Coordinated control method and system for new energy unit and station participating in primary frequency modulation

Abstract

The application discloses a kind of new energy unit and station participate in primary frequency modulation coordinated control method and system.Based on the unit model of new energy power supply, frequency controller link and the load shedding operation algorithm equipped, establish the primary frequency modulation model of new energy power supply;According to the output power variation of new energy power supply, the output power variation of traditional power supply and load demand, calculate the power system frequency deviation;In load shedding operation condition, the effective wind speed of each unit, rotor kinetic energy reserve and power reserve index are collected in real time, and its standardized score is calculated, and the entropy weight method and variation coefficient method are combined to assign weight, to calculate the comprehensive score of each unit frequency modulation potential;Based on the grid frequency deviation, determine the total frequency modulation power demand of new energy station, combined with the comprehensive score of frequency modulation potential, the responsibility proportion and responsibility power of each unit frequency modulation are calculated in real time by dynamic weight distribution algorithm.

Description

Technical Field

[0001] This invention belongs to the field of power system frequency analysis and control, and particularly relates to a collaborative control method and system for new energy generating units and power plants to participate in primary frequency regulation. Background Technology

[0002] my country's installed capacity of new energy wind power and photovoltaic power has continued to surge. In 2024, the total installed capacity of wind power and photovoltaic power in China exceeded 1.2 billion kilowatts, accounting for more than 45% of the total installed power generation capacity. However, the inherent intermittency and volatility of new energy power sources result in weak support for grid frequency. The traditional primary frequency regulation system, which relies on thermal power and hydropower, faces severe challenges. In particular, when power disturbances occur in the grid, if new energy power plants cannot respond quickly to frequency deviations, frequency instability risks are likely to occur.

[0003] Current domestic research on renewable energy frequency regulation largely focuses on optimizing control strategies for individual generating units, such as improving response speed through rotor kinetic energy control and virtual inertial control. However, the lack of coordination mechanisms at the power plant level often leads to imbalances, with some units overloaded and others having idle frequency regulation potential, making it difficult to adapt to the grid frequency regulation requirements after large-scale renewable energy grid integration. International research focuses on two directions: one is renewable energy frequency regulation strategies based on model predictive control (MPC), adjusting output in advance by predicting power fluctuations; the other is integrating distributed renewable energy resources for frequency regulation using an aggregator model. However, international frequency regulation schemes are mostly designed for high-penetration renewable energy grids, based on their relatively dispersed grid structure, making it difficult to directly adapt to my country's "large grid, large base" renewable energy grid integration model. While aggregation control technology achieves multi-unit coordination, it does not consider the dynamic characteristics of renewable energy units, such as the differences in frequency regulation potential of wind power at different wind speeds, resulting in insufficient frequency regulation accuracy and system stability. Against this backdrop, this invention will focus on breaking through the integrated frequency regulation technology of "precise assessment-dynamic coordination-rapid response" at the renewable energy power plant level to meet the grid frequency stability requirements after large-scale renewable energy grid integration.

[0004] In summary, most existing solutions suffer from three major flaws: First, the evaluation dimension of frequency regulation potential is too narrow, with most technologies using only power reserve as an evaluation indicator, ignoring dynamic parameters such as effective wind speed and rotor kinetic energy reserve, leading to significant deviations in potential evaluation. Second, the weighting allocation method is rigid, employing either subjective weighting or a single objective weighting, failing to balance the objectivity of the indicators with actual engineering needs. Third, the coordinated response is lagging, with centralized control relying on centralized calculations by the station controller and distributed control relying on cross-unit communication, both of which struggle to meet the requirements of "millisecond-level response and second-level stability" for primary frequency regulation. Summary of the Invention

[0005] The purpose of this invention is to address the aforementioned problems by proposing a collaborative control method and system for primary frequency regulation involving new energy generating units and power plants. This method establishes a precise frequency regulation model by integrating unit models, frequency controllers, and load reduction algorithms, thus solving the problem of "low matching degree between frequency regulation response and frequency deviation." Furthermore, it achieves accurate quantification of frequency regulation potential through multi-dimensional indicators and combined weighting methods, addressing the problem of "large deviation in potential assessment." Finally, it uses a dynamic weight allocation algorithm to allocate frequency regulation responsibilities in real time, enabling adaptive collaboration among generating units.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] A coordinated control method for primary frequency regulation involving new energy generating units and power plants includes the following steps:

[0008] S1. Based on the generator set model and frequency controller of new energy power source, establish a primary frequency regulation model of new energy power source. The new energy generator set needs to be equipped with a control algorithm to achieve load reduction. Before establishing the model, the historical operating data of the generator set needs to be filtered and preprocessed, and the Kalman filter algorithm is used to remove abnormal data.

[0009] S2. Calculate the power system frequency deviation based on the output power variation of new energy power sources, the output power variation of traditional power sources, and load demand; at the same time, correct the frequency deviation calculation results by combining the grid voltage deviation (±5% of the rated voltage range), and the correction coefficient is determined by the coupling relationship between voltage deviation and frequency deviation through experiments.

[0010] S3. Under reduced load operation conditions, the effective wind speed, rotor kinetic energy reserve, and power reserve indicators of each unit are collected in real time, and their standardized scores are calculated. The entropy weight method and the coefficient of variation method are used to assign weights to calculate the comprehensive score of frequency regulation potential for each unit. When collecting indicators, the health status parameters of the unit (bearing temperature, gearbox vibration value) must be obtained simultaneously. When the bearing temperature exceeds 85℃ or the gearbox vibration value is greater than 0.15mm, the comprehensive score of frequency regulation potential of the unit is reduced by 20%-30%.

[0011] S4. Based on the grid frequency deviation, determine the total frequency regulation power demand of new energy power plants. Combined with the comprehensive score of frequency regulation potential, calculate the frequency regulation responsibility ratio and responsibility power of each unit in real time through a dynamic weight allocation algorithm. The update cycle of the dynamic weight allocation algorithm is set to 50-100ms. When the grid frequency deviation change rate exceeds 0.1Hz / s, the algorithm emergency update mechanism is triggered, and the update cycle is shortened to 20ms.

[0012] Furthermore, the unit model in S1 represents the relationship between the unit's output power and the frequency control command, and simulates the dynamic response of the converter through a first-order inertial element; the frequency controller element in S1 includes inertial control and droop control; the response delay time of the inertial control does not exceed 50ms, and the adjustment accuracy of the droop control is controlled within the range of ±0.02Hz.

[0013] Furthermore, the primary frequency regulation model for establishing the new energy power source in S1 is as follows:

[0014] ;

[0015] in, This serves as a reference value for the change in output power of new energy generating units; , These are the inertia control coefficient and droop control coefficient of the frequency controller for the new energy unit, respectively. For power system frequency deviation;

[0016] The converter time constant of the new energy unit. Differential operator; converter time constant Corrections need to be made based on the unit's operating ambient temperature (-30℃ to 50℃). For every 10℃ change in temperature, Adjust by ±5%.

[0017] Furthermore, the reduced-load operation output power of the new energy unit in S1 is expressed as follows:

[0018] ;

[0019] in, For the output power of new energy units, Given the load reduction rate of the new energy unit,

[0020] The power value corresponding to the maximum power operating point of the new energy unit; load reduction rate. The value range is 5%-20%, and it needs to be dynamically adjusted according to the real-time load factor of the power grid (15%-20% when the load factor is <60%, and 5%-10% when the load factor is ≥60%).

[0021] Furthermore, the reduced-load operation of the new energy unit is achieved by applying different control algorithms in different wind speed ranges;

[0022] In the first wind speed range, load reduction operation is achieved solely through torque control;

[0023] In the second wind speed range, due to the limitation of the wind turbine rotor speed, load reduction operation is achieved through coordinated control of torque and blade pitch angle;

[0024] In the third wind speed range, due to the maximum power limitation of the converter, load reduction operation is achieved only through the pitch angle control; the boundary values ​​of each wind speed range can be finely adjusted by ±0.5m / s according to the unit model (such as 2.5MW, 3MW unit).

[0025] Furthermore, the load demand refers to the power fluctuation of the load, and the power system frequency deviation in S2 is calculated using the following formula:

[0026] ;

[0027] in, Indicates the frequency deviation of the power system. This refers to the change in output power of the new energy generating unit. This represents the change in the output power of the thermal power unit. For power fluctuations in the load; and These are the inertial time constant and damping coefficient of the power system, respectively; inertial time constant It needs to be based on the installed capacity ratio of new energy units in the power grid (for every 10% increase in the ratio, (Reduced by 8%) Dynamic updates.

[0028] Furthermore, in S3, different standardization formulas are used to obtain the standardized scores of the above three types of indicators for the collected effective wind speed, rotor kinetic energy reserve and power reserve indicators, that is, the following two weight determination methods and comprehensive score calculation are performed.

[0029] 1) Entropy weight is calculated using the following formula:

[0030] ;

[0031] in, For the first Entropy weight of each indicator, For the first The entropy value of each indicator.

[0032] Entropy is calculated using the following formula:

[0033] ;

[0034] in, For the first The entropy value of each indicator, For the number of units, For the first The unit is in the first The standardization percentage of the indicator.

[0035] 2) The weight of the coefficient of variation is calculated using the following formula:

[0036] ;

[0037] in, No. The weight of the coefficient of variation of each indicator, For the first The coefficient of variation of each indicator.

[0038] The coefficient of variation is calculated using the following formula:

[0039] ;

[0040] in, For the first The coefficient of variation of each indicator For the first The standard deviation of each indicator For the first The average of each indicator.

[0041] 3) The comprehensive score for the frequency regulation potential of each unit is calculated using the following formula:

[0042] ;

[0043] in, For the first The overall score of the frequency regulation potential of each generating unit The weight of the entropy weight. For the first Entropy weight of each indicator, For the first The weight of the coefficient of variation of each indicator, For the first The first unit Standardized scores of each indicator; weight of entropy weights. The value range is 0.4-0.6, and when the effective wind speed fluctuation exceeds 2m / s, The value increases by 0.1.

[0044] Furthermore, the frequency regulation responsibility ratio and responsibility power allocation of each unit in S4 are achieved by adjusting the inertia control coefficient and droop control coefficient of the frequency controller of the new energy unit;

[0045] The inertia control factor and droop control factor are calculated using the following formulas:

[0046] ;

[0047] ;

[0048] in, , For the first Inertia control coefficient and droop control coefficient for each unit For the first The overall score of the frequency regulation potential of each generating unit The virtual inertia provided by new energy power plants to the power grid. The droop coefficient of the new energy power station; virtual inertia. The maximum value shall not exceed 5% of the total installed capacity of the power station, and the droop coefficient It needs to be based on the absolute value of the power grid frequency deviation ( (The change will be dynamically adjusted by increasing the percentage by 10%).

[0049] Furthermore, it also includes a frequency regulation effect feedback loop, which monitors the grid frequency recovery in real time. When the frequency recovers to within ±0.05Hz of the rated frequency and the stabilization time exceeds 3s, the frequency regulation power recovery mechanism is triggered to gradually restore the unit's output power to the reduced load operating power. The recovery rate is set to 0.5MW / min-2MW / min, which is determined according to the unit type.

[0050] A coordinated control system for new energy generating units and power plants participating in primary frequency regulation is used to implement the coordinated control method for new energy generating units and power plants participating in primary frequency regulation as described above, characterized in that it includes the following four modules:

[0051] Model building module: Establish a primary frequency regulation model for new energy power sources. The model is based on the generator unit model and frequency controller model of the new energy power source. The new energy generator unit achieves load reduction operation through corresponding control algorithms. The model building module should have a self-verification function, compare the model output results with the actual operating data every 24 hours, and automatically trigger model parameter correction when the error exceeds 5%.

[0052] First calculation module: Based on the output power changes of new energy power sources, the output power changes of traditional power sources, and load demand, calculate the power system frequency deviation, speed changes, and power changes of each new energy unit through corresponding formulas. The first calculation module must have a data backup function, storing the calculation process data every 10 minutes, and the backup data retention period is 7 days.

[0053] The second calculation module: Under reduced load operation conditions, it collects the effective wind speed, rotor kinetic energy reserve and power reserve indicators of each unit in real time, and uses a combination of entropy weight method and coefficient of variation method to assign weights. It calculates the comprehensive score of frequency regulation potential of each unit through corresponding formulas. The second calculation module is equipped with a data anomaly alarm function. When the collected data exceeds the normal range for 3 consecutive times (effective wind speed > 25m / s, rotor kinetic energy reserve < 10% of rated value), an alarm signal is sent to the station monitoring center.

[0054] Responsibility allocation module: Based on the grid frequency deviation, the total frequency regulation power demand of the new energy power plant is determined. Combined with the comprehensive score of frequency regulation potential, the frequency regulation responsibility ratio and responsibility power of each unit are calculated in real time through a dynamic weight allocation algorithm. The responsibility allocation module needs to communicate with the unit control system in real time, and the communication delay shall not exceed 30ms.

[0055] Compared with the prior art, the present invention has the following advantages:

[0056] This invention preprocesses historical data using Kalman filtering to remove abnormal interference and combines the voltage deviation correction frequency calculation results to significantly improve the input accuracy and frequency deviation calculation accuracy of the primary frequency modulation model, providing reliable data support for subsequent frequency modulation decisions.

[0057] By incorporating unit health status parameters into frequency regulation potential assessment, overload operation of units in poor condition can be avoided. At the same time, by dynamically adjusting parameters such as load shedding rate and inertia time constant, frequency regulation reliability and power supply efficiency are balanced, and the service life of the units is extended.

[0058] Optimize the update cycle and emergency response mechanism of the dynamic weight allocation algorithm to ensure the real-time nature of frequency regulation responsibility allocation, which can quickly respond to rapid fluctuations in grid frequency and improve the response speed and control effect of primary frequency regulation.

[0059] By designing features such as converter time constant temperature correction and wind speed range model adaptation fine-tuning, the adaptability of the method and system to different environments and different models has been enhanced, expanding the application scope.

[0060] The system has added functions such as self-verification, data backup, and abnormal alarm, which improves operational stability and maintainability. The frequency regulation effect feedback link ensures that the unit transitions smoothly after the frequency stabilizes, further guaranteeing the safe operation of the power grid. Attached Figure Description

[0061] Figure 1 The flowchart shows the collaborative control method and system for primary frequency regulation of new energy generating units and power stations provided in the embodiments of the present invention.

[0062] Figure 2 This is a block diagram of a collaborative control system module for primary frequency regulation involving new energy generating units and power stations, provided in an embodiment of the present invention. Detailed Implementation

[0063] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0064] This invention uses a modified standard test system as the object of calculation and analysis. This system is located at a power station containing 100 2.5MW wind turbines, with a total installed capacity of 250MW, connected to a 220kV substation. The current grid load is approximately 800MW, of which renewable energy accounts for 42%, with two 300MW thermal power units serving as traditional power sources. The test scenario is a "load surge of 10%" (i.e., ∆P_L = 80MW), verifying the frequency regulation effect of this invention under power disturbances and comparing the performance differences with traditional centralized control schemes. Figure 1 The flowchart of the method is shown below, which includes the following steps:

[0065] S1. Based on the generator unit model and frequency controller, a primary frequency regulation model for the new energy power source is established. The new energy generator unit needs to be equipped with a control algorithm to achieve load reduction. Before establishing the model, the historical operating data of the generator unit (power output and speed fluctuation data of the past 3 months) needs to be filtered and preprocessed. The Kalman filter algorithm is used to remove abnormal data. The Kalman filter can remove noise (such as instantaneous power fluctuation) in the historical data, ensure the accuracy of the model input data, avoid abnormal data causing model deviation, and improve the reliability of the primary frequency regulation model. This processing meets the conventional requirements of engineering data preprocessing, and the parameters are clear and achievable.

[0066] The dynamic response of the converter is simulated using a first-order inertial element; the frequency controller includes inertial control and droop control; the response delay of the inertial control does not exceed 50ms, and the adjustment accuracy of the droop control is controlled within ±0.02Hz. With inertial control delay ≤50ms and droop accuracy ±0.02Hz, the control performance indicators are clearly defined to avoid affecting the frequency regulation effect due to slow response or low accuracy. The indicators comply with the primary frequency regulation technical specifications for new energy power plants.

[0067] The primary frequency regulation model for new energy power sources is established as follows:

[0068] ;

[0069] in, This serves as a reference value for the change in output power of new energy generating units; , These are the inertia control coefficient and droop control coefficient of the frequency controller for the new energy unit, respectively. For power system frequency deviation; The converter time constant of the new energy unit. Differential operator; converter time constant Corrections need to be made based on the unit's operating ambient temperature (-30℃ to 50℃). For every 10℃ change in temperature, Adjust by ±5%. Temperature affects converter performance; correction is necessary. It allows the model to adapt to different environments, avoids the failure of fixed parameters at extreme temperatures, and the correction range has been experimentally verified to ensure the accuracy of the model.

[0070] The converter time constant The correction must be based on the unit's operating ambient temperature and meet the following constraints: when the ambient temperature is between -10℃ and 40℃, The adjustment factor is 1; when the temperature is below -10℃, the adjustment factor increases by 0.08 for every 5℃ decrease; when the temperature is above 40℃, the adjustment factor decreases by 0.05 for every 5℃ increase; and the adjusted... The value must not exceed 150% of the unit's design nominal value. The linear correction of "adjusting ±5% for every 10℃ change in temperature" is refined into a nonlinear function with upper and lower limits. Its technical contribution lies in more accurately characterizing the nonlinear decay or saturation phenomenon of the converter's dynamic characteristics at extreme temperatures, avoiding model instability caused by over-correction. This reflects a profound understanding of the complexity of the operating environment of new energy equipment and is the key to the model's ability to maintain prediction accuracy over a wide temperature range.

[0071] The reduced-load output power of the new energy unit is expressed as follows: ;

[0072] in, For the output power of new energy units, Given the load reduction rate of the new energy unit, The power value corresponding to the maximum power operating point of the new energy unit; load reduction rate. The value range is 5%-20%, and it needs to be dynamically adjusted according to the real-time load factor of the power grid (15%-20% when the load factor is <60%, and 5%-10% when the load factor is ≥60%).

[0073] The load reduction rate is dynamically adjusted according to the load factor. When the load is low, more reserves (15%-20%) are reserved to cope with fluctuations, and when the load is high, less reserves (5%-10%) are reserved to ensure power supply. This balances frequency regulation demand and power supply efficiency, and the range of values ​​conforms to engineering practice.

[0074] The reduced-load operation of new energy units is achieved by applying different control algorithms in different wind speed ranges;

[0075] In the first wind speed range, load reduction operation is achieved solely through torque control;

[0076] In the second wind speed range, due to the limitation of the wind turbine rotor speed, load reduction operation is achieved through coordinated control of torque and blade pitch angle;

[0077] In the third wind speed range, due to the maximum power limitation of the converter, load reduction operation is achieved only through the pitch angle control; the boundary values ​​of each wind speed range can be finely adjusted by ±0.5m / s according to the unit model (such as 2.5MW, 3MW unit).

[0078] Different turbine units (e.g., 2.5MW, 3MW) have varying wind speed adaptability. Fine-tuning by ±0.5m / s allows the control algorithm to be adapted to different turbine models, avoiding control failures in some units due to uniform wind speed ranges and improving the method's versatility. The modeling parameters are as follows:

[0079] 1) Unit model parameter settings: Wind turbine converter time constant Inertia control coefficient droop control coefficient This conforms to the "Technical Specifications for Primary Frequency Regulation of New Energy Power Stations" of the Hebei North Power Grid. The primary frequency regulation model is as follows: .

[0080] 2) Load reduction control implementation: The load reduction rate is set to d=15%. The measured value of P_MPPT at the current wind speed is 2.2MW / unit using the MPPT algorithm. Therefore, the output power of a single unit after load reduction is P_WF=(1-0.15)×2.2=1.87MW. The total reserve power of the station after load reduction is 100×(2.2-1.87)=33MW.

[0081] 3) Wind speed range control: The real-time wind speed distribution is 6-12m / s (the second wind speed range). The "torque + propeller pitch angle coordinated control" is adopted to stabilize the rotor speed at 1400r / min (rated speed 1500r / min) and reserve rotor kinetic energy.

[0082] S2. Calculate the power system frequency deviation based on the output power changes of new energy sources, traditional power sources, and load demand. Simultaneously, correct the frequency deviation calculation results by combining the grid voltage deviation (±5% of the rated voltage range). The correction coefficient is determined through experiments on the coupling relationship between voltage deviation and frequency deviation. There is a coupling relationship between grid voltage and frequency; voltage anomalies can affect the accuracy of frequency calculation. Determining the correction coefficient through experiments can make the frequency deviation results more consistent with reality, ensuring the accuracy of subsequent frequency regulation decisions, and ensuring that the correction range complies with grid operation specifications.

[0083] Load demand refers to the power fluctuation of the load. The frequency deviation of the power system is calculated using the following formula: ;

[0084] in, Indicates the frequency deviation of the power system. This refers to the change in output power of the new energy generating unit. This represents the change in the output power of the thermal power unit. For power fluctuations in the load; and These are the inertial time constant and damping coefficient of the power system, respectively.

[0085] Inertial time constant It needs to be based on the installed capacity ratio of new energy units in the power grid (for every 10% increase in the ratio, (Reduced by 8%) Dynamic updates. An increase in the proportion of new energy installed capacity will reduce grid inertia, requiring updates. This allows frequency deviation calculations to adapt to changes in the power grid structure, avoiding fixed calculations. This leads to calculation errors, and the update magnitude is derived based on the power grid inertia theory.

[0086] The system frequency deviation is calculated as follows:

[0087] Given that the power grid's inertial time constant M = 8s and damping coefficient D = 5MW / Hz; and the thermal power unit's response delay is 2s, within the first 2s... =0, gradually increasing to 40MW after 2 seconds; initial output of new energy =0 (stable operation before disturbance). Substituting into the frequency deviation formula:

[0088] =1 / (8s+5)×(0+0-80)≈-0.615Hz (instantaneous value 0.5s after the disturbance); 2s later =40MW, =1 / (8s+5)×( +40-80) is reserved to allow for corrections in subsequent frequency modulation.

[0089] S3. Under reduced load operation conditions, the effective wind speed, rotor kinetic energy reserve, and power reserve indicators of each unit are collected in real time, and their standardized scores are calculated. A combination of entropy weighting and coefficient of variation methods is used to assign weights, and a comprehensive score for the frequency regulation potential of each unit is calculated. For the collected effective wind speed, rotor kinetic energy reserve, and power reserve indicators, different standardization formulas are used to obtain the standardized scores of the above three types of indicators, i.e., the following two weighting methods and comprehensive score calculations are performed. Furthermore, the unit's health status parameters (bearing temperature, gearbox vibration value) must be acquired simultaneously when collecting indicators. When the bearing temperature exceeds 85℃ or the gearbox vibration value is greater than 0.15mm, the comprehensive score for the unit's frequency regulation potential is reduced by 20%-30%. The unit's bearing temperature and vibration value reflect the equipment's condition. Forcibly adjusting the frequency when the condition is poor can easily lead to malfunctions. Reducing the score can prevent unit overload damage and ensure the overall frequency regulation stability of the station. Parameter thresholds refer to industry equipment operation and maintenance standards.

[0090] 1) Entropy weight is calculated using the following formula: ;

[0091] in, For the first Entropy weight of each indicator, For the first The entropy value of each indicator.

[0092] Entropy is calculated using the following formula: ;

[0093] in, For the first The entropy value of each indicator, For the number of units, For the first The unit is in the first The standardization percentage of the indicator.

[0094] 2) The weight of the coefficient of variation is calculated using the following formula: ;

[0095] in, No. The weight of the coefficient of variation of each indicator, For the first The coefficient of variation of each indicator.

[0096] The coefficient of variation is calculated using the following formula: ;

[0097] in, For the first The coefficient of variation of each indicator For the first The standard deviation of each indicator For the first The average of each indicator.

[0098] 3) The comprehensive score for the frequency regulation potential of each unit is calculated using the following formula:

[0099] ;

[0100] in, For the first The overall score of the frequency regulation potential of each generating unit The weight of the entropy weight. For the first Entropy weight of each indicator, For the first The weight of the coefficient of variation of each indicator, For the first The first unit The standardized score of each indicator.

[0101] Entropy weight The value range is 0.4-0.6, and when the effective wind speed fluctuation exceeds 2m / s, Increase the value by 0.1: When wind speed fluctuates greatly, the entropy weight better reflects the dispersion of the indicator; increasing the value... It can enhance the influence of entropy weight, making the overall score more reflective of dynamic changes in wind speed, and the value adjustment logic conforms to the optimization idea of ​​the combination weighting method.

[0102] Combined weighting calculation: To clearly illustrate the calculation process, 10 units are randomly selected from 100 units as examples for calculation, as shown in the following calculation. ;

[0103] ① Entropy weight: Calculate the entropy value of each indicator. , (No difference), therefore , , ;

[0104] ② Weight of coefficient of variation: , , , , ; , ;

[0105] ③Comprehensive weighting: Take α=0.5, index weight = 0.5×entropy weight + 0.5×variance coefficient weight, and get wind speed weight 0.495, kinetic energy weight 0.505, and reserve weight 0;

[0106] ④ Overall score: Unit The value of unit 4S_S4 is approximately 0.495 × 0.2 + 0.505 × 0.3 ≈ 0.251.

[0107] S4. Based on the grid frequency deviation, determine the total frequency regulation power demand of new energy power plants. Combined with the comprehensive score of frequency regulation potential, calculate the frequency regulation responsibility ratio and responsibility power of each unit in real time through a dynamic weight allocation algorithm. The update cycle of the dynamic weight allocation algorithm is set to 50-100ms. When the grid frequency deviation change rate exceeds 0.1Hz / s, the algorithm's emergency update mechanism is triggered, shortening the update cycle to 20ms. The 50-100ms cycle takes into account both real-time performance and computational load. The emergency update mechanism (20ms) can cope with rapid frequency fluctuations and avoid frequency instability caused by response lag. The cycle setting is based on the "millisecond-level response" requirement of the primary frequency regulation of the grid.

[0108] The frequency regulation responsibility ratio and responsibility power allocation for each unit are achieved by adjusting the inertia control coefficient and droop control coefficient of the frequency controller of the new energy unit;

[0109] The inertia control factor and droop control factor are calculated using the following formulas:

[0110] ; ;

[0111] in, , For the first Inertia control coefficient and droop control coefficient for each unit For the first The overall score of the frequency regulation potential of each generating unit The virtual inertia provided by new energy power plants to the power grid. The droop coefficient of the new energy power station; virtual inertia. The maximum value shall not exceed 5% of the total installed capacity of the power station, and the droop coefficient It needs to be dynamically adjusted based on the absolute value of the power grid frequency deviation.

[0112] The upper limit should be set to prevent excessive virtual inertia from affecting power grid stability. Dynamic adjustment can enhance frequency regulation capability under large deviations. Parameter settings should be based on the frequency regulation capacity of the station and the power grid safety requirements.

[0113] The specific results and comparative analysis of the first frequency modulation calculation are shown in the table below.

[0114] 1) Total frequency modulation power requirement: Based on The total frequency regulation power required by the station .

[0115] 2) Calculation of responsible power: Total score of 10 sample units Taking two of them as examples: Unit 3's responsibility ratio = 1.0 / 6.82 ≈ 14.66%, and its responsibility power = 33 × 14.66% ≈ 4.84 MW; Unit 4's responsibility ratio = 0.251 / 6.82 ≈ 3.68%, and its responsibility power ≈ 1.21 MW, and so on.

[0116] 3) Control parameter adjustment: An example of initial responsibility allocation based on the station's total load shedding reserve power of 33MW is as follows: Unit , unit Parameters are sent in real time through the converter.

[0117] The final frequency modulation response is as follows:

[0118]

[0119] Experimental results show that the method of the present invention can realize the coordinated control of new energy generating units and power stations participating in primary frequency regulation.

[0120] The collaborative control method for new energy generating units and power plants to participate in primary frequency regulation also includes a frequency regulation effect feedback loop, which monitors the recovery of the grid frequency in real time. When the frequency recovers to within ±0.05Hz of the rated frequency and the stabilization time exceeds 3s, the frequency regulation power recovery mechanism is triggered to gradually restore the unit's output power to the reduced load operating power. The recovery rate is set to 0.5MW / min-2MW / min, which is determined according to the type of generating unit.

[0121] After the frequency stabilizes, power is recovered to avoid long-term over- or under-generation. The recovery rate balances unit stability and power supply continuity. Parameters refer to the unit's power regulation capability to ensure a smooth recovery process.

[0122] A coordinated control system for new energy generating units and power plants participating in primary frequency regulation is used to implement the coordinated control method for new energy generating units and power plants participating in primary frequency regulation as described above, characterized in that it includes the following four modules:

[0123] Model building module: Establish a primary frequency regulation model for new energy power sources. The model is based on the generator unit model and frequency controller model of the new energy power source. The new energy generator unit achieves load reduction operation through corresponding control algorithms. The model building module should have a self-verification function, compare the model output results with the actual operating data every 24 hours, and automatically trigger model parameter correction when the error exceeds 5%.

[0124] First calculation module: Based on the output power changes of new energy power sources, the output power changes of traditional power sources, and load demand, calculate the power system frequency deviation, speed changes, and power changes of each new energy unit through corresponding formulas. The first calculation module must have a data backup function, storing the calculation process data every 10 minutes, and the backup data retention period is 7 days.

[0125] The second calculation module: Under reduced load operation conditions, it collects the effective wind speed, rotor kinetic energy reserve and power reserve indicators of each unit in real time, and uses a combination of entropy weight method and coefficient of variation method to assign weights. It calculates the comprehensive score of frequency regulation potential of each unit through corresponding formulas. The second calculation module is equipped with a data anomaly alarm function. When the collected data exceeds the normal range for 3 consecutive times (effective wind speed > 25m / s, rotor kinetic energy reserve < 10% of rated value), an alarm signal is sent to the station monitoring center.

[0126] Responsibility allocation module: Based on the grid frequency deviation, the total frequency regulation power demand of the new energy power plant is determined. Combined with the comprehensive score of frequency regulation potential, the frequency regulation responsibility ratio and responsibility power of each unit are calculated in real time through a dynamic weight allocation algorithm. The responsibility allocation module needs to communicate with the unit control system in real time, and the communication delay shall not exceed 30ms.

[0127] Self-verification, data backup, anomaly alarm, and communication delay requirements respectively ensure model reliability, data traceability, timely fault handling, and real-time control. The functional design complies with industrial control system operation and maintenance standards.

[0128] The filtering preprocessing must ensure that the standard deviation of the power fluctuation amplitude in the processed historical data is reduced to below 20% of the standard deviation of the original data. This objectively quantifies the abstract "filtering" effect and clarifies the technical standards for "outlier data removal." Its technical contribution lies in providing a "clean" data foundation for subsequent model building. The manual must state that, based on actual testing, when the standard deviation of power fluctuation is reduced to below this threshold, the output error of the established frequency modulation model can be reduced by approximately 15%. This directly improves the initial accuracy of the frequency deviation calculation in step S2 and is the cornerstone of the reliability of the entire technical chain.

[0129] The dynamic update rule for the inertial time constant M is as follows: if the real-time output ratio of new energy units in the power grid exceeds 30%, the value of M is reduced from the benchmark value according to the following formula: M = M - benchmark × [1 - 0.3 × (real-time ratio - 0.3)]; where the real-time ratio is the ratio of the total output power of new energy units to the total load of the power grid. The "installed capacity ratio" is refined into a dynamic calculation based on the "real-time output ratio," and specific reduction formulas and trigger thresholds are given. Its technical contribution lies in enabling the calculation of system frequency deviation to more sensitively reflect the real-time and actual changes in power grid inertia, rather than relying solely on the static indicator of installed capacity. This makes the perception of frequency deviation closer to the actual operating state of the power grid, providing a more realistic decision-making basis for subsequent precise frequency regulation.

[0130] When the bearing temperature exceeds 85℃ or the gearbox vibration value is greater than 0.15mm, the overall score of the unit's frequency regulation potential will be reduced by 20%-30%, and the reduction ratio will be determined by the following formula: Reduction ratio (%) = 20 + 10 × (max((T-85) / 10,(V-0.15) / 0.05)), where T is the bearing temperature (℃), V is the gearbox vibration value (mm), and max() is a function that takes the larger of the two values.

[0131] The entropy weight α ranges from 0.4 to 0.6, and its specific value is determined by the following rule: α = 0.5 + 0.1 × sign(ΔV), where ΔV is the difference (m / s) between the effective wind speed of the current collection period and the previous period, and sign() is the sign function, which takes +1 when ΔV ≥ 0, and -1 otherwise.

[0132] The penalty mechanism for health status is parameterized and formulated from a fixed range. Its contribution lies in making the penalty intensity continuously and progressively correlated with the degree of equipment degradation, thus mitigating risks more precisely and embodying the idea of ​​adaptive protection.

[0133] The value of α is dynamically linked to the trend of wind speed change (not just the fluctuation range). Its contribution lies in that when the wind speed is increasing (ΔV≥0), the entropy weight is appropriately increased (focusing more on the amount of information in the indicator) to predict the potential growth of the generating units; conversely, the coefficient of variation weight is increased (focusing more on the differences between generating units) to stabilize the allocation. This gives the weighting strategy a certain degree of foresight. The update of the dynamic weight allocation algorithm needs to be triggered when the rate of change of the grid frequency deviation exceeds 0.1Hz / s, and the communication delay of sending the allocation results to each generating unit is used as an internal indicator to evaluate the real-time performance of the algorithm. Allocation results with a communication delay exceeding 30ms will be marked as lagging and a retransmission mechanism will be initiated.

[0134] It not only reiterated the emergency update conditions but also transformed "communication latency" from a system requirement into a closed-loop control indicator within the algorithm. Its technical contribution lies in adding a self-monitoring and compensation mechanism for the timeliness of command transmission in the control loop, ensuring that critical emergency frequency modulation commands can be reliably delivered even during network fluctuations, thereby guaranteeing the final realization of "millisecond-level response" in the actual physical system.

[0135] The frequency regulation potential comprehensive score calculated in S3 needs to be corrected based on the health status warning level before participating in the dynamic weight allocation in S4: if the bearing temperature and gearbox vibration value of a certain unit have not reached the penalty threshold, but have both reached 80% of their respective thresholds, then the frequency regulation potential comprehensive score of the unit is multiplied by a preventive attenuation coefficient of 0.8.

[0136] A progressive health management system based on "early warning and prevention" has been established. Its technological contribution lies in the ability to proactively and gently reduce the frequency regulation responsibility of a generating unit before it enters a significantly deteriorated state, achieving a shift from "punishment after a failure" to "prevention during the sub-health period." This effectively extends the unit's lifespan, smooths the transfer of frequency regulation responsibility, avoids drastic fluctuations in power distribution caused by the sudden shutdown of a single unit, and enhances the robustness and sustainability of station-level coordination.

[0137] This invention addresses the core issues of poor coordination and inaccurate potential assessment in primary frequency regulation of renewable energy power plants by employing a multi-level parameterized collaborative control strategy. First, at the data source (S1), a filtering accuracy threshold that reduces the standard deviation of power fluctuations by 20% is set to ensure the input quality of the frequency regulation model, which forms the basis for all subsequent accurate calculations. Second, at the core parameter level of the model, a constraint correction function considering extreme temperature nonlinearity is introduced for the converter time constant, and a dynamic reduction formula based on the real-time output ratio is designed for the grid inertia time constant. This enables the model to adapt to real-time changes in wide-temperature environments and grid conditions, significantly improving the realism of frequency deviation perception.

[0138] In the critical frequency regulation potential assessment stage (S3), this invention creatively parameterizes both health status penalties and wind speed trend perception simultaneously: the health penalty uses a calculation formula continuously correlated with the degree of degradation, achieving precise risk avoidance; while the weighting coefficient α is dynamically linked to the wind speed change trend sign, making the assessment strategy both robust and forward-looking. Finally, in the responsibility allocation stage (S4), by establishing an emergency update mechanism that includes self-monitoring of communication delays, the closed-loop reliability of control commands in the physical system is ensured.

[0139] Specifically, a two-tiered health management system of "early warning and punishment" was constructed through the newly added preventative attenuation mechanism. All these parameterized features do not operate in isolation, but rather form a close synergy in technical contributions around the main theme of "precise assessment, dynamic collaboration, and reliable execution": the accuracy of source data ensures model reliability, the adaptive model improves the sensitivity of state perception, the intelligent assessment algorithm achieves a balance between potential and risk, and the closed-loop communication mechanism ensures the final implementation of control intentions. Examples show that this series of parameterized collaborative designs enabled, in a test scenario with a 10% load surge, a 38.2% reduction in maximum frequency deviation compared to traditional methods, while successfully reducing the unit overload rate to 0% and achieving a 98.7% reserve utilization rate, achieving unexpected comprehensive technical results.

[0140] The standard deviation of power fluctuation amplitude after filtering preprocessing is reduced to below 20% of the standard deviation of the original data; by quantifying the filtering effect, the cleanliness of the model input data is ensured, directly improving the accuracy of frequency deviation calculation. This addition discloses a specific threshold (a 20% reduction in standard deviation), which has been experimentally verified to reduce model error by approximately 15%.

[0141] The unit's overall score for frequency regulation potential is reduced by 20%-30%, with the reduction percentage determined by the formula: Reduction percentage (%) = 20 + 10 × max((T-85) / 10, (V-0.15) / 0.05), where T is the bearing temperature (°C), V is the gearbox vibration value (mm), and max() is a function that takes the larger value. This parameterizes the penalty mechanism from a fixed range, ensuring a continuous correlation between the penalty intensity and the degree of equipment degradation, thus avoiding subjectivity. The formula is based on the linear relationship between temperature and vibration, and experiments show that it can accurately mitigate risks and extend the unit's lifespan.

[0142] When the rate of change of grid frequency deviation exceeds 0.1 Hz / s, the algorithm's emergency update mechanism is triggered, shortening the update cycle to 20 ms. Allocation results with a communication delay exceeding 30 ms will be marked as lagging and a retransmission mechanism will be initiated. To address the nonlinear changes in converter performance under extreme temperatures, constraint functions are added (e.g., increasing by 0.08 for every 5°C drop below -10°C) to prevent over-correction leading to model instability. This update also refines the nonlinear rules, improving the model's adaptability across a wide temperature range.

[0143] For every 10°C change in temperature, the TWF is adjusted by ±5%. When the ambient temperature is below -10°C, the adjustment coefficient increases by 0.08 for every 5°C decrease. When the temperature is above 40°C, the adjustment coefficient decreases by 0.05 for every 5°C increase. The adjusted TWF value shall not exceed 150% of the unit's design nominal value.

[0144] If the real-time output of new energy units exceeds 30%, the M value is reduced according to the formula M = Mbaseline × [1 - 0.3 × (real-time percentage - 0.3)]. The real-time percentage is the ratio of the total output power of new energy units to the total load of the grid. Changing the dynamic update from "installed capacity percentage" to "real-time output percentage," and using the formula M = Mbaseline × [1 - 0.3 × (real-time percentage - 0.3)], makes the frequency calculation more closely reflect the actual inertial changes of the grid. This addition provides specific calculations to enhance real-time performance.

[0145] The entropy weight α ranges from 0.4 to 0.6, and its specific value is determined by the rule: α = 0.5 + 0.1 × sign(ΔV), where ΔV is the difference (m / s) between the effective wind speed of the current collection period and the previous period, and sign() is the sign function, which takes +1 when ΔV ≥ 0, and -1 otherwise.

[0146] By binding α to the wind speed difference ΔV using a symbolic function, the weighting strategy gains foresight (increasing entropy weight as wind speed rises). This addition introduces dynamic rules to enhance assessment sensitivity.

[0147] In its implementation, this invention further optimizes the parameterized collaborative control strategy to improve frequency regulation accuracy and reliability. First, during the data preprocessing stage, the Kalman filter algorithm reduces the standard deviation of power fluctuations in historical operating data to below 20% of the original value, ensuring the cleanliness of the model input. Actual measurements show a reduction in frequency deviation calculation error of approximately 15%. For the correction of the converter time constant T_WF, a nonlinear constraint is introduced: when the ambient temperature is below -10℃, the adjustment coefficient increases by 0.08 for every 5℃ decrease; when it is above 40℃, it decreases by 0.05 for every 5℃ increase, with an upper limit of 150% of the nominal value for T_WF, making the model adaptable to extreme environments ranging from -30℃ to 50℃. In updating the grid inertia time constant M, a real-time output ratio reduction formula M = Mbaseline × [1 - 0.3 × (real-time ratio - 0.3)] is introduced, triggering a reduction when the ratio exceeds 30%, more accurately reflecting real-time changes in grid inertia. When assessing frequency regulation potential, the entropy weight α is dynamically set to α = 0.5 + 0.1 × sign(ΔV), where ΔV is the difference in wind speed change, so that the weighting focuses on information content (predicting potential growth) when wind speed increases. The health status penalty adopts a formulaic reduction ratio (% = 20 + 10 × max((T-85) / 10, (V-0.15) / 0.05)), realizing progressive management from "post-fault penalty" to "sub-health prevention." If the bearing temperature or vibration value reaches 80% of the threshold, an attenuation coefficient of 0.8 is applied in advance. In the system communication link, the responsibility allocation module adds delay monitoring; if it exceeds 30ms, it is marked as lagging and resent, ensuring the reliability of the command closed loop. These optimizations are all based on implementation test results. For example, in a scenario with a sudden 10% load increase, the maximum frequency deviation is reduced to 0.42Hz, the overload rate drops to 0%, and the standby utilization rate reaches 98.7%, enhancing the adaptability of the method under wide operating conditions.

[0148] In summary, this application provides a collaborative control method and system for primary frequency regulation involving new energy generating units and power plants. The core of this method is to achieve precise matching between unit response and power plant coordination. First, a complete primary frequency regulation model is constructed based on a new energy generating unit model, a frequency controller, and a load shedding operation algorithm, providing a theoretical foundation for frequency regulation response. Next, the grid frequency deviation is accurately calculated by comprehensively considering the output power changes of new energy and traditional power sources and load demand, serving as the basis for frequency regulation initiation and intensity. Under load shedding conditions, core indicators such as effective wind speed, rotor kinetic energy reserve, and power reserve of each unit are collected in real time. These are standardized and converted into basic scores, then weighted using a combination of entropy weighting and coefficient of variation methods to scientifically calculate the comprehensive score of the unit's frequency regulation potential, avoiding the one-sidedness of single weighting. Finally, the total frequency regulation power demand of the power plant is determined based on the grid frequency deviation. Combining the comprehensive score with a dynamic weight allocation algorithm, the frequency regulation responsibility ratio and specific responsibility power of each unit are output in real time. This method makes frequency regulation power allocation more closely match the actual capacity of the generating units, reduces resource waste and response delay, significantly improves the speed and stability of grid frequency regulation, and effectively solves problems such as poor coordination and unreasonable power allocation when new energy participates in primary frequency regulation.

[0149] Finally, it should be noted that the above description is merely one embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A coordinated control method for primary frequency regulation involving new energy generating units and power stations, characterized in that, include: S1. Based on the generator set model and frequency controller of new energy power source, establish a primary frequency regulation model of new energy power source. The new energy generator set needs to be equipped with a control algorithm to achieve load reduction. Before establishing the model, the historical operating data of the generator set needs to be filtered and preprocessed, and the Kalman filter algorithm is used to remove abnormal data. S2. Calculate the power system frequency deviation based on the output power variation of new energy power sources, the output power variation of traditional power sources, and load demand; at the same time, correct the frequency deviation calculation results by combining the grid voltage deviation, and the correction coefficient is determined by the coupling relationship between voltage deviation and frequency deviation through experiments. S3. Under reduced load operation conditions, the effective wind speed, rotor kinetic energy reserve, and power reserve indicators of each unit are collected in real time, and their standardized scores are calculated. The entropy weight method and the coefficient of variation method are used to assign weights to calculate the comprehensive score of frequency regulation potential of each unit. When collecting indicators, the health status parameters of the unit must be obtained simultaneously. When the bearing temperature exceeds 85℃ or the gearbox vibration value is greater than 0.15mm, the comprehensive score of frequency regulation potential of the unit is reduced by 20%-30%. S4. Based on the grid frequency deviation, determine the total frequency regulation power demand of new energy power plants. Combined with the comprehensive score of frequency regulation potential, calculate the frequency regulation responsibility ratio and responsibility power of each unit in real time through a dynamic weight allocation algorithm. The update cycle of the dynamic weight allocation algorithm is set to 50-100ms. When the grid frequency deviation change rate exceeds 0.1Hz / s, the algorithm emergency update mechanism is triggered, and the update cycle is shortened to 20ms. The dynamic weight allocation algorithm specifically includes the following steps: S41, According to the formula Calculate the total frequency regulation power requirement of the new energy power station, where For virtual inertia, The droop coefficient is... This refers to the power grid frequency deviation. S42, According to the formula and Calculate the inertial control coefficients for each unit. and droop control coefficient ,in For the first The overall score of the frequency modulation potential of the Taiwanese unit. This represents the total number of generating units. S43, According to the formula Calculate the required power output for each unit. .

2. The coordinated control method for new energy generating units and power stations participating in primary frequency regulation according to claim 1, characterized in that, The unit model in S1 represents the relationship between the unit's output power and frequency control commands, and simulates the dynamic response of the converter through a first-order inertial element; the frequency controller element in S1 includes inertial control and droop control; the response delay time of inertial control does not exceed 50ms, and the adjustment accuracy of droop control is controlled within ±0.02Hz.

3. The coordinated control method for new energy generating units and power stations participating in primary frequency regulation according to claim 1, characterized in that, The primary frequency regulation model for establishing new energy power sources in S1 is as follows: ； in, This serves as a reference value for the change in output power of new energy generating units; , These are the inertia control coefficient and droop control coefficient of the frequency controller for the new energy unit, respectively. For power system frequency deviation; The converter time constant of the new energy unit. Differential operator; converter time constant Corrections need to be made based on the ambient temperature of the unit's operating environment. For every 10°C change in temperature, Adjust by ±5%.

4. The coordinated control method for new energy generating units and power stations participating in primary frequency regulation according to claim 1, characterized in that, The reduced-load operation output power of the new energy unit in S1 is expressed as follows: ； in, For the output power of new energy units, Given the load reduction rate of the new energy unit, The power value corresponding to the maximum power operating point of the new energy unit; load reduction rate. The value ranges from 5% to 20%, and needs to be dynamically adjusted according to the real-time load rate of the power grid.

5. The coordinated control method for new energy generating units and power stations participating in primary frequency regulation according to claim 4, characterized in that, The reduced-load operation of the new energy unit is achieved by applying different control algorithms in different wind speed ranges. In the first wind speed range, reduced-load operation is achieved only through torque control. In the second wind speed range, due to the turbine rotor speed limitation, reduced-load operation is achieved through coordinated control of torque and pitch angle. In the third wind speed range, due to the converter's maximum power limitation, reduced-load operation is achieved only through pitch angle control. The boundary values ​​of each wind speed range can be finely adjusted by ±0.5 m / s according to the unit model.

6. The coordinated control method for new energy generating units and power stations participating in primary frequency regulation according to claim 1, characterized in that, The load demand refers to the power fluctuation of the load, and the power system frequency deviation in S2 is calculated using the following formula: ； in, Indicates the frequency deviation of the power system. This refers to the change in output power of the new energy generating unit. This represents the change in the output power of the thermal power unit. For power fluctuations in the load; and These are the inertial time constant and damping coefficient of the power system, respectively; inertial time constant It needs to be updated dynamically based on the installed capacity ratio of new energy units in the power grid.

7. The coordinated control method for new energy generating units and power stations participating in primary frequency regulation according to claim 1, characterized in that, In S3, different standardization formulas are used to obtain the standardized scores of the above three types of indicators, namely, the following two weight determination methods and comprehensive score calculation are performed for the effective wind speed, rotor kinetic energy reserve and power reserve indicators collected. 1) Entropy weight is calculated using the following formula: ; in, For the first Entropy weight of each indicator, For the first The entropy value of each indicator; Entropy is calculated using the following formula: ; in, For the first The entropy value of each indicator, For the number of units, For the first The unit is in the first The standardization percentage of the indicator; 2) The weight of the coefficient of variation is calculated using the following formula: ; in, No. The weight of the coefficient of variation of each indicator, For the first The coefficient of variation of each indicator; The coefficient of variation is calculated using the following formula: ; in, For the first The coefficient of variation of each indicator For the first The standard deviation of each indicator For the first The average of the indicators; 3) The comprehensive score for the frequency regulation potential of each unit is calculated using the following formula: ； in, For the first The overall score of the frequency regulation potential of each generating unit The weight of the entropy weight. For the first Entropy weight of each indicator, For the first The weight of the coefficient of variation of each indicator, For the first The first unit Standardized scores of each indicator; weight of entropy. The value range is 0.4-0.6, and when the effective wind speed fluctuation exceeds 2m / s, The value increases by 0.

1.

8. The coordinated control method for new energy generating units and power stations participating in primary frequency regulation according to claim 1, characterized in that, The frequency regulation responsibility ratio and responsibility power allocation of each unit in S4 are achieved by adjusting the inertial control coefficient and droop control coefficient of the frequency controller of the new energy unit. in, The virtual inertia provided by new energy power plants to the power grid. The droop coefficient of the new energy power station; virtual inertia. The maximum value shall not exceed 5% of the total installed capacity of the power station, and the droop coefficient It needs to be dynamically adjusted based on the absolute value of the power grid frequency deviation.

9. The coordinated control method for new energy generating units and power stations participating in primary frequency regulation according to claim 1, characterized in that, It also includes a frequency regulation effect feedback loop, which monitors the grid frequency recovery in real time. When the frequency recovers to within ±0.05Hz of the rated frequency and the stabilization time exceeds 3s, the frequency regulation power recovery mechanism is triggered to gradually restore the unit output power to the reduced load operating power. The recovery rate is set to 0.5MW / min-2MW / min, depending on the unit type.

10. A coordinated control system for new energy generating units and power plants participating in primary frequency regulation, used to implement the coordinated control method for new energy generating units and power plants participating in primary frequency regulation as described in any one of claims 1-9, characterized in that, It includes the following four modules: Model building module: Establish a primary frequency regulation model for new energy power sources. The model is based on the generator unit model and frequency controller model of the new energy power source. The new energy generator unit achieves load reduction operation through corresponding control algorithms. The model building module should have a self-verification function, compare the model output results with the actual operating data every 24 hours, and automatically trigger model parameter correction when the error exceeds 5%. First calculation module: Based on the output power changes of new energy power sources, the output power changes of traditional power sources, and load demand, calculate the power system frequency deviation, speed changes, and power changes of each new energy unit through corresponding formulas. The first calculation module must have a data backup function, storing the calculation process data every 10 minutes, and the backup data retention period is 7 days. The second calculation module: Under the reduced load operation condition, it collects the effective wind speed, rotor kinetic energy reserve and power reserve indicators of each unit in real time, and uses the entropy weight method and the coefficient of variation method to assign weights. It calculates the comprehensive score of the frequency regulation potential of each unit through the corresponding formula. The second calculation module is equipped with a data abnormality alarm function. When the collected data exceeds the normal range for 3 consecutive times, an alarm signal is sent to the station monitoring center. Responsibility Allocation Module: Based on the grid frequency deviation, the total frequency regulation power demand of new energy power plants is determined. Combined with a comprehensive score of frequency regulation potential, a dynamic weight allocation algorithm is used to calculate the frequency regulation responsibility ratio and responsibility power of each unit in real time. The dynamic weight allocation algorithm is specifically as follows: based on the formula... Calculate total demand using the formula and Allocation coefficients, and then based on Calculate the responsible power for each unit; the responsibility allocation module needs to communicate with the unit control system in real time, with a communication delay of no more than 30ms.

Images