A method and device for coordinating primary frequency modulation and inertia response of a wind farm

Abstract

The application discloses a kind of wind farm primary frequency modulation and inertia response coordination control method, comprising: real-time acquisition wind farm grid-connected point frequency and the operating state data of unit, calculate frequency deviation and frequency change rate;When meeting the set condition, trigger frequency modulation response, calculate dynamic adaptive frequency modulation coefficient according to the operating state data of each wind turbine, calculate the total increment of all field frequency modulation power;Multi-objective optimization allocation model is constructed;Improved particle swarm optimization algorithm is used to solve multi-objective optimization allocation model, and the power distribution value of each wind turbine is obtained;The power distribution value is issued to each wind turbine, and each wind turbine realizes inertia response by converter torque control first, and then realizes primary frequency modulation by adjusting speed set point.The method adapts to the operating characteristics of wind power generation, effectively makes up for the industry short board that wind power lacks natural moment of inertia and poor coordination of frequency modulation response, and realizes the orderly collaborative control of inertia response and primary frequency modulation.

Description

Technical Field

[0001] This invention relates to the field of wind power generation technology, specifically to a method and apparatus for coordinated control of primary frequency regulation and inertial response in a wind farm. Background Technology

[0002] Wind energy, as a sustainable new energy source, is receiving increasing attention from countries around the world due to its pollution-free and renewable nature. In recent years, my country's wind power industry has developed rapidly, with the installed capacity of wind power continuously increasing. As more and more large-capacity wind farms are directly connected to the power grid, the interaction between wind power and the power grid is becoming increasingly complex.

[0003] In wind power technology, variable-speed constant-frequency wind turbines are currently the main commercially available type. These turbines connect to the grid via converters, decoupling the rotor from the system frequency and preventing them from automatically providing inertia support when the system frequency changes. Furthermore, most wind turbines currently operate near their maximum power point using maximum wind energy capture control, which fails to provide frequency regulation reserve capacity during active power adjustments, further increasing the frequency regulation pressure on the system. Today, frequency control in wind farms is receiving increasing attention from power companies, leading to related technical requirements. Wind farms possessing their own frequency regulation capabilities and participating in grid frequency adjustments are key characteristics of grid-friendly wind farms. With a high proportion of wind power connected to the grid, the rotational inertia and primary frequency regulation capabilities of traditional synchronous turbines are continuously declining, posing a serious challenge to grid frequency stability. Currently, wind farms participating in grid frequency support mainly face the following problems:

[0004] 1. Separation of inertia response and primary frequency control: Inertia response relies heavily on converter fast torque control, while primary frequency control relies heavily on speed / pitch angle adjustment. Independent control of the two can easily lead to power command conflicts and regulation overshoot, affecting frequency support performance.

[0005] 2. Fixed frequency regulation coefficient and poor adaptability: Traditional primary frequency regulation often uses a fixed droop coefficient, which does not take into account the differences in frequency regulation margin under different wind speeds and unit health conditions. This can easily lead to overshoot in low-margin operating conditions and insufficient support in high-margin operating conditions.

[0006] 3. Power allocation focuses only on response performance, neglecting unit safety and economy: Existing allocation methods mostly aim for optimal response time, without considering unit fatigue load, cumulative damage and differences in operation and maintenance costs. Long-term operation will aggravate equipment wear and tear and increase operation and maintenance costs.

[0007] 4. Lack of multi-objective collaborative optimization mechanism: It is difficult to simultaneously take into account frequency regulation response speed, unit safety and operation economy, and it is impossible to achieve a balance between frequency support performance and the full life cycle benefits of wind farm.

[0008] Therefore, there is an urgent need for a coordinated control method for primary frequency regulation and inertial response in wind farms to solve the above problems. Summary of the Invention

[0009] The purpose of this invention is to provide a method and device for coordinated control of primary frequency regulation and inertia response in wind farms. This method is adapted to the operating characteristics of wind power generation and effectively compensates for the industry shortcomings of wind power grid connection, such as the lack of natural rotational inertia and poor coordination of frequency regulation response. Through a hierarchical response mechanism that prioritizes torque and then speed, coordinated control of inertia response and primary frequency regulation is achieved, avoiding regulation conflicts and overshoot, and improving frequency support effect.

[0010] This invention is achieved through the following technical solution:

[0011] In a first aspect, embodiments of the present invention provide a coordinated control method for primary frequency regulation and inertial response in a wind farm, comprising:

[0012] Real-time data collection of wind farm grid connection frequency and operating status of each wind turbine unit; calculation of frequency deviation and frequency change rate; and preprocessing of the collected data.

[0013] When the frequency deviation exceeds the preset dead zone and the total active power of the wind farm is greater than the set value of the total rated capacity, the frequency regulation response is triggered. The dynamic adaptive frequency regulation coefficient is calculated based on the operating status data of each wind turbine. The total increment of frequency regulation power of the entire field is calculated based on the frequency deviation, the dynamic adaptive frequency regulation coefficient and the real-time output active power of the wind turbine.

[0014] A multi-objective optimization allocation model is constructed with the total increment of frequency modulation power across the entire field as a constraint. The multi-objective optimization allocation model takes frequency modulation response time, fatigue load increment and frequency modulation economy as optimization objectives, and power balance, single unit power regulation capability, rotor speed safety and fatigue damage threshold as constraints.

[0015] An improved particle swarm optimization algorithm is used to solve the multi-objective optimization allocation model to obtain the power allocation value for each wind turbine.

[0016] The power allocation value is sent to each wind turbine. Each wind turbine first achieves inertial response through converter torque control, and then achieves primary frequency regulation by adjusting the speed setpoint, thus completing the coordinated control of the power of the entire field.

[0017] Furthermore, the real-time acquisition of wind farm grid connection frequency and operating status data of each wind turbine, calculation of frequency deviation and frequency change rate, and preprocessing of the acquired data specifically include:

[0018] Real-time acquisition of wind farm grid connection frequency, setting of sampling period, and calculation of frequency deviation and frequency change rate;

[0019] Collect wind speed, rotor speed, current active power, main shaft torque, tower inference, cumulative fatigue damage and operation and maintenance cost coefficient for each wind turbine unit;

[0020] The collected data is filtered to remove outliers.

[0021] Furthermore, the calculation of the dynamic adaptive frequency regulation coefficient based on the operating status data of each wind turbine specifically includes:

[0022] The available frequency control margin for each unit is calculated based on the current wind speed, rotational speed, and power. The available frequency control margin includes both upward and downward margins.

[0023] The dynamic adaptive frequency modulation coefficient is obtained by multiplying the margin correction coefficient, fatigue damage correction coefficient, and economic correction coefficient with the basic frequency modulation coefficient.

[0024] Furthermore, the objective function of the multi-objective optimization allocation model is:

[0025] ;

[0026] Wherein, J1 is the frequency modulation response time target, J2 is the fatigue load increment target, and J3 is the frequency modulation economy target. , , These are the weighting coefficients corresponding to the frequency regulation response time target, the fatigue load increment target, and the frequency regulation economy target, respectively, satisfying... .

[0027] Furthermore, the method of using an improved particle swarm optimization algorithm to solve the multi-objective optimization allocation model specifically includes:

[0028] The frequency regulation power increment of each wind turbine is used as the particle position code, and one particle corresponds to a set of power distribution schemes for the entire field.

[0029] A weighted comprehensive objective function with the goals of frequency modulation response time, fatigue load increment, and frequency modulation economy is used as the fitness function to calculate the fitness value of each particle.

[0030] By introducing dynamic inertia weights, adaptive mutation mechanisms, and constraint penalty mechanisms, the optimal power allocation value that satisfies all constraints is obtained by iteratively updating the optimal position of individual particles and the global optimal position, and by taking the maximum number of iterations or the change in the global optimal fitness less than a preset threshold as the termination condition.

[0031] Furthermore, the dynamic inertia weight The calculation formula is:

[0032] ;

[0033] in, This represents the current iteration number. The maximum number of iterations, , These are the maximum and minimum values ​​of the inertia weight, respectively.

[0034] Furthermore, the constraint penalty mechanism applies a penalty term to the fitness function for particles that violate constraints on power balance, rotor speed safety, or fatigue damage threshold, ensuring that the solution meets all constraints.

[0035] Secondly, another embodiment of the present invention provides a coordinated control device for primary frequency regulation and inertia response in a wind farm, used to implement the method described in the above embodiments, the device comprising:

[0036] The data acquisition and preprocessing module is used to collect real-time data on the frequency of the wind farm's grid connection point and the operating status of each wind turbine, calculate the frequency deviation and frequency change rate, and preprocess the collected data.

[0037] The calculation module is used to trigger the frequency regulation response when the frequency deviation exceeds the preset dead zone and the total active power of the wind farm is greater than the set value of the total rated capacity. It calculates the dynamic adaptive frequency regulation coefficient based on the operating status data of each wind turbine, and calculates the total frequency regulation power increment of the entire field based on the frequency deviation, the dynamic adaptive frequency regulation coefficient and the real-time output active power of the wind turbine.

[0038] The model building module constructs a multi-objective optimization allocation model with the total increment of frequency modulation power across the entire field as a constraint. The multi-objective optimization allocation model takes frequency modulation response time, fatigue load increment and frequency modulation economy as optimization objectives, and power balance constraint, single unit power regulation capability, rotor speed safety and fatigue damage threshold as constraint conditions.

[0039] The model solving module uses an improved particle swarm optimization algorithm to solve the multi-objective optimization allocation model and obtain the power allocation value for each wind turbine.

[0040] The hierarchical coordination response module is used to send power allocation values ​​to each wind turbine. Each wind turbine first achieves inertial response through converter torque control, and then achieves primary frequency regulation by adjusting the speed setpoint, thus completing the coordinated control of the power of the entire field.

[0041] Furthermore, the data acquisition and preprocessing module includes an acquisition unit and a preprocessing unit. The acquisition unit is used to acquire the grid connection frequency of the wind farm in real time, set the sampling period, calculate the frequency deviation and frequency change rate, and acquire the wind speed, rotor speed, current active power, main shaft torque, tower inference, cumulative fatigue damage and operation and maintenance cost coefficient of each wind turbine. The preprocessing unit is used to filter the acquired data and remove outliers.

[0042] Furthermore, the calculation module includes an adaptive frequency regulation coefficient calculation unit. The adaptive frequency regulation coefficient calculation unit calculates the available frequency regulation margin for each unit based on the current wind speed, rotational speed, and power. The available frequency regulation margin includes an upward adjustment margin and a downward adjustment margin. The dynamic adaptive frequency regulation coefficient is obtained by multiplying the margin correction coefficient, fatigue damage correction coefficient, and economic correction coefficient with the basic frequency regulation coefficient.

[0043] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0044] The present invention provides a method and apparatus for coordinated control of primary frequency regulation and inertial response in a wind farm, which has the following advantages:

[0045] 1. It can adapt to the operating characteristics of wind power generation, effectively making up for the industry's shortcomings of wind power grid connection, such as the lack of natural rotational inertia and poor frequency regulation response coordination. By coordinating the operating conditions, output fluctuations, and equipment health status of each unit in the wind farm, it achieves orderly and coordinated control of inertia response and primary frequency regulation. This not only enables rapid response to grid frequency fluctuations and strengthens the frequency support capability of wind power clusters for grid connection, but also meets the relevant technical specifications for wind power grid connection. It prioritizes rapid inertia support through converter torque control, and then completes continuous frequency regulation through speed adjustment, effectively suppressing grid frequency abrupt changes and secondary drops, and significantly improving the stability and smoothness of wind farm support for grid frequency.

[0046] 2. With frequency regulation response time, fatigue load increment, and frequency regulation economy as optimization objectives, and power balance, single unit regulation capability, rotor speed safety, and fatigue damage threshold as constraints, a multi-objective optimization allocation model is constructed. Under the premise of meeting the total frequency regulation power requirements of the entire field, more regulation can be allocated to units with large margin, high health, and low cost, so as to achieve differentiated and precise allocation, effectively shorten response time, reduce fatigue damage, reduce operation and maintenance costs, and achieve the optimal synergy among the three.

[0047] 3. By introducing dynamic inertia weights, adaptive mutation mechanisms, and constraint penalty mechanisms into the particle swarm optimization algorithm, we can balance global search and local optimization capabilities, avoid getting trapped in local optima, and ensure that the solution strictly meets all constraints. This allows for the rapid output of the optimal power allocation scheme, meeting the real-time requirements of frequency modulation control and improving the robustness and engineering practicality of the control strategy.

[0048] 4. No new hardware equipment is required. It can be implemented in the existing wind farm main control system only by upgrading the control logic. It is suitable for the renovation of existing wind farms with different turbine models and different operating years. The renovation cycle is short, the cost is low, and the compatibility is good. It can effectively improve the grid frequency stability level under large-scale wind power grid connection and has important engineering application value and promotion prospects. Attached Figure Description

[0049] To more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of the present invention and should not be considered as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort. In the drawings:

[0050] Figure 1 A flowchart illustrating a coordinated control method for primary frequency regulation and inertial response in a wind farm, provided in the first embodiment of the present invention;

[0051] Figure 2 This is a schematic diagram of the structure of a coordinated control device for primary frequency regulation and inertia response in a wind farm, provided in another embodiment of the present invention. Detailed Implementation

[0052] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.

[0053] like Figure 1 As shown, the first embodiment of the present invention provides a coordinated control method for primary frequency regulation and inertial response in a wind farm, comprising the following steps:

[0054] Real-time data collection of wind farm grid connection frequency and operating status of each wind turbine unit; calculation of frequency deviation and frequency change rate; and preprocessing of the collected data.

[0055] When the frequency deviation exceeds the preset dead zone and the total active power of the wind farm is greater than the set value of the total rated capacity, the frequency regulation response is triggered. The dynamic adaptive frequency regulation coefficient is calculated based on the operating status data of each wind turbine. The total increment of frequency regulation power of the entire field is calculated based on the frequency deviation, the dynamic adaptive frequency regulation coefficient and the real-time output active power of the wind turbine.

[0056] A multi-objective optimization allocation model is constructed with the total increment of frequency modulation power across the entire field as a constraint. The multi-objective optimization allocation model takes frequency modulation response time, fatigue load increment and frequency modulation economy as optimization objectives, and power balance constraint, single unit power regulation capability, rotor speed safety and fatigue damage threshold as constraint conditions.

[0057] An improved particle swarm optimization algorithm is used to solve the multi-objective optimization allocation model to obtain the power allocation value for each wind turbine.

[0058] The power allocation value is sent to each wind turbine. Each wind turbine first achieves inertial response through converter torque control, and then achieves primary frequency regulation by adjusting the speed setpoint, thus completing the coordinated control of the power of the entire field.

[0059] This embodiment takes a wind farm with an installed capacity of 114MW (38 units × 3.0MW) as an example to illustrate the coordinated control method of primary frequency regulation and inertia response of the wind farm provided by the embodiment of the present invention.

[0060] Data acquisition includes grid-side data acquisition and generator-side operational status data acquisition. Grid-side data acquisition includes real-time acquisition of the wind farm's grid connection frequency. With the sampling period t set to 100ms, the frequency deviation was calculated. The formula is:

[0061] ;

[0062] Calculate the rate of change of frequency The formula is:

[0063] ;

[0064] in, The sampling interval is denoted as .

[0065] Unit-side operating status data acquisition: For each wind turbine i, i=1,2,3…N, N=38, collect the following data: wind speed Rotor speed Current active power Spindle torque Tower-like reasoning Cumulative fatigue damage and maintenance cost coefficient The operation and maintenance cost coefficient is related to the unit's operating years and maintenance history, and its value ranges from 1 to 3. The higher the value, the higher the operation and maintenance cost.

[0066] The collected data is low-pass filtered to remove abrupt outliers, such as wind speed exceeding the cut-out wind speed or power exceeding 1.2 times the rated capacity, to ensure data reliability.

[0067] When the frequency deviation exceeds the preset dead zone and the total active power of the wind farm exceeds the set value of the total rated capacity, the frequency regulation response is triggered, specifically including:

[0068] | |> , As a preset frequency dead zone, this embodiment sets the frequency dead zone to ±0.05Hz; and ,in, This represents the current total active power of the wind farm. The total capacity is set, triggering a frequency modulation response.

[0069] Based on the current wind speed, engine speed, and power, the available frequency regulation margin for each unit is calculated as follows:

[0070] Increase margin :

[0071] ,

[0072] in, This represents the maximum active power that the unit can output at the current wind speed.

[0073] Reduce margin :

[0074] ,

[0075] in, The minimum active power required for safe operation of the unit.

[0076] The dynamic adaptive frequency regulation coefficient is calculated based on the operating status data of each wind turbine. The formula is:

[0077]

[0078] in, To fix the base coefficient, the value ranges from 10 to 50. In this embodiment... , This is a margin correction factor; the larger the available frequency regulation margin of the unit, the better. The larger the value, the range is 0.8~1.2. This is a fatigue damage correction factor; the higher the cumulative damage to the unit, the greater the fatigue damage. The smaller the value, the more likely it is to be in the range of 0.7 to 1.0. This is an economic adjustment factor; the higher the operation and maintenance cost factor, the better. The smaller the value, the range is 0.8~1.0.

[0079] By combining the grid frequency deviation and the dynamic frequency regulation coefficient, the total increase in frequency regulation power across the entire field is calculated. :

[0080] ,

[0081] And set a field-wide amplitude limit constraint: ,in, To preset the frequency modulation power limit for the entire field, the value is generally set to 6% to 10% of the total rated capacity.

[0082] Specific methods for constructing multi-objective optimization allocation models include:

[0083] Construct a multi-objective optimization function F with the objectives of response time, fatigue load increment, and frequency modulation economy:

[0084] ;

[0085] in, The target for frequency regulation response time is to minimize the weighted sum of the response delays between the unit power regulation command and the actual power. The target for fatigue load increment is the weighted sum and minimization of the fatigue damage increment caused by changes in spindle torque and tower thrust. To achieve the goal of frequency regulation economy, the objective is to minimize the product of the unit operation and maintenance cost coefficient and the power regulation amount; , , These are the weighting coefficients corresponding to the frequency regulation response time target, the fatigue load increment target, and the frequency regulation economy target, respectively, satisfying... It can be configured according to the operational needs of the wind farm. In this embodiment, , =0.3, .

[0086] Constraints:

[0087] Power balance constraints:

[0088] ,

[0089] in, This represents the power allocation increment for the i-th wind turbine unit.

[0090] Single-unit power regulation capability constraints:

[0091] .

[0092] Rotor speed safety constraints: ;

[0093] in, This represents the speed adjustment amount.

[0094] Fatigue damage threshold constraint:

[0095]

[0096] in, This represents the increase in fatigue damage caused by this frequency modulation. The preset cumulative fatigue damage threshold for the unit.

[0097] By constructing a multi-objective optimization model with frequency regulation response time, fatigue load increment, and frequency regulation economy as optimization objectives, the model avoids the disadvantages caused by a single objective: ensuring rapid frequency support response, suppressing fatigue damage caused by main shaft torque and tower thrust, and reducing the operation and maintenance costs of the frequency regulation process, achieving optimal synergy among the three. Combining multiple hard constraints such as single-unit power regulation capability, rotor speed safety, and fatigue damage threshold, the model allocates more regulation to units with large margins, high health, and low cost, while limiting the output of units with high fatigue and small margins, under the premise of meeting the total frequency regulation power requirements of the entire field. This prevents local units from being under heavy load for a long time and accumulating damage, thus extending unit life. Multi-objective optimization achieves smooth, balanced, and controllable power distribution across the entire field, avoiding power abrupt changes or overshoot caused by proportional distribution, reducing the risk of secondary frequency drops or oscillations, and improving grid frequency stability. The model comprehensively considers real-time operating conditions, unit health differences, and economic costs, and can still output reasonable allocation schemes under extreme conditions such as low wind speed, high fatigue, and high cost, improving the robustness and engineering practicality of the control strategy.

[0098] After constructing the multi-objective optimization allocation model, an improved particle swarm optimization algorithm is used to iteratively solve the model, and finally obtain the optimal frequency regulation power increment allocation value of each wind turbine that satisfies all constraints. The specific solution process is as follows:

[0099] The frequency regulation power increment required for each wind turbine in the entire wind farm is encoded as a particle position variable. Assuming there are N wind turbines in the wind farm, the position of a single particle... It can be represented as:

[0100]

[0101] in, The particle position represents the power allocation increment for the i-th wind turbine, with each dimension corresponding to the power adjustment of one turbine. Each particle represents a power allocation scheme for all wind turbines in the field. The particle velocity represents the adjustment step size of the power increment in each dimension, used to control the search direction and magnitude of the solution during the iteration process.

[0102] A multi-objective weighted composite objective function is adopted as the particle fitness function, and the objective function expression is:

[0103] .

[0104] The particle swarm size is set to 50-100; in this embodiment, the particle swarm size is set to 60. The maximum number of iterations is set to 80-120; in this embodiment, the maximum number of iterations is 100. Particle positions are randomly initialized within the available frequency regulation margin of each unit, i.e.:

[0105] ,

[0106] in, To reduce the margin of the unit, Increase the margin for the unit; set the initial particle velocity to 0 to ensure that the initial search starts from within the feasible region.

[0107] To ensure the solution satisfies all model constraints, a constraint penalty mechanism is introduced. For particles that violate power balance constraints, single-machine power regulation capability constraints, rotor speed safety constraints, or fatigue damage threshold constraints, a penalty term is applied to their fitness function, significantly increasing the fitness value of infeasible solutions and thus gradually eliminating them during iteration. The penalized fitness function is shown below. for:

[0108] ,

[0109] in, The penalty coefficient is typically set between 500 and 2000. In this embodiment, it is set to 1000. V represents the constraint violation amount, ensuring that the algorithm always searches for the optimal solution within the feasible region.

[0110] A dynamic inertia weight strategy is introduced to balance the algorithm's global search capability with its local optimization capability. Dynamic inertia weight. The calculation formula is:

[0111] ;

[0112] in, This represents the current iteration number. The maximum number of iterations, , These are the maximum and minimum values ​​of the inertia weight, respectively. , =0.4. The inertia weight is larger in the early stages of iteration, which is beneficial for global large-scale search; the inertia weight decreases in the later stages of iteration, enhancing the ability for fine-grained local search and improving convergence accuracy.

[0113] The particle velocity and position are iteratively updated according to the standard particle swarm optimization formula. To avoid the algorithm getting trapped in local optima, an adaptive mutation mechanism is introduced. During the iteration process, every preset number of iterations (e.g., 10-20 times), a small number of particles (e.g., 5%-10%) are randomly selected, and their positions are subjected to a small-range random perturbation.

[0114]

[0115] in, For the new position of the particle, This is the particle's old position. The perturbation step size decreases with the number of iterations. The random numbers are generated according to a standard normal distribution. This mechanism enhances particle swarm diversity and improves the algorithm's ability to escape local optima.

[0116] During the iteration process, the individual optimal position and the global optimal position of each particle are continuously updated. Iteration stops when any of the following termination conditions are met:

[0117] 1. Reach the preset maximum number of iterations;

[0118] 2. The change in the global optimal fitness value is less than the preset threshold in multiple consecutive iterations.

[0119] After the iteration is completed, the global optimal position is output, which is the optimal frequency regulation power increment allocation value of each wind turbine, as the basis for subsequent power allocation.

[0120] The optimal frequency regulation power increment allocation value for each wind turbine is sent to each unit; within 0~1s, the unit responds quickly through converter torque control, releasing rotor kinetic energy to achieve inertial response support; within 1~15s, the speed setpoint is adjusted, and the output is continuously controlled by the pitch angle to continuously output frequency regulation power, completing one frequency regulation; after the frequency regulation is completed, the unit operation data is fed back in real time, and the rotor speed is gradually restored according to the preset recovery curve to avoid secondary frequency drop and ensure control accuracy.

[0121] The wind farm primary frequency regulation and inertia response coordinated control method proposed in this invention has the following advantages compared with the prior art:

[0122] 1. It can adapt to the operating characteristics of wind power generation, effectively making up for the industry's shortcomings of wind power grid connection, such as the lack of natural rotational inertia and poor frequency regulation response coordination. By coordinating the operating conditions, output fluctuations, and equipment health status of each unit in the wind farm, it achieves orderly and coordinated control of inertia response and primary frequency regulation. This not only enables rapid response to grid frequency fluctuations and strengthens the frequency support capability of wind power clusters for grid connection, but also meets the relevant technical specifications for wind power grid connection. It prioritizes rapid inertia support through converter torque control, and then completes continuous frequency regulation through speed adjustment, effectively suppressing grid frequency abrupt changes and secondary drops, and significantly improving the stability and smoothness of wind farm support for grid frequency.

[0123] 2. With frequency regulation response time, fatigue load increment, and frequency regulation economy as optimization objectives, and power balance, single unit regulation capability, rotor speed safety, and fatigue damage threshold as constraints, a multi-objective optimization allocation model is constructed. Under the premise of meeting the total frequency regulation power requirements of the entire field, more regulation can be allocated to units with large margin, high health, and low cost, so as to achieve differentiated and precise allocation, effectively shorten response time, reduce fatigue damage, reduce operation and maintenance costs, and achieve the optimal synergy among the three.

[0124] 3. By introducing dynamic inertia weights, adaptive mutation mechanisms, and constraint penalty mechanisms into the particle swarm optimization algorithm, we can balance global search and local optimization capabilities, avoid getting trapped in local optima, and ensure that the solution strictly meets all constraints. This allows for the rapid output of the optimal power allocation scheme, meeting the real-time requirements of frequency modulation control and improving the robustness and engineering practicality of the control strategy.

[0125] 4. No new hardware equipment is required. It can be implemented in the existing wind farm main control system only by upgrading the control logic. It is suitable for the renovation of existing wind farms with different turbine models and different operating years. The renovation cycle is short, the cost is low, and the compatibility is good. It can effectively improve the grid frequency stability level under large-scale wind power grid connection and has important engineering application value and promotion prospects.

[0126] like Figure 2 As shown in the embodiment of the present invention, a coordinated control device for primary frequency regulation and inertial response of a wind farm is provided, used to implement the coordinated control method for primary frequency regulation and inertial response of a wind farm described in the first embodiment above. The device includes:

[0127] The data acquisition and preprocessing module is used to collect real-time data on the frequency of the wind farm's grid connection point and the operating status of each wind turbine, calculate the frequency deviation and frequency change rate, and preprocess the collected data.

[0128] The calculation module is used to trigger the frequency regulation response when the frequency deviation exceeds the preset dead zone and the total active power of the wind farm is greater than the set value of the total rated capacity. It calculates the dynamic adaptive frequency regulation coefficient based on the operating status data of each wind turbine, and calculates the total frequency regulation power increment of the entire field based on the frequency deviation, the dynamic adaptive frequency regulation coefficient and the real-time output active power of the wind turbine.

[0129] The model building module constructs a multi-objective optimization allocation model with the total increment of frequency modulation power across the entire field as a constraint. The multi-objective optimization allocation model takes frequency modulation response time, fatigue load increment and frequency modulation economy as optimization objectives, and power balance constraint, single unit power regulation capability, rotor speed safety and fatigue damage threshold as constraint conditions.

[0130] The model solving module uses an improved particle swarm optimization algorithm to solve the multi-objective optimization allocation model and obtain the power allocation value for each wind turbine.

[0131] The hierarchical coordination response module is used to send power allocation values ​​to each wind turbine. Each wind turbine first achieves inertial response through converter torque control, and then achieves primary frequency regulation by adjusting the speed setpoint, thus completing the coordinated control of the power of the entire field.

[0132] The data acquisition and preprocessing module includes an acquisition unit and a preprocessing unit. The acquisition unit is used to collect the grid connection frequency of the wind farm in real time, set the sampling period, calculate the frequency deviation and frequency change rate, and collect the wind speed, rotor speed, current active power, main shaft torque, tower inference, cumulative fatigue damage and operation and maintenance cost coefficient of each wind turbine. The preprocessing unit is used to filter the collected data and remove outliers.

[0133] The calculation module includes an adaptive frequency regulation coefficient calculation unit. The adaptive frequency regulation coefficient calculation unit calculates the available frequency regulation margin for each unit based on the current wind speed, rotational speed and power. The available frequency regulation margin includes an upward adjustment margin and a downward adjustment margin. The dynamic adaptive frequency regulation coefficient is obtained by multiplying the margin correction coefficient, fatigue damage correction coefficient and economic correction coefficient with the basic frequency regulation coefficient.

[0134] The execution process of each module can be carried out according to the steps of the coordinated control method for primary frequency regulation and inertia response of a wind farm provided in the first embodiment, and will not be described in detail in this embodiment.

[0135] The coordinated control device and method for primary frequency regulation and inertia response of a wind farm provided in this embodiment of the invention are based on the same inventive concept and have the same beneficial effects, and will not be described again here.

[0136] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. A coordinated control method for primary frequency regulation and inertial response in a wind farm, characterized in that, include: Real-time data collection of wind farm grid connection frequency and operating status of each wind turbine unit; calculation of frequency deviation and frequency change rate; and preprocessing of the collected data. When the frequency deviation exceeds the preset dead zone and the total active power of the wind farm is greater than the set value of the total rated capacity, the frequency regulation response is triggered. The dynamic adaptive frequency regulation coefficient is calculated based on the operating status data of each wind turbine. The total increment of frequency regulation power of the entire field is calculated based on the frequency deviation, the dynamic adaptive frequency regulation coefficient and the real-time output active power of the wind turbine. A multi-objective optimization allocation model is constructed with the total increment of frequency modulation power across the entire field as a constraint. The multi-objective optimization allocation model takes frequency modulation response time, fatigue load increment and frequency modulation economy as optimization objectives, and power balance, single unit power regulation capability, rotor speed safety and fatigue damage threshold as constraints. An improved particle swarm optimization algorithm is used to solve the multi-objective optimization allocation model to obtain the power allocation value for each wind turbine. The power allocation value is sent to each wind turbine. Each wind turbine first achieves inertial response through converter torque control, and then achieves primary frequency regulation by adjusting the speed setpoint, thus completing the coordinated control of the power of the entire field.

2. The coordinated control method for primary frequency regulation and inertial response of a wind farm according to claim 1, characterized in that, The process involves real-time acquisition of wind farm grid connection frequency and operating status data for each wind turbine, calculation of frequency deviation and frequency change rate, and preprocessing of the acquired data, specifically including: Real-time acquisition of wind farm grid connection frequency, setting of sampling period, and calculation of frequency deviation and frequency change rate; Collect wind speed, rotor speed, current active power, main shaft torque, tower inference, cumulative fatigue damage and operation and maintenance cost coefficient for each wind turbine unit; The collected data is filtered to remove outliers.

3. The coordinated control method for primary frequency regulation and inertial response of a wind farm according to claim 1, characterized in that, The calculation of the dynamic adaptive frequency regulation coefficient based on the operating status data of each wind turbine specifically includes: The available frequency control margin for each unit is calculated based on the current wind speed, rotational speed, and power. The available frequency control margin includes both upward and downward margins. The dynamic adaptive frequency modulation coefficient is obtained by multiplying the margin correction coefficient, fatigue damage correction coefficient, and economic correction coefficient with the basic frequency modulation coefficient.

4. The coordinated control method for primary frequency regulation and inertial response of a wind farm according to claim 1, characterized in that, The objective function of the multi-objective optimization allocation model is: ； Wherein, J1 is the frequency modulation response time target, J2 is the fatigue load increment target, and J3 is the frequency modulation economy target. , , These are the weighting coefficients corresponding to the frequency regulation response time target, the fatigue load increment target, and the frequency regulation economy target, respectively, satisfying... .

5. The coordinated control method for primary frequency regulation and inertial response of a wind farm according to claim 4, characterized in that, The method of using an improved particle swarm optimization algorithm to solve the multi-objective optimal allocation model specifically includes: The frequency regulation power increment of each wind turbine is used as the particle position code, and one particle corresponds to a set of power distribution schemes for the entire field. A weighted comprehensive objective function with the goals of frequency modulation response time, fatigue load increment, and frequency modulation economy is used as the fitness function to calculate the fitness value of each particle. By introducing dynamic inertia weights, adaptive mutation mechanisms, and constraint penalty mechanisms, the optimal power allocation value that satisfies all constraints is obtained by iteratively updating the optimal position of individual particles and the global optimal position, and by taking the maximum number of iterations or the change in the global optimal fitness less than a preset threshold as the termination condition.

6. The coordinated control method for primary frequency regulation and inertial response of a wind farm according to claim 5, characterized in that, The dynamic inertia weight The calculation formula is: ; in, This represents the current iteration number. The maximum number of iterations, , These are the maximum and minimum values ​​of the inertia weight, respectively.

7. The coordinated control method for primary frequency regulation and inertial response of a wind farm according to claim 6, characterized in that, The constraint penalty mechanism applies a penalty term to the fitness function for particles that violate constraints on power balance, rotor speed safety, or fatigue damage threshold, ensuring that the solution meets all constraints.

8. A coordinated control device for primary frequency regulation and inertial response in a wind farm, characterized in that, The apparatus for implementing the method as described in any one of claims 1-7 comprises: The data acquisition and preprocessing module is used to collect real-time data on the frequency of the wind farm's grid connection point and the operating status of each wind turbine, calculate the frequency deviation and frequency change rate, and preprocess the collected data. The calculation module is used to trigger the frequency regulation response when the frequency deviation exceeds the preset dead zone and the total active power of the wind farm is greater than the set value of the total rated capacity. It calculates the dynamic adaptive frequency regulation coefficient based on the operating status data of each wind turbine, and calculates the total frequency regulation power increment of the entire field based on the frequency deviation, the dynamic adaptive frequency regulation coefficient and the real-time output active power of the wind turbine. The model building module constructs a multi-objective optimization allocation model with the total increment of frequency modulation power across the entire field as a constraint. The multi-objective optimization allocation model takes frequency modulation response time, fatigue load increment and frequency modulation economy as optimization objectives, and power balance, single-unit power regulation capability, rotor speed safety and fatigue damage threshold as constraints. The model solving module uses an improved particle swarm optimization algorithm to solve the multi-objective optimization allocation model and obtain the power allocation value for each wind turbine. The hierarchical coordination response module is used to send power allocation values ​​to each wind turbine. Each wind turbine first achieves inertial response through converter torque control, and then achieves primary frequency regulation by adjusting the speed setpoint, thus completing the coordinated control of the power of the entire field.

9. The coordinated control device for primary frequency regulation and inertial response of a wind farm according to claim 8, characterized in that, The data acquisition and preprocessing module includes an acquisition unit and a preprocessing unit. The acquisition unit is used to acquire the grid connection frequency of the wind farm in real time, set the sampling period, calculate the frequency deviation and frequency change rate, and acquire the wind speed, rotor speed, current active power, main shaft torque, tower inference, cumulative fatigue damage and operation and maintenance cost coefficient of each wind turbine. The preprocessing unit is used to filter the acquired data and remove outliers.

10. The coordinated control device for primary frequency regulation and inertial response of a wind farm according to claim 8, characterized in that, The calculation module includes an adaptive frequency regulation coefficient calculation unit. The adaptive frequency regulation coefficient calculation unit calculates the available frequency regulation margin for each unit based on the current wind speed, rotational speed and power. The available frequency regulation margin includes an upward adjustment margin and a downward adjustment margin. The dynamic adaptive frequency regulation coefficient is obtained by multiplying the margin correction coefficient, fatigue damage correction coefficient and economic correction coefficient with the base frequency regulation coefficient.

Images