct a fatigue damage assessment model.

[0125] It should be noted that the actual installed unit control strategy refers to the operating control logic preset at the factory or determined after on-site commissioning of the wind turbine, including the triggering conditions, execution range and adjustment rules of actions such as yaw, pitch, start and stop.

[0126] Unit safety design rules refer to the safe operation guidelines for units formulated based on industry standards, design specifications and manufacturer requirements, covering core indicators such as the design life of key components, fatigue damage limits, and structural strength limits.

[0127] The fatigue damage assessment model refers to a mathematical model that integrates unit control strategy constraints and safety design rules to quantitatively analyze the relationship between damage equivalent load and fatigue risk.

[0128] In this embodiment of the invention, a fatigue damage assessment method is established by combining the control strategy of the actual installed unit with the unit's safety design requirements. Specifically, it is necessary to first identify the core parameters of the yaw action in the actual installed unit's control strategy, and clarify the boundary values ​​and action constraints of the yaw angle in the model; then, by combining the design life of key components, the upper limit of fatigue damage, and the safety redundancy coefficient in the unit's safety design rules, the damage assessment standards and risk level classification basis of the model are determined; subsequently, the above constraints and standards are integrated to establish a quantitative mapping relationship of "damage equivalent load - yaw angle - fatigue risk level", forming a fatigue damage assessment model.

[0129] Step S2032: Input the target damage equivalent load data into the fatigue damage assessment model to perform fatigue damage assessment and generate assessment results.

[0130] In this embodiment of the invention, the damage equivalent load value and its corresponding wind condition-yaw condition parameters contained in the target damage equivalent load data are input into the fatigue damage assessment model in the format required by the model. The model will call the built-in "damage equivalent load yaw angle fatigue risk level" mapping relationship, combine the yaw constraint in the actual installed unit control strategy and the damage threshold in the unit safety design rules, calculate the total fatigue damage ratio corresponding to the load data, match the corresponding risk level, and output the yaw limit angle for the risk level to form the assessment result.

[0131] Step S2033: Extract the yaw limit angle considering unit fatigue damage from the evaluation results, and input the yaw limit angle into the unit-level yaw damage analysis dataset for storage.

[0132] It should be noted that the crew-level yaw damage analysis dataset refers to a collection of fatigue damage data related to yaw actions, constructed at the level of a single crew unit.

[0133] In this embodiment of the invention, the yaw limit angles corresponding to the fatigue risk levels are selected from the evaluation results, and these yaw limit angles are imported into the field-level health status operation evaluation dataset to complete the data classification and storage. Specifically, within the original yaw constraints (referring to the allowable yaw angle range of the actual installed unit's control strategy), yaw limit angles are divided into levels I / II / III based on DEL data. The average DEL value corresponding to the rated wind speed (e.g., 12 m / s) + the original yaw limit angle (±30°) is used as the baseline DEL0. The allowable DEL threshold range for each level is set as follows: Level I: Maximum allowable 1.1DEL0; Level II: 0.9DEL0; Level III: 0.8DEL0. Fitting curves of DEL and yaw angle under different wind speeds are then established. Based on the set threshold ranges for different levels, the limit yaw angles for the corresponding levels are deduced to form the yaw limit angles.

[0134] It is worth mentioning that the yaw limit angle is a further limitation based on fatigue analysis within the original allowable relative yaw angle range of the unit's control strategy. It is divided into three levels: Level I (Yaw Open Level), Level II (Yaw Protected Level), and Level III (Yaw Limit Protected Level). Level I allows the largest range of relative yaw angles, while Level III allows the smallest. Setting different yaw range levels serves two purposes: firstly, to avoid excessively large relative yaw angles during wake control under specific wind conditions, which could significantly increase fatigue damage to the unit; and secondly, to reduce the risk of fatigue damage to the yaw mechanism or other components of units with poor operational health by lowering the allowable yaw range.

[0135] Step S204: With the yaw limit angle as a constraint and the goal of optimizing the combined power generation and fatigue damage of the wind farm cluster, the wind farm wake yaw optimization model is iteratively solved to obtain the wind turbine-level yaw control matrix for active wake control.

[0136] Specifically, step S204 includes:

[0137] Step S2041: Based on the yaw limit angle, construct the first constraint condition.

[0138] It should be noted that the first constraint condition refers to the core condition used to constrain the yaw action of the wind turbine, with the yaw limit angle as the optimization variable.

[0139] In this embodiment of the invention, the yaw angle is used as the optimization variable and the field-level health status operation assessment dataset—yaw limit angle data—is used as the optimization variable constraint to form the first constraint condition.

[0140] Step S2042: Based on the preset unit-level yaw damage analysis dataset, construct the second constraint condition according to the maximum value of the target damage equivalent load of a single unit in the optimization process.

[0141] It should be noted that the second constraint is a constraint constructed based on the maximum value of the target damage equivalent load in the unit-level yaw damage analysis dataset for a single unit.

[0142] The maximum value of the target damage equivalent load for a single unit is the maximum value of the target damage equivalent load for a single unit under all past wind and yaw conditions, extracted from the unit-level yaw damage analysis dataset.

[0143] In this embodiment of the invention, a second constraint condition is formed by applying a target constraint to the maximum DEL value of a single unit during the optimization control process based on the unit-level yaw damage analysis dataset.

[0144] Step S2043: Based on the first and second constraints, and with the optimization objective of optimizing the combined power generation and fatigue damage of the wind farm cluster, the wake yaw optimization model of the wind farm is solved iteratively.

[0145] In this embodiment of the invention, the first and second constraints are embedded as boundary conditions for model solving into the wind farm wake yaw optimization model. Then, dual optimization objectives are set, namely, maximizing the power generation of the wind farm group and minimizing the loss of the equivalent load of wind farm damage. Subsequently, the NSGA-II intelligent optimization algorithm is used to continuously adjust the yaw angle parameters of each unit within the constraints. After each iteration, the corresponding power generation and damage load loss values ​​are calculated. The parameter combination is dynamically optimized by comparing the iteration results until the difference between the two iteration results is less than a preset threshold. Finally, the wind farm-level yaw control matrix for active wake control is output. The matrix form is set as [unit number, wind speed, wind direction, turbulence intensity, yaw angle].

[0146] Specifically, the NSGA-II intelligent optimization algorithm is adapted to multi-objective optimization with conflicting objectives such as "power-fatigue damage". Genetic algorithms (GA) are preferentially used in the single-objective stage, while NSGA-II is extended to the multi-objective stage. The two have highly consistent structures, low development costs, and are well-suited to the nonlinear, multivariable, and conflicting objective optimization requirements of wake yaw control. Its optimization process structure is as follows: initialization - objective function calculation (maximum output power for wake calculation + minimum DEL for database matching) - fast non-dominated sorting - congestion calculation - selection and retention - crossover / mutation - iteration termination.

[0147] Step S2044: Determine the field-level yaw control matrix for active wake control of the wind turbine based on the solution results.

[0148] It should be noted that the solution result refers to the output data obtained after iteratively solving the wind farm wake yaw optimization model, that is, the output field-level yaw control matrix used for active wake control, and the matrix form is set as [unit number, wind speed, wind direction, turbulence intensity, yaw angle].

[0149] In this embodiment of the invention, the optimal yaw angle of each unit under different wind conditions (wind speed, wind direction) is extracted from the iterative solution results. Then, according to the mapping relationship of "unit number - wind condition parameter - optimal yaw angle", the data is organized into a structured matrix form, namely [unit number, wind speed, wind direction, turbulence intensity, yaw angle].

[0150] Step S205: Collect initial operating condition data of key components of the wind turbine, and perform data cleaning and preprocessing on the initial operating condition data to generate target operating condition data.

[0151] It should be noted that key components refer to the core components that play a decisive role in the safe operation and lifespan of wind turbine units and are prone to fatigue damage during yaw control.

[0152] Initial operating condition data refers to the original operating status data of key components of wind turbine units, including parameters such as load, temperature, and vibration that reflect the real-time stress and operating status of the components.

[0153] Data cleaning and preprocessing refers to a series of processing operations performed on initial operating data.

[0154] Target operating condition data refers to standardized data that accurately reflects the operating status of key components after data cleaning and preprocessing.

[0155] In this embodiment of the invention, relevant operating condition data (i.e., initial operating condition data) of key components of each wind turbine are collected from the SCADA system. Data preprocessing work such as interpolation, filling gaps, normalization, and outlier removal is performed on the initial operating condition data to obtain the target operating condition data.

[0156] Step S206: Divide the target working condition data into a training set and a test set.

[0157] It should be noted that the training set refers to the dataset that is partitioned from the target working condition data and used for model training.

[0158] The test set refers to a dataset that is partitioned from the target operating condition data and used to evaluate the model's performance.

[0159] In this embodiment of the invention, the target working condition data is divided into a training set and a test set according to a ratio of 7:3.

[0160] Step S207: Input the target working condition data corresponding to the training set into the preset initial working condition-load association model for training, and generate the updated working condition-load association model.

[0161] It should be noted that the initial operating condition-load correlation model refers to a pre-built basic model used to describe the correlation between the operating condition parameters and loads of key components.

[0162] The updated working condition-load correlation model refers to the working condition-load correlation model obtained after training and optimization using training set data.

[0163] In this embodiment of the invention, the target working condition data in the training set is matched with the corresponding load data to form a sample pair of "working condition parameter-load value". The sample pair is then input into the preset initial working condition-load association model according to the format required by the model. The model iteratively learns the mapping relationship between working condition parameters and load, and continuously adjusts the internal parameters of the model to reduce the error between the predicted load and the actual load until the prediction accuracy of the model on the training set reaches the preset threshold, and finally the updated working condition-load association model is obtained.

[0164] Step S208: Use the target working condition data corresponding to the test set to test the updated working condition-load association model and generate the target working condition-load association model.

[0165] It should be noted that the target working condition-load correlation model refers to the final model obtained after verification and optimization using a test set, which can accurately predict the load borne by key components based on their working condition parameters within a preset accuracy range.

[0166] In this embodiment of the invention, the target working condition data in the test set is input into the updated working condition-load association model, and the model outputs the corresponding predicted load value. The predicted load value is then compared with the actual load value in the test set, and the error index between the two is calculated. If the error index meets the preset accuracy requirements, the updated working condition-load association model is directly determined as the target working condition-load association model. If the error does not meet the requirements, the model parameters are fine-tuned based on the error feedback in the test set, and the model is verified again with the test set data until the model prediction accuracy meets the requirements, and finally the target working condition-load association model is generated.

[0167] Step S209: Input the target working condition data of each key component into the target working condition-load correlation model, and output the dynamic load time series of each key component.

[0168] It should be noted that dynamic load time series refers to a continuous set of data arranged in chronological order, showing the loads on key components at different times.

[0169] In this embodiment of the invention, target operating condition data of key components of each unit are acquired in real time and input into the target operating condition-load association model. The model will calculate and output the load change data of the component in the corresponding time dimension in real time according to the preset operating condition and load mapping relationship. After these data are arranged in time order, a dynamic load time series is formed.

[0170] It is worth noting that SCADA data lacks direct load values ​​(such as blade flapping moment and gearbox input shaft torque). Therefore, a mapping model between operating condition parameters and dynamic load time series needs to be established. A mechanistic model + data-driven approach is employed, establishing correlation models based on aerodynamic-mechanical principles for key components (such as blade roots and towers). Machine learning methods are used to improve model accuracy with a training set and to verify model accuracy with a test set, converting the dynamic load data. If measured load data (such as strain gauges or load testing systems) is available for the wind farm, it is prioritized for training.

[0171] Step S210: Count the number of cycles corresponding to the dynamic load time series to obtain the load spectrum.

[0172] It should be noted that the number of cycles refers to the total number of times a load cycle of a specific amplitude occurs in a dynamic load time series. A load cycle refers to the complete process of the load changing from one value to another extreme value and then returning to the initial value.

[0173] Load spectrum refers to a graph that describes the cyclic characteristics of loads on key components.

[0174] In this embodiment of the invention, the rainflow counting method is used to count the number of cycles corresponding to different load amplitudes in order to construct a load spectrum.

[0175] Step S211: Based on the equivalent damage principle and load spectrum, calculate the equivalent damage life of each key component.

[0176] It should be noted that the equivalent damage principle means that fatigue damage caused by loads of different amplitudes and different cycles applied to components can be equivalently converted into damage caused by the corresponding number of cycles under a certain reference load.

[0177] Damage equivalent life (DEL_operation) refers to the life value corresponding to the application of a reference load when the fatigue damage of a key component under the actual load spectrum is equivalently converted based on the principle of equivalent damage.

[0178] In this embodiment of the invention, after constructing the load spectrum, the damage equivalent life DEL_operation of the key components is calculated based on the equivalent damage principle.

[0179] Step S212: Based on the damage equivalent life, stress-life curve and linear damage accumulation principle of the key components, determine the remaining life prediction information of the key components and input it into the field-level health status operation assessment dataset.

[0180] It should be noted that the principle of linear damage accumulation means that under multiple load cycles, the total fatigue damage of a component is equal to the sum of the damage generated in each load cycle. When the total damage reaches 1, the component will fail due to fatigue.

[0181] Remaining life prediction information refers to the time or number of cycles that a component can continue to operate normally from the current moment until fatigue failure, calculated by a specific prediction model based on real-time operating data, load time series data, stress-life curves, and the principle of linear damage accumulation of key components.

[0182] The field-level health status operation assessment dataset refers to a structured collection that covers all wind farm units and integrates health status-related data, including yaw limit angle and remaining life prediction of key components.

[0183] In this embodiment of the invention, the damage equivalent life of the key components is combined with the component stress-life curve (SN curve) and Miner's rule to accumulate the damage amount of the running period, and compare it with the total allowable damage of the design to obtain the remaining life, thus obtaining the field-level health status operation assessment dataset - the prediction of the remaining life of key components.

[0184] Specifically, the total fatigue life of key components under reference stress is obtained through stress-life curves. Then, combining this with the linear damage accumulation principle, the remaining life is calculated based on the calculated unit DEL_operation, compared with the total allowable damage and unit operating years in international standards or unit design schemes. This yields the final remaining life prediction information for key components. The specific formula is:

[0185]

[0186] In the formula, Total permissible damage; This is the cumulative damage amount for the segments that have been run. This refers to the number of years the equipment has been in operation.

[0187] Furthermore, the remaining life prediction information of key components is imported into the wind farm yaw status database for storage. The database consists of three parts: first, a unit-level yaw damage analysis dataset; second, a farm-level health status operation assessment dataset, which includes three levels of yaw limit angles and the remaining life of key components, and is periodically updated, calculated, and stored in the database; and third, a farm-level yaw control matrix.

[0188] Step S213: When the predicted value of the remaining lifetime prediction information is less than the preset lifetime threshold, the field-level yaw control matrix is ​​updated.

[0189] In some optional implementations, step S213 above includes:

[0190] Step S2131: Compare the predicted value of the remaining lifespan prediction information with the preset lifespan threshold.

[0191] It should be noted that the predicted value refers to the core value extracted from the remaining life prediction information, which quantifies the remaining safe operating time of key components.

[0192] The preset lifespan threshold refers to the critical lifespan value set in advance based on the design fatigue life of key components, industry operation and maintenance standards, equipment safety redundancy requirements, and operation and maintenance economy.

[0193] In this embodiment of the invention, the predicted value of the remaining life prediction information needs to be compared with the preset life threshold in real time to further determine whether the control strategy update start conditions for the remaining life of the unit's key components are met.

[0194] Step S2132: When the predicted value of the remaining lifetime prediction information is greater than or equal to the preset lifetime threshold, the current field-level yaw control matrix is ​​maintained.

[0195] It should be noted that the current wind farm-level yaw control matrix refers to the matrix that the wind farm is currently executing to coordinate the yaw angles of each turbine.

[0196] In this embodiment of the invention, after comparing the remaining life prediction value with the preset life threshold, if it is determined that the remaining life of the key components is still sufficient (i.e., the prediction value is not lower than the threshold), it indicates that the unit operation mode under the current field-level yaw control matrix has not caused excessive fatigue damage to the components, and there is no need to adjust the control strategy, that is, there is no need to start the control strategy for the remaining life of the key components of the unit.

[0197] Step S2133: When the predicted value of the remaining life prediction information is less than the preset life threshold, the unit corresponding to the key component of the predicted value is marked.

[0198] It should be noted that marking a wind turbine unit refers to the operation of identifying wind turbine units with critical components that have insufficient remaining lifespan (predicted value is lower than the preset lifespan threshold) in the operation and maintenance system.

[0199] In this embodiment of the invention, after comparing the remaining life prediction value with the preset life threshold, if the prediction value is lower than the threshold, it indicates that the component has entered a maintenance cycle that requires close attention, and subsequent maintenance procedures need to be triggered, and the unit with abnormal fatigue damage of key components is marked.

[0200] Step S2134: Upgrade the restriction level corresponding to the yaw angle limit of the marked unit, update the first constraint condition, the second constraint condition and the optimization objective, and jump to execute the step of iteratively solving the wind farm wake yaw optimization model until the target farm-level yaw control matrix is ​​obtained.

[0201] It should be noted that the restriction level refers to the graded control standard for the yaw angle adjustment range of wind turbine units. Different levels correspond to different angle adjustment thresholds, and the higher the level, the stricter the angle restriction.

[0202] The target field-level yaw control matrix refers to the yaw control matrix that is finally obtained after updating the constraints and optimizing the objectives and iterating again to adapt to the current unit state (including the upgrade of the marked unit angle limit).

[0203] In this embodiment of the invention, during the yaw re-optimization process, the yaw allowable angle limit level for the marked aircraft group is upgraded (e.g., upgraded from Level I (yaw open level) to Level II (yaw protected level) control), while the yaw allowable angle limit levels for the remaining aircraft groups remain unchanged during the re-optimization process. The target constraints are updated during the yaw re-optimization process, using the DEL_operation dataset of the marked aircraft group as a benchmark. During the optimization process, the maximum DEL constraint value for the marked aircraft group is replaced by DEL_operation instead of DEL_simulation. The process then jumps back to the iterative solution process in step S2043, performs yaw optimization calculations again, and outputs the updated field-level yaw control matrix, i.e., the target field-level yaw control matrix.

[0204] This application improves upon the shortcomings of existing active wake control methods that only design active wake control strategies based on the single objective of optimal wind farm power output. It establishes a multi-objective optimized yaw control strategy that considers unit fatigue damage under yaw conditions. This method can effectively reduce the increase of fatigue load while optimizing the overall output power of the wind farm.

[0205] The wind farm-level yaw state database proposed in this invention can be continuously used for updating and optimizing yaw control strategies. This avoids the need to repeatedly simulate fatigue loads when recalculating wake control strategies. The fatigue load situation under different operating conditions can be determined simply by calling the relevant data during recalculation. Therefore, this invention can save a lot of computing resources, especially for the dynamic optimization and updating of wake control strategies.

[0206] Furthermore, regular assessments of turbine service quality are conducted using wind farm operation data, and the farm-level health status assessment dataset (including yaw limit angle and remaining life predictions for key components) is updated in real time. Fatigue damage assessments and remaining life predictions are performed on key components implementing wake control (such as yaw mechanisms), ensuring that the active wake control strategy fully considers the operational status of key components. This is combined with database data for regular updates and optimizations considering health status. Fatigue damage to key components is set as a new constraint, and by optimizing yaw angles and reducing yaw frequency, wake control is prevented from directly or indirectly reducing the service life of some components.

[0207] This embodiment also provides an active wake yaw control device that takes into account the fatigue damage of wind turbine units. This device is used to implement the above embodiments and preferred embodiments, and details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0208] This embodiment provides an active wake yaw control device that takes into account the fatigue damage of wind turbine units, such as... Figure 5 As shown, it includes:

[0209] Module 301 is used to construct a wind farm wake yaw optimization model for wind turbines by using wind condition data, actual coordinate information, wind turbine operating parameters and preset engineering wake models.

[0210] The calculation module 302 is used to calculate the structural load of the wind turbine, obtain the load time series data of the wind turbine, and calculate the target damage equivalent load data of each load time series data.

[0211] The evaluation module 303 is used to perform fatigue damage assessment on the equivalent load data of each target damage, and determine the yaw limit angle of each wind turbine unit in the fatigue damage based on the evaluation results.

[0212] The solver module 304 is used to iteratively solve the wind farm wake yaw optimization model with yaw limit angle as constraint and the goal of optimizing the combined power generation and fatigue damage of the wind farm group, so as to obtain the field-level yaw control matrix of the wind turbine for active wake control.

[0213] In some alternative implementations, the evaluation module 303 includes:

[0214] The first building unit is used to construct a fatigue damage assessment model based on the actual installed unit control strategy and unit safety design rules;

[0215] The evaluation unit is used to input the target damage equivalent load data into the fatigue damage evaluation model to perform fatigue damage evaluation and generate evaluation results.

[0216] The storage unit is used to extract the yaw limit angle considering unit fatigue damage from the evaluation results and input the yaw limit angle into the unit-level yaw damage analysis dataset for storage.

[0217] In some alternative implementations, the solver module 304 includes:

[0218] The second building unit is used to construct the first constraint condition based on the yaw limit angle;

[0219] The third building unit is used to construct the second constraint condition based on the preset unit-level yaw damage analysis dataset and according to the maximum value of the target damage equivalent load of a single unit in the optimization process.

[0220] The solution unit is used to iteratively solve the wind farm wake yaw optimization model based on the first and second constraints, with the optimization objective of optimizing the combined power generation and fatigue damage of the wind farm group;

[0221] The matrix unit is used to determine the field-level yaw control matrix for active wake control of the wind turbine based on the solution results.

[0222] In some alternative implementations, it also includes:

[0223] The cleaning and preprocessing unit is used to collect initial operating condition data of key components of the wind turbine, and to perform data cleaning and preprocessing on the initial operating condition data to generate target operating condition data.

[0224] The partitioning unit is used to divide the target working condition data into training and testing sets;

[0225] The training unit is used to input the target working condition data corresponding to the training set into the preset initial working condition-load association model for training, and generate an updated working condition-load association model.

[0226] The test unit is used to test the updated working condition-load association model using the target working condition data corresponding to the test set, and to generate the target working condition-load association model.

[0227] The input model unit is used to input the target working condition data of each key component into the target working condition-load association model and output the dynamic load time series of each key component.

[0228] The statistical unit is used to count the number of cycles corresponding to the dynamic load time series to obtain the load spectrum;

[0229] The lifetime calculation unit is used to calculate the damage equivalent lifetime of each key component based on the equivalent damage principle and load spectrum.

[0230] The prediction unit is used to determine the remaining life prediction information of key components based on the damage equivalent life, stress-life curve and linear damage accumulation principle of the key components, and input it into the field-level health status operation assessment dataset.

[0231] The update unit is used to update the field-level yaw control matrix when the predicted value of the remaining lifetime prediction information is less than the preset lifetime threshold.

[0232] In some optional implementations, the updating unit includes:

[0233] The comparison subunit is used to compare the predicted value of the remaining lifetime prediction information with a preset lifetime threshold.

[0234] The holding sub-unit is used to maintain the current field-level yaw control matrix when the predicted value of the remaining lifetime prediction information is greater than or equal to the preset lifetime threshold.

[0235] The marking subunit is used to mark the key components of the unit corresponding to the predicted value when the predicted value of the remaining life prediction information is less than the preset life threshold.

[0236] The update sub-unit is used to upgrade the yaw angle limit corresponding to the marked unit, and update the first constraint, the second constraint and the optimization objective. Then, it jumps to execute the step of iteratively solving the wind farm wake yaw optimization model until the target farm-level yaw control matrix is ​​obtained.

[0237] The active wake yaw control device considering wind turbine fatigue damage provided in this embodiment of the invention can execute the active wake yaw control method considering wind turbine fatigue damage provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method. Further functional descriptions of the above modules and units are the same as in the corresponding embodiments described above, and will not be repeated here.

[0238] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0239] The following is a detailed reference. Figure 6 This diagram illustrates a structural schematic suitable for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 401, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 402 or a program loaded from memory 408 into random access memory (RAM) 403. The RAM 403 also stores various programs and data required for the operation of the electronic device. The processor 401, ROM 402, and RAM 403 are interconnected via a bus 404. An input / output (I / O) interface 405 is also connected to the bus 404.

[0240] Typically, the following devices can be connected to I / O interface 405: input devices 406 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 407 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 408 including, for example, magnetic tapes, hard disks, etc.; and communication devices 409. Communication device 409 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 6 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0241] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 409, or installed from a memory 408, or installed from a ROM 402. When the computer program is executed by the processor 401, it performs the functions defined in the active wake yaw control method considering wind turbine fatigue damage of the embodiments of the present invention.

[0242] Figure 6 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0243] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the active wake yaw control method considering wind turbine fatigue damage shown in the above embodiments is implemented.

[0244] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0245] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. An active wake yaw control method considering fatigue damage of wind turbine units, characterized in that, The method includes: Using wind condition data, actual coordinate information, wind turbine operating parameters, and a preset engineering wake model, a wind farm wake yaw optimization model for the wind turbine is constructed. Calculate the structural load of the wind turbine to obtain the load time series data of the wind turbine, and calculate the target damage equivalent load data of each load time series data. Fatigue damage assessment is performed on the equivalent load data of each target damage, and the yaw limit angle of each wind turbine unit in the fatigue damage is determined based on the assessment results. Using the yaw limit angle as a constraint and the combined optimization of power generation and fatigue damage of the wind farm as the objective, the wake yaw optimization model of the wind farm is iteratively solved to obtain the farm-level yaw control matrix of the wind turbine for active wake control.

2. The method according to claim 1, characterized in that, The process of performing fatigue damage assessment on the equivalent load data of each target damage, and determining the yaw limit angle for fatigue damage of each unit based on the assessment results, includes: Based on the actual installed unit control strategy and unit safety design rules, a fatigue damage assessment model is constructed. The equivalent load data of the target damage is input into the fatigue damage assessment model to perform fatigue damage assessment and generate assessment results. Extract the yaw limit angle that takes into account the fatigue damage of the unit from the evaluation results, and input the yaw limit angle into the unit-level yaw damage analysis dataset for storage.

3. The method according to claim 1, characterized in that, The wind farm wake yaw optimization model is iteratively solved using the yaw limit angle as a constraint and the combined optimization of power generation and fatigue damage of the wind farm as the objective. This yields the wind turbine-level yaw control matrix for active wake control, including: Based on the yaw limit angle, a first constraint condition is constructed; Based on the pre-set unit-level yaw damage analysis dataset, a second constraint condition is constructed according to the maximum value of the target damage equivalent load of a single unit during the optimization process. Based on the first and second constraints, the wind farm wake yaw optimization model is iteratively solved with the optimization objective of optimizing the combined power generation and fatigue damage of the wind farm cluster. Based on the solution results, the field-level yaw control matrix for active wake control of the wind turbine is determined.

4. The method according to claim 3, characterized in that, Also includes: Initial operating condition data of key components of the wind turbine are collected, and the initial operating condition data is cleaned and preprocessed to generate target operating condition data. The target operating condition data is divided into a training set and a test set; The target working condition data corresponding to the training set is input into the preset initial working condition-load association model for training, and an updated working condition-load association model is generated. The updated working condition-load association model is tested using the target working condition data corresponding to the test set to generate the target working condition-load association model; Input the target working condition data of each of the key components into the target working condition-load correlation model, and output the dynamic load time series of each of the key components; The load spectrum is obtained by counting the number of cycles corresponding to the dynamic load time series. Based on the equivalent damage principle and the load spectrum, the equivalent damage life of each of the key components is calculated. Based on the damage equivalent life, stress-life curve and linear damage accumulation principle of the key components, the remaining life prediction information of the key components is determined and input into the field-level health status operation assessment dataset. When the predicted value of the remaining lifetime prediction information is less than the preset lifetime threshold, the field-level yaw control matrix is ​​updated.

5. The method according to claim 4, characterized in that, When the predicted value of the remaining lifetime prediction information is less than a preset lifetime threshold, the field-level yaw control matrix is ​​updated, including: The predicted value of the remaining lifespan prediction information is compared with a preset lifespan threshold. When the predicted value of the remaining lifetime prediction information is greater than or equal to the preset lifetime threshold, the current field-level yaw control matrix is ​​maintained. When the predicted value of the remaining life prediction information is less than the preset life threshold, the unit of the key component corresponding to the predicted value is marked. The yaw angle limit of the marked unit is upgraded to the corresponding restriction level, and the first constraint, the second constraint and the optimization objective are updated. Then, the step of iteratively solving the yaw optimization model of the wind farm wake is executed until the target farm-level yaw control matrix is ​​obtained.

6. An active wake yaw control device considering fatigue damage of wind turbine units, characterized in that, The device includes: The construction module is used to construct the wind farm wake yaw optimization model of the wind turbine by using wind condition data, actual coordinate information, wind turbine operating parameters and preset engineering wake model. The calculation module is used to calculate the structural load of the wind turbine, obtain the load time series data of the wind turbine, and calculate the target damage equivalent load data of each load time series data. The evaluation module is used to perform fatigue damage assessment on the equivalent load data of each target damage, and determine the yaw limit angle of each wind turbine unit in the fatigue damage based on the evaluation results. The solution module is used to iteratively solve the wind farm wake yaw optimization model with the yaw limit angle as a constraint and the combined optimization of power generation and fatigue damage of the wind farm group as the objective, so as to obtain the wind turbine's farm-level yaw control matrix for active wake control.

7. The apparatus according to claim 6, characterized in that, The evaluation module includes: Based on the actual installed unit control strategy and unit safety design rules, a fatigue damage assessment model is constructed. The evaluation unit is used to input the target damage equivalent load data into the fatigue damage evaluation model to perform fatigue damage evaluation and generate evaluation results. The storage unit is used to extract the yaw limit angle considering the fatigue damage of the unit from the evaluation results, and input the yaw limit angle into the unit-level yaw damage analysis dataset for storage.

8. An electronic device, characterized in that, include: The system includes a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to perform the active wake yaw control method considering wind turbine fatigue damage as described in any one of claims 1 to 5.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to execute the active wake yaw control method considering wind turbine fatigue damage as described in any one of claims 1 to 5.

10. A computer program product, characterized in that, Includes computer instructions for causing a computer to execute the active wake yaw control method considering wind turbine fatigue damage as described in any one of claims 1 to 5.

Images