A new energy power station virtual inertia adaptive control method

Abstract

This invention discloses a virtual inertia adaptive control method for new energy power plants, belonging to the field of power system control technology. It includes acquiring the local operating status parameters of each power generation unit in the new energy power plant and the grid parameters at the grid connection point; calculating the unit's operational health based on the local operating status parameters; monitoring the frequency change rate at the grid connection point in real time, and triggering a local virtual inertia response when the absolute value of the frequency change rate exceeds a fast response threshold dynamically generated based on the unit's operational health; calculating the grid interaction strength index based on the grid parameters; integrating the unit's operational health and the grid interaction strength index to generate a safe adjustment range for the virtual inertia parameters; distributing the safe adjustment range to each power generation unit for local parameter fine-tuning; and generating the final control command. This invention uses unit operational health and the power plant-level grid interaction strength index, combined with a dynamic decoupling optimization algorithm, to achieve adaptive adjustment of the virtual inertia parameters.

Description

Technical Field

[0001] This invention relates to the field of power system control technology, and in particular to a virtual inertia adaptive control method for new energy power plants. Background Technology

[0002] As renewable energy accounts for an increasingly larger share of the global energy mix, new energy power plants, represented by wind and solar power, are being integrated into the grid on a large scale. These renewable energy sources are connected to the grid via power electronic inverters. Since they lack the rotational inertia of traditional synchronous generators, this leads to a decrease in the overall inertia level of the power system, weakened disturbance immunity, and a severe challenge to frequency stability. To address this issue, virtual inertia control technology has emerged. By simulating the inertial response characteristics of synchronous generators, it enables inverters to actively provide or absorb power when grid frequency disturbances occur, thus providing frequency support for the grid.

[0003] In existing technologies, most virtual inertia control methods for new energy power plants employ fixed-parameter control strategies. The control system typically pre-sets one or a set of fixed virtual inertia coefficients and damping coefficients based on offline simulation analysis or typical grid operating conditions. When the frequency change rate at the grid connection point exceeds a fixed trigger threshold, the inverter adjusts its power according to these preset parameters to provide a virtual inertia response. This control method can provide a certain degree of frequency support under specific design conditions.

[0004] However, the aforementioned existing technical solutions have certain limitations. First, the operating state of the power grid is dynamic, and its strength and stability margin are not constant. Using fixed parameters cannot adapt to changes in the strength of the power grid, and may result in insufficient support when the grid is weak, or unnecessary power fluctuations when the grid is strong. Second, the operating states of each power generation unit within the new energy power plant, such as equipment temperature, DC-side energy reserves, and the degree of device aging, also change in real time. Fixed parameter control strategies ignore the instantaneous adjustment capabilities and safety margins of the equipment itself, and may cause excessive adjustment shocks to equipment in poor condition, affecting its service life and operational safety. Summary of the Invention

[0005] To address the aforementioned issues, this invention provides a virtual inertia adaptive control method for new energy power plants. By employing unit operational health and power plant-level grid interaction intensity index, combined with a dynamic decoupling optimization algorithm, the method can achieve adaptive adjustment of virtual inertia parameters.

[0006] The above objectives can be achieved through the following approach:

[0007] A virtual inertia adaptive control method for a new energy power plant includes: acquiring the local operating status parameters of each power generation unit in the new energy power plant and the grid parameters of the grid connection point; calculating the unit's operating health based on the local operating status parameters; monitoring the frequency change rate of the grid connection point in real time, and triggering a local virtual inertia response when the absolute value of the frequency change rate exceeds the fast response threshold dynamically generated based on the unit's operating health; calculating the grid interaction strength index based on the grid parameters; integrating the unit's operating health and the grid interaction strength index to generate a parameter safety adjustment range for the virtual inertia; distributing the parameter safety adjustment range to each power generation unit for local parameter fine-tuning, and generating the final control command.

[0008] Optionally, the step of calculating the unit operation health of each power generation unit based on the local operation status parameters includes: extracting multiple health indicator parameters from the local operation status parameters; determining the weight coefficient of each health indicator parameter according to its influence on the equipment regulation capability, and weighting and fusing the multiple health indicator parameters to generate a health score; obtaining the equipment safe operation boundary, and correcting the health score to obtain the unit operation health of each power generation unit.

[0009] Optionally, the fast response threshold dynamically generated based on the unit's operational health includes: determining a dynamic adjustment coefficient that is inversely proportional to the unit's operational health; obtaining a benchmark threshold characterizing the normal operating conditions of the power grid; and multiplying the dynamic adjustment coefficient by the benchmark threshold to calculate the fast response threshold.

[0010] Optionally, the step of calculating the power plant-level grid interaction strength index based on the grid parameters includes: extracting frequency stability parameters and voltage stability parameters from the grid parameters; performing time series trend analysis on the frequency stability parameters and voltage stability parameters to generate grid strength trend values; and mapping the grid strength trend values ​​to an index to calculate the power plant-level grid interaction strength index.

[0011] Optionally, the step of integrating the unit's operational health and the power plant-level grid interaction strength index to generate the parameter safety adjustment range of the virtual inertia through a preset dynamic decoupling optimization algorithm includes: decoupling the unit's operational health and the power plant-level grid interaction strength index into equipment regulation capability factors and grid demand factors to obtain independent influencing factors; obtaining optimized control parameters based on the independent influencing factors through a preset dynamic decoupling optimization algorithm; obtaining equipment physical characteristic safety constraints, and defining the parameter safety adjustment range of the virtual inertia around the optimized control parameters.

[0012] Optionally, each power generation unit performs local parameter fine-tuning within the parameter safety adjustment range based on its own unit operating health, and generates the final control command by: calculating the local adjustment weight based on the unit operating health; applying the local adjustment weight to perform interpolation calculation between the upper and lower limits of the parameter safety adjustment range to determine the virtual inertia parameter; and generating the final control command based on the virtual inertia parameter.

[0013] Optionally, the method further includes: collecting historical local operating status parameters, grid parameters at the grid connection point, final control commands, and grid frequency feedback signals after executing control commands to form a historical operating dataset; updating the calculation model for calculating the unit's operating health and the algorithm for calculating the power plant-level grid interaction strength index based on the historical operating dataset; and optimizing the fast response threshold and the safety adjustment range by applying the updated calculation model and algorithm.

[0014] Optionally, the step of decoupling the unit operating health and the power plant-level grid interaction intensity index into an equipment regulation capability factor and a grid demand factor to obtain independent influencing factors includes: constructing a joint feature vector from the unit operating health and the power plant-level grid interaction intensity index; processing the joint feature vector and projecting it onto a mutually orthogonal feature space; and separating the equipment regulation capability factor, which characterizes the intrinsic capabilities of the equipment, and the grid demand factor, which characterizes the external demands of the grid, in the feature space to obtain independent influencing factors.

[0015] Optionally, calculating the local adjustment weight based on the unit's operational health includes: quantifying the unit's operational health into an adjustable margin for the power generation unit; and calculating the local adjustment weight based on the adjustable margin.

[0016] Based on the same inventive concept, this invention also provides a virtual inertia adaptive control system for a new energy power plant. The system includes: a data acquisition module for acquiring local operating status parameters of each power generation unit in the new energy power plant and grid parameters of the grid connection point; a unit health assessment module for calculating the unit operating health of each power generation unit based on the local operating status parameters, wherein the unit operating health characterizes the instantaneous adjustment capability and safe operating boundary of the power generation unit; and a virtual inertia triggering module for real-time monitoring of the frequency change rate at the grid connection point. When the absolute value of the frequency change rate exceeds a fast response threshold dynamically generated based on the unit operating health, a triggering mechanism is activated. The system triggers a local virtual inertia response; a power grid strength assessment module calculates a power plant-level power grid interaction strength index based on the power grid parameters, the power plant-level power grid interaction strength index characterizing the trend of power grid strength change at the grid connection point; a dynamic decoupling optimization module integrates the unit's operational health and the power plant-level power grid interaction strength index, and generates a parameter safety adjustment range for the virtual inertia through a preset dynamic decoupling optimization algorithm; a parameter collaborative execution module distributes the parameter safety adjustment range to each power generation unit, and each power generation unit performs local parameter fine-tuning within the parameter safety adjustment range based on its own unit operational health, generating the final control command.

[0017] Compared with the prior art, the present invention has the following advantages:

[0018] This invention achieves bidirectional adaptive adjustment of the virtual inertia control strategy from both the supply and demand sides through dual dynamic assessment of the power generation unit's own health status and the grid strength at the connection point. This method ensures that control commands meet the real-time support requirements of the grid without exceeding the equipment's own safety tolerance, thereby improving grid frequency stability while also considering the operational economy and long-term reliability of the power generation equipment, thus forming a positive interaction between the power plant and the grid.

[0019] This invention proposes a hierarchical and collaborative control architecture within a power station. The station-level controller generates a safe adjustment range for parameters based on global information, and each power generation unit then autonomously fine-tunes these ranges according to its own status. This control mode combines the global optimization of centralized control with the flexibility of distributed control, enabling the utilization of the adjustment potential of each healthy unit while avoiding excessive strain on poorly performing units. It solves the problems of low adjustment efficiency and uneven equipment load caused by traditional control methods, thereby improving the overall control robustness and comprehensive benefits of the power station.

[0020] This invention introduces a self-learning and optimization closed loop driven by historical data, enabling the core model and algorithm of the control system to continuously iterate and improve itself. This mechanism allows the control strategy to automatically adapt to the long-term evolution of grid characteristics and the natural degradation of equipment performance, ensuring the advanced nature and effectiveness of the control method during long-term operation. This gives the new energy power plant an evolutionary capability similar to that of a living organism, enabling it to continuously maintain the supporting performance of the power grid.

[0021] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a flowchart illustrating a virtual inertia adaptive control method for a new energy power plant according to an embodiment of the present invention.

[0024] Figure 2 This is a dynamic response threshold diagram based on the unit's operational health in an embodiment of the present invention.

[0025] Figure 3 This is a diagram showing the safe adjustment range of virtual inertia parameters according to an embodiment of the present invention.

[0026] Figure 4 This is a schematic diagram of the structure of a virtual inertia adaptive control system for a new energy power plant according to an embodiment of the present invention. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] Reference Figure 1One embodiment of the present invention proposes a virtual inertia adaptive control method for new energy power plants. By using the unit operation health degree and the power plant-level grid interaction intensity index, combined with a dynamic decoupling optimization algorithm, the virtual inertia parameters can be adaptively adjusted.

[0029] The method described in this embodiment specifically includes:

[0030] S1. Obtain the local operating status parameters of each power generation unit in the new energy power station and the grid parameters of the grid connection point;

[0031] Specifically, the first step is to install or utilize existing sensors and monitoring systems within each power generation unit of the new energy power plant, such as photovoltaic inverters or wind turbine generators. These devices are used to collect local operating status parameters reflecting the unit's current operating condition and regulation capabilities in real time. These parameters may include the temperature of key components, DC-side voltage and current to characterize energy reserves, and internally recorded operating hours or fault information to aid in assessing device aging. Simultaneously, high-precision grid monitoring devices, such as power quality analyzers or phasor measurement units, need to be deployed at the common access point (grid connection point) connecting the power plant to the main grid. These devices measure grid parameters at the grid connection point in real time, primarily including the instantaneous frequency and voltage of the grid, which are crucial for assessing grid stability. All collected local operating status parameters and grid connection point parameters are transmitted in real time to the data acquisition module via the power plant's internal communication network, providing the necessary raw data input for subsequent steps to calculate the unit's operational health and the power plant-level grid interaction strength index.

[0032] S2. Calculate the unit operating health of each power generation unit based on the local operating status parameters. The unit operating health characterizes the instantaneous adjustment capability and safe operating boundary of the power generation unit.

[0033] Optionally, the calculation of the unit operating health of each power generation unit based on the local operating status parameters includes:

[0034] Extract multiple health indicator parameters from the local operating status parameters;

[0035] The weighting coefficients of each health indicator parameter are determined based on their impact on the equipment's adjustment capability, and the multiple health indicator parameters are weighted and fused to generate a health score.

[0036] The safe operating boundary of the equipment is obtained, and the health score is corrected to obtain the unit operating health of each power generation unit.

[0037] Specifically, the first step is to extract and calculate multiple health indicators that directly reflect the equipment's operating condition from a massive amount of local operating status parameters. These indicators are normalized or standardized values, such as the equipment temperature deviation, DC-side energy reserve adequacy, or quantified values ​​of device aging. Next, based on engineering experience or offline analysis, the severity of the impact of each health indicator on the instantaneous regulation capability of the power generation unit needs to be determined, and weighting coefficients are set accordingly. Weighting coefficients are dimensionless values ​​representing relative importance. Subsequently, a comprehensive health score is generated by weighted fusion of multiple health indicators. This weighted fusion process can be expressed as:

[0038]

[0039] in It is a health score. It is the first The current value of each health indicator parameter, It is its corresponding weight coefficient, and all The sum is usually 1, ensuring Maintain with Within the same normalized interval. Finally, since the above scores primarily reflect the overall condition of the equipment, a safety verification must also be introduced. It is necessary to obtain the clearly defined safe operating boundaries of the power generation unit, such as physical limits like maximum permissible operating temperature and minimum DC voltage threshold. By... The final unit operating health is obtained by comparing and adjusting against these safe operating boundaries. This adjustment process ensures that even if the overall score is acceptable, if any critical parameter has reached a safe boundary, the final unit operating health will be set to an extremely low value to prevent the equipment from performing dangerous adjustment actions.

[0040] For example, the calculation process of the unit's operational health is illustrated using a photovoltaic inverter unit. First, three health indicator parameters are extracted from its local operating status parameters, respectively... Power module temperature specifications DC side voltage margin index and The aging indicators corresponding to the cumulative operating hours of the equipment. Assuming that their impact on instantaneous power regulation capability is analyzed, their weighting coefficients are determined as follows: At a certain moment, the measured This indicates that the temperature is still good. Indicates sufficient DC energy. This indicates the equipment is relatively new. At this point, the health score obtained through weighted fusion is... Subsequently, the system obtains the inverter's safe operating boundary, assuming its power module's absolute maximum allowable temperature is 90 degrees Celsius. During correction, the system checks the current real-time temperature at 88 degrees Celsius, which does not violate the safety boundary; therefore, the correction passes, and the unit's operational health is ultimately determined to be 0.77. If at another moment, although... The calculated value is still 0.77, but the real-time temperature reaches 91 degrees Celsius. At this point, the safety operating boundary of the equipment has been breached, and the correction mechanism will be activated to forcibly correct the final unit operating health to 0, in order to indicate that its instantaneous adjustment capability has been lost, thereby preventing it from participating in the virtual inertia response.

[0041] S3. Real-time monitoring of the frequency change rate of the grid connection point. When the absolute value of the frequency change rate exceeds the fast response threshold dynamically generated based on the unit's operational health, a local virtual inertia response is triggered.

[0042] Optionally, the fast response threshold dynamically generated based on the unit's operational health includes:

[0043] A dynamic adjustment coefficient that is inversely proportional to the operational health of the unit is determined based on the unit's operational health status.

[0044] Obtain the baseline threshold that characterizes the normal operating conditions of the power grid;

[0045] The fast response threshold is calculated by multiplying the dynamic adjustment coefficient by the baseline threshold.

[0046] Specifically, this first relies on the unit operational health status of each power generation unit obtained in the previous step. Based on the unit operational health status, a dynamic adjustment coefficient that is inversely proportional to it needs to be determined. This means that the higher the unit operational health status, the better the equipment condition and the stronger the regulation capability, and the lower its corresponding dynamic adjustment coefficient; conversely, the lower the health status, the higher the dynamic adjustment coefficient. The dynamic response threshold based on the unit operational health status is as follows: Figure 2 As shown. This coefficient It is a dimensionless regulating factor that is related to the unit's operational health. The functional relationship is as follows:

[0047] ,

[0048] At the same time, it is necessary to obtain a benchmark threshold that characterizes the normal operating conditions of the power grid. This is a fixed absolute value of the frequency change rate, such as 0.5 Hz / s, representing the upper limit of frequency fluctuation acceptable to the power grid during normal operation without requiring additional inertia support. Finally, the dynamic adjustment coefficient is multiplied by the baseline threshold to calculate the final fast response threshold:

[0049] .

[0050] For example, the system acquires a baseline threshold characterizing the normal operating conditions of the power grid. It is set to 0.5 Hz / s. This is the preset dynamic adjustment coefficient of the control system. With unit operational health The inverse relationship between them is Assuming The value range is [0.8, 1.8] corresponding to the interval [1.0, 0]. Here, it is simplified to [0.8, 1.8] to ensure the coefficient is meaningful. At a certain moment, the No. 1 power generation unit in the power station is in excellent condition. Calculate its unit operational health. At this point, the system determines a dynamic adjustment coefficient for it. Therefore, the fast response threshold for unit 1 is... Meanwhile, due to slight overheating, the operational health of generator unit No. 2 was affected. The system determines a dynamic adjustment coefficient for it. The fast response threshold of unit 2 is .

[0051] S4. Calculate the power station-level grid interaction strength index based on the grid parameters. The power station-level grid interaction strength index characterizes the trend of grid strength change at the grid connection point.

[0052] Optionally, the calculation of the power plant-level grid interaction strength index based on the grid parameters includes:

[0053] Extract frequency stability parameters and voltage stability parameters from the power grid parameters;

[0054] Time series trend analysis is performed on the frequency stability parameters and voltage stability parameters to generate power grid strength trend values;

[0055] The power grid strength trend value is mapped to an index to calculate the power plant-level power grid interaction strength index.

[0056] Specifically, it is first necessary to continuously sample the instantaneous frequency of the grid connection point at a high frequency. and instantaneous voltage These data are all acquired in real time by the high-precision power grid monitoring devices deployed in the above steps. Next, the fluctuations of these two power grid parameters need to be quantified to reflect the changing trends in power grid strength. Frequency change rate It is an important indicator of frequency stability; the larger its absolute value, the higher the urgency of a disturbance in the power grid. Simultaneously, the amplitude of voltage changes... The depth of voltage dips also reflects the robustness of the power grid. To obtain a comprehensive index... This requires a weighted fusion of the frequency change rate and the voltage change amplitude. This fusion can be achieved using the following formula:

[0057] ,

[0058] in The power plant-level grid interaction strength index. and These are the weighting coefficients for the absolute value of the rate of change of frequency and the amplitude of voltage change, respectively. They are dimensionless preset values ​​used to balance the importance of these two parameters in characterizing grid strength. All parameters should be normalized to ensure they are calculated within the same dimensional space. The resulting power station-level grid interaction strength index is a quantitative indicator; a higher value indicates a more severe disturbance in the grid and a more urgent need for virtual inertia support provided by renewable energy power stations.

[0059] For example, assume that the instantaneous frequency and instantaneous voltage of the grid parameters at the grid connection point are monitored in real time. Let the absolute value of the rate of change of frequency be... The normalized value is Voltage variation amplitude The normalized value is Preset weighting coefficients and This indicates a higher level of focus on the rate of frequency change. At a certain moment, a sudden large-capacity load shedding occurs in the power grid, causing a sharp rise in frequency. The normalized value of the absolute value of the rate of frequency change... achieve However, the voltage fluctuation is small, and the normalized value of the voltage change amplitude is... for At this time, the power plant-level grid interaction strength index At another moment, a near-end short-circuit fault occurred in the power grid, causing a severe instantaneous voltage drop. achieve However, due to the rapid protective action, the frequency change rate is not large. for At this time, the power plant-level grid interaction strength index In this way, the power plant-level grid interaction strength index can comprehensively reflect the grid's dual requirements for frequency stability and voltage stability.

[0060] S5. Integrate the unit's operational health status with the power plant-level grid interaction strength index, and generate the parameter safety adjustment range of the virtual inertia through a preset dynamic decoupling optimization algorithm;

[0061] Optionally, the parameter safety adjustment range for generating the virtual inertia by integrating the unit's operational health and the power plant-level grid interaction strength index through a preset dynamic decoupling optimization algorithm includes:

[0062] The unit's operational health and the power plant-level grid interaction intensity index are decoupled into equipment regulation capability factor and grid demand factor to obtain independent influencing factors;

[0063] Based on the independent influencing factors, optimized control parameters are obtained through a preset dynamic decoupling optimization algorithm;

[0064] Obtain the safety constraints of the physical characteristics of the equipment, and define the parameter safety adjustment range of the virtual inertia around the optimized control parameters.

[0065] Specifically, the unit's operational health status is first obtained from the station-level controller. Interaction strength index with power plant-level power grid And decouple them into equipment regulation capability factors. and grid demand factor These are two independent quantities. The purpose of decoupling is to eliminate potential coupling effects between unit operational health and the power plant-level grid interaction strength index, ensuring that subsequent optimization is based purely on the intrinsic capabilities of the equipment and external grid demands. The decoupling process can employ multivariate statistical methods such as principal component analysis or independent component analysis. Next, these independent influencing factors... and The input is a preset dynamic decoupling optimization algorithm. The core of this algorithm is to solve for the optimal control parameters based on the real-time states of the two factors. The optimized control parameters are the station-level optimal virtual inertia parameters, satisfying both safety and support requirements. Finally, it is necessary to obtain the equipment's physical characteristic safety constraints, i.e., the absolute physical limits of each power generation unit in terms of virtual inertia parameters, such as the maximum power change rate limit. The system, based on the optimized control parameters and referring to the equipment's physical characteristic safety constraints, defines the safe adjustment range of the obtained virtual inertia parameters. This range ensures that the parameters are within... The unit meets the safety operation requirements of the worst-case health unit and is in The location can provide the maximum effective support indicated by the optimal control parameters, and the safe adjustment range of the virtual inertia parameter is as follows: Figure 3 As shown.

[0066] For example, suppose the decoupling operation will affect the health of the unit operation. Projection as The power plant-level grid interaction strength index Projection as At a certain moment, the power station's High, It's also very high. The independent impact factor obtained after decoupling is... At a high value, These values ​​are also high. Based on these two high values, the dynamic decoupling optimization algorithm will generate a relatively high optimization control parameter. ,For example Simultaneously, it obtains safety constraints on the equipment's physical characteristics, such as the absolute upper limit of the power change rate, and parameters like inertia. The physical cap is 5.0. Around Define the safe adjustment range of the parameters, for example, [1.0, 4.5]. At another moment, Lower It's also lower. After decoupling At a low value, The value is low. The algorithm will generate lower optimized control parameters. ,For example At this time, surrounding Furthermore, the safety adjustment range of parameters is determined by referring to the safety constraints of the physical characteristics of the equipment, for example, [0.5, 2.5].

[0067] Optionally, the step of decoupling the unit's operational health and the power plant-level grid interaction intensity index into equipment regulation capability factors and grid demand factors to obtain independent influencing factors includes:

[0068] The unit's operational health and the power plant-level grid interaction strength index are constructed into a joint feature vector;

[0069] Process the joint feature vector and project it onto a mutually orthogonal feature space;

[0070] In the feature space, the equipment regulation capability factor, which characterizes the inherent capabilities of the equipment, and the power grid demand factor, which represents the external demands of the power grid, are separated to obtain independent influencing factors.

[0071] Specifically, the first step is to check the health status of the unit. Interaction strength index with the power plant-level power grid Construct as a joint feature vector The joint feature vector can be expressed as:

[0072] ,

[0073] in, The unit's operational health status is obtained by the station-level controller. It is the power station-level grid interaction strength index obtained by the station-level controller. This represents the vector transpose operation. Next, the joint eigenvector is processed using multivariate statistical analysis methods such as principal component analysis or independent component analysis. The core operation is to project it onto mutually orthogonal feature spaces. This projection transformation transforms two indicators that might otherwise be correlated. and This is transformed into new, independent, orthogonal feature components. Finally, within the mutually orthogonal feature space, the device regulation capability factor, which characterizes the inherent capabilities of the device, is precisely separated. and the grid demand factor characterizing external grid demand These factors collectively constitute independent influencing factors. The principle behind this decoupling operation is that it utilizes orthogonal transformation to decouple the supply side... and demand side The information is mathematically decoupled to ensure... It only reflects the pure supply capacity of the equipment in an instantaneous adjustment, while It only reflects the purely external demand of the power grid for inertia support.

[0074] For example, assume that the station-level controller obtains the unit's operational health status. The power plant-level grid interaction strength index is 0.9. The value is 0.8. First, they are constructed into a joint feature vector. Next, the dynamic decoupling optimization module runs principal component analysis to process and project the vector. By projecting it onto mutually orthogonal feature spaces, the algorithm separates the equipment adjustment capability factor, which characterizes the equipment's inherent capabilities. The grid demand factor, which is at a high value, also represents the external demand of the power grid. It is also at a high value. Based on and All values ​​are high, and the dynamic decoupling optimization algorithm yields a higher optimal control parameter. Under another working condition, if Lower, for example, 0.3, while It is also relatively low, for example, 0.2, obtained after decoupling. and All are at low values. At this point, the algorithm will generate lower optimized control parameters. ,For example This decoupling mechanism ensures that the independent influencing factors relied upon by the station-level controller are purely supply and demand information, avoiding, for example... High times may be related to A certain inherent correlation can bias the optimization results, thereby affecting subsequent optimization efforts. The defined parameter safety adjustment range can more accurately match the equipment safety limits and the real-time needs of the power grid.

[0075] S6. The parameter safety adjustment range is sent to each power generation unit. Each power generation unit performs local parameter fine-tuning within the parameter safety adjustment range based on its own unit operating health, and generates the final control command.

[0076] Optionally, each power generation unit performs local parameter fine-tuning within the parameter safety adjustment range based on its own unit operational health, generating final control commands including:

[0077] The local adjustment weight is calculated based on the health status of the unit.

[0078] The virtual inertia parameter is determined by interpolating between the upper and lower limits of the parameter safety adjustment range using the local adjustment weights.

[0079] Based on the virtual inertia parameters, the final control command is generated.

[0080] Specifically, firstly, the safety adjustment range of parameters issued by the receiving station-level controller of each power generation unit. It is a station-level virtual inertia parameter Global limitations. Simultaneously, each power generation unit obtains its own unit operational health status from the above steps. Next, the local controller of the power generation unit performs local parameter fine-tuning based on the received parameter safety adjustment range and its own unit operational health. The purpose of local parameter fine-tuning is to select the optimal virtual inertia parameter according to its own state within the safety range determined at the station level. The principle of fine-tuning is... The higher the unit, the more options are allowed for selection. The closer to the upper limit of the parameter's safe adjustment range To provide stronger inertia support; and For lower-value units, the preference is to select those closer to the lower limit. of This is to protect the equipment. Such fine-tuning can be achieved using linear mapping or piecewise functions, for example:

[0081] ,

[0082] in, These are the virtual inertia parameters that were ultimately determined for this power generation unit. and These are the upper and lower limits of the safety adjustment range for parameters issued at the station level. This is a normalized unit health score used for fine-tuning, with a value ranging from [0,1]. This formula ensures... Always in Within the specified range, and positively correlated with device health. Finally, the local controller will apply this adaptive... Substitute this into the calculation of the virtual inertia controller, and combine it with the real-time monitored frequency deviation at the grid connection point. Generate the final control commands. The instruction will be rapidly injected into the power grid to provide instantaneous power support.

[0083] For example, suppose the station-level controller issues a parameter safety adjustment range based on the power grid and overall health. =[1.5, 4.5]. Health status of generator unit #1 within the power plant. Better, after normalization The condition of generator unit No. 2 is slightly worse; after normalization... Calculate the virtual inertia parameters for each element according to the fine-tuning formula: Element 1 Unit 2: Unit 1, due to its high health rating, was given a value of 4.2, close to the upper limit of 4.5, to provide strong inertia support; while Unit 2 was given a value of 2.7, which met the basic requirements of the station level and also protected the equipment.

[0084] Optionally, calculating the local adjustment weight based on the unit's operational health includes:

[0085] Based on the operational health of the unit, it is quantified as the adjustable margin of the power generation unit;

[0086] Based on the adjustable margin, the local adjustment weight is calculated.

[0087] Specifically, the first step is to obtain the calculated unit operational health status. Unit operational health It is a comprehensive indicator that characterizes the instantaneous regulation capability and safe operating boundary of a power generation unit. Its value is typically between [0,1], with higher values ​​indicating stronger available regulation capability. This indicates the unit's operational health. Directly quantified as the adjustable margin of the power generation unit ,Right now Adjustable margin It is a dimensionless value that intuitively reflects the maximum inertia support potential that the unit can release without violating safety boundaries. Then, based on this adjustable margin... The local adjustment weights used for fine-tuning local parameters are calculated. Adjusting weights locally Its function is to adjust the safety range of parameters issued at the station level. The final virtual inertia parameters mapped to this unit Since a larger adjustment margin indicates a healthier device, it should be allowed to choose a larger virtual inertia parameter to provide stronger support; therefore, local adjustment of weights is necessary. Should be with adjustable margin They are positively correlated. The most direct method is to adjust the margin. After normalization, it is used as the local adjustment weight. ,Right now ,make sure It lies between [0,1].

[0088] For example, a power generation unit calculates its unit operational health. The value is 0.8. Based on the definition in this step, it is first quantified as the adjustable margin of the power generation unit. This means that the unit also has This potential can be used for virtual inertia support without triggering security protections. Then, based on this... The local adjustment weights are calculated. In another scenario, another power generation unit, despite having a high overall health score, experienced an excessively high temperature in its power module, ultimately resulting in a revised unit operational health score because it exceeded safety limits. It is set to 0.2. At this point, its adjustable margin... Local weight adjustment This kind of passage Quantified as This method effectively transforms parameters into adjustment factors within the control strategy. It enables the global parameter range issued at the station level to be fine-tuned locally based on the actual health status of each generating unit. Healthy units receive higher adjustment weights, thus selecting larger inertia parameters to enhance system support; unhealthy units receive lower adjustment weights, thus selecting smaller inertia parameters to ensure their safe operation, achieving both refined control commands and guaranteed safe operation.

[0089] Optionally, the method further includes:

[0090] Collect historical local operating status parameters, grid parameters at the grid connection point, final control commands, and grid frequency feedback signals after executing control commands to form a historical operating dataset;

[0091] Based on the historical operation dataset, update the calculation model for calculating the operational health of the unit and the algorithm for calculating the power plant-level grid interaction strength index;

[0092] The updated computational model and algorithm are applied to optimize the fast response threshold and the safety adjustment range.

[0093] Specifically, the first step is to continuously sample and store operational data at high frequencies. This includes local operating status parameters of each power generation unit, grid parameters at the grid connection point, the final control commands issued by the system, and the grid frequency feedback signals monitored after the commands are executed. This time-aligned data collectively forms a historical operational dataset. The system sets trigger conditions or periodically initiates a self-learning program, which uses model training algorithms such as neural networks or regression analysis to perform in-depth analysis of the historical operational dataset. The core objective of this analysis is to update the key models and algorithms in the control system. Firstly, the self-learning program updates the calculation model used to assess the operational health of the units, for example, by adjusting the weighting coefficients of health indicator parameters. This weighting factor is used for health scoring. The score is calculated based on the current values ​​of various health indicator parameters. Its corresponding weight coefficient The result of the weighted fusion. Secondly, the self-learning program updates the algorithm used to calculate the power plant-level grid interaction strength index, for example, by correcting the weighting coefficients in the fusion formula. and These weights are used to balance the importance of the normalized absolute value of the rate of change of frequency and the magnitude of the voltage change in the exponential calculation. Finally, the system applies the updated calculation model and algorithm for real-time optimization. On the one hand, the updated unit health calculation model will output a more accurate health status, thereby optimizing the fast response threshold. The threshold It is achieved by dynamically adjusting the coefficients based on the updated health status. Compared with the benchmark threshold characterizing normal operating conditions The results are obtained through multiplication. On the other hand, the updated unit operating health and power plant-level grid interaction strength index are input into the dynamic decoupling optimization algorithm to generate a more accurate parameter safety adjustment range.

[0094] For example, in the initial stage of system operation, the DC-side voltage margin index has a certain weight in the model used to calculate the operational health of the units. It was set to 0.3. After several months of operation, historical running datasets were collected. The self-learning program, through analysis, discovered that when the DC-side voltage margin index... At lower levels, the grid frequency feedback signal corresponding to the final control command issued indicates poor support and the equipment is prone to triggering protection. System judgment The impact was underestimated; regression analysis automatically adjusted the values. The value is increased to 0.5. Once this updated calculation model is applied, in subsequent runs, if the DC voltage margin of a certain unit is low, its... The final unit health score will be calculated lower because of the dynamic adjustment factor. This unit is inversely proportional to health level. It will increase. Based on the fast response threshold. It is determined by the dynamic adjustment coefficient. and benchmark threshold The computational relationship obtained by multiplication is the fast response threshold of this unit. The frequency will be optimized and increased. This means that the unit will only be triggered when there are more severe frequency fluctuations in the power grid, thus protecting equipment under low DC voltage conditions. Simultaneously, if historical data shows that the impact of the frequency change rate on stability is underestimated under specific power grid conditions, the self-learning program will adjust the frequency change rate weighting coefficient. Automatically increased. Applying this updated algorithm, the calculated power plant-level grid interaction strength index will be higher in subsequent similar grid disturbances, enabling the dynamic decoupling optimization algorithm to generate an updated upper limit for the parameter safety adjustment range. This enables power plants to provide stronger support when the power grid needs it.

[0095] Based on the same inventive concept, such as Figure 4 As shown, the present invention also provides a virtual inertia adaptive control system for new energy power plants, the system comprising:

[0096] The data acquisition module is used to obtain the local operating status parameters of each power generation unit in the new energy power plant and the grid parameters of the grid connection point;

[0097] The unit health assessment module is used to calculate the unit operating health of each power generation unit based on the local operating status parameters. The unit operating health characterizes the instantaneous adjustment capability and safe operating boundary of the power generation unit.

[0098] The virtual inertia triggering module is used to monitor the frequency change rate of the grid connection point in real time. When the absolute value of the frequency change rate exceeds the fast response threshold dynamically generated based on the unit's operational health, it triggers a local virtual inertia response.

[0099] The power grid strength assessment module is used to calculate the power station-level power grid interaction strength index based on the power grid parameters. The power station-level power grid interaction strength index characterizes the trend of power grid strength change at the grid connection point.

[0100] The dynamic decoupling optimization module is used to integrate the unit's operational health and the power plant-level grid interaction intensity index, and generate the parameter safety adjustment range of the virtual inertia through a preset dynamic decoupling optimization algorithm.

[0101] The parameter coordination execution module is used to distribute the parameter safety adjustment range to each power generation unit. Each power generation unit performs local parameter fine-tuning within the parameter safety adjustment range based on its own unit operating health, and generates the final control command.

[0102] It should be noted that the electrical connections between the various units described above do not necessarily represent direct or indirect connections. Any indirect connection method can be applied to the embodiments of the present invention as long as it achieves the purpose of the present invention. The above descriptions are merely exemplary embodiments of the present invention and should not be construed as limiting the scope of the present invention.

[0103] All equivalent changes and modifications made in accordance with the teachings of this invention are still within the scope of this invention. Those skilled in the art will readily conceive of other embodiments of this invention upon considering the specification and the disclosure of practical truth. This application is intended to cover any variations, uses, or adaptations of this invention that follow the general principles of this invention and include common knowledge or conventional techniques in the art not described herein.

Claims

1. A virtual inertia adaptive control method for a new energy power plant, characterized in that, The method includes: Obtain the local operating status parameters of each power generation unit in the new energy power plant and the grid parameters of the grid connection point; The unit operating health of each power generation unit is calculated based on the local operating status parameters. The unit operating health characterizes the instantaneous adjustment capability and safe operating boundary of the power generation unit. Real-time monitoring of the frequency change rate at the grid connection point; when the absolute value of the frequency change rate exceeds the fast response threshold dynamically generated based on the unit's operational health, a local virtual inertia response is triggered. The power station-level grid interaction strength index is calculated based on the grid parameters. The power station-level grid interaction strength index characterizes the trend of grid strength change at the grid connection point. By integrating the unit's operational health and the power plant-level grid interaction strength index, a preset dynamic decoupling optimization algorithm is used to generate the parameter safety adjustment range of the virtual inertia. This includes: decoupling the unit's operational health and the power plant-level grid interaction strength index into equipment regulation capability factors and grid demand factors to obtain independent influencing factors; obtaining optimized control parameters based on the independent influencing factors using the preset dynamic decoupling optimization algorithm; obtaining equipment physical characteristic safety constraints, and defining the parameter safety adjustment range of the virtual inertia around the optimized control parameters, including: constructing a joint feature vector from the unit's operational health and the power plant-level grid interaction strength index; processing the joint feature vector and projecting it onto a mutually orthogonal feature space; separating the equipment regulation capability factor characterizing the equipment's intrinsic capabilities and the grid demand factor representing the external grid demand in the feature space to obtain independent influencing factors. The parameter safety adjustment range is sent to each power generation unit. Each power generation unit performs local parameter fine-tuning within the parameter safety adjustment range based on its own unit operating health, and generates the final control command.

2. The virtual inertia adaptive control method for a new energy power plant according to claim 1, characterized in that, The unit operating health status of each power generation unit calculated based on the local operating status parameters includes: Extract multiple health indicator parameters from the local operating status parameters; The weighting coefficients of each health indicator parameter are determined based on their impact on the equipment's adjustment capability, and the multiple health indicator parameters are weighted and fused to generate a health score. The safe operating boundary of the equipment is obtained, and the health score is corrected to obtain the unit operating health of each power generation unit.

3. The virtual inertia adaptive control method for a new energy power plant according to claim 1, characterized in that, The fast response threshold dynamically generated based on the unit's operational health includes: A dynamic adjustment coefficient that is inversely proportional to the operational health of the unit is determined based on the unit's operational health status. Obtain the baseline threshold that characterizes the normal operating conditions of the power grid; The fast response threshold is calculated by multiplying the dynamic adjustment coefficient by the baseline threshold.

4. The virtual inertia adaptive control method for a new energy power plant according to claim 1, characterized in that, The power plant-level grid interaction strength index calculated based on the grid parameters includes: Extract frequency stability parameters and voltage stability parameters from the power grid parameters; Time series trend analysis is performed on the frequency stability parameters and voltage stability parameters to generate power grid strength trend values; The power grid strength trend value is mapped to an index to calculate the power plant-level power grid interaction strength index.

5. The virtual inertia adaptive control method for a new energy power plant according to claim 1, characterized in that, Each power generation unit performs local parameter fine-tuning within the parameter safety adjustment range based on its own unit operational health, and generates final control commands including: The local adjustment weight is calculated based on the health status of the unit. The virtual inertia parameter is determined by interpolating between the upper and lower limits of the parameter safety adjustment range using the local adjustment weights. Based on the virtual inertia parameters, the final control command is generated.

6. The virtual inertia adaptive control method for a new energy power plant according to claim 1, characterized in that, The method further includes: Collect historical local operating status parameters, grid parameters at the grid connection point, final control commands, and grid frequency feedback signals after executing control commands to form a historical operating dataset; Based on the historical operation dataset, update the calculation model for calculating the operational health of the unit and the algorithm for calculating the power plant-level grid interaction strength index; The updated computational model and algorithm are applied to optimize the fast response threshold and the safety adjustment range.

7. The virtual inertia adaptive control method for a new energy power plant according to claim 5, characterized in that, The calculation of the local adjustment weight based on the unit's operational health includes: Based on the operational health of the unit, it is quantified as the adjustable margin of the power generation unit; Based on the adjustable margin, the local adjustment weight is calculated.

8. A virtual inertia adaptive control system for a new energy power plant, applied to the virtual inertia adaptive control method for a new energy power plant as described in any one of claims 1-7, characterized in that, The system includes: The data acquisition module is used to obtain the local operating status parameters of each power generation unit in the new energy power plant and the grid parameters of the grid connection point; The unit health assessment module is used to calculate the unit operating health of each power generation unit based on the local operating status parameters. The unit operating health characterizes the instantaneous adjustment capability and safe operating boundary of the power generation unit. The virtual inertia triggering module is used to monitor the frequency change rate of the grid connection point in real time. When the absolute value of the frequency change rate exceeds the fast response threshold dynamically generated based on the unit's operational health, it triggers a local virtual inertia response. The power grid strength assessment module is used to calculate the power station-level power grid interaction strength index based on the power grid parameters. The power station-level power grid interaction strength index characterizes the trend of power grid strength change at the grid connection point. A dynamic decoupling optimization module is used to integrate the unit's operational health and the power plant-level grid interaction strength index, and generate the parameter safety adjustment range of the virtual inertia through a preset dynamic decoupling optimization algorithm. This includes: decoupling the unit's operational health and the power plant-level grid interaction strength index into equipment regulation capability factors and grid demand factors to obtain independent influencing factors; obtaining optimized control parameters based on the independent influencing factors through the preset dynamic decoupling optimization algorithm; obtaining equipment physical characteristic safety constraints, and defining the parameter safety adjustment range of the virtual inertia around the optimized control parameters, including: constructing a joint feature vector from the unit's operational health and the power plant-level grid interaction strength index; processing the joint feature vector and projecting it onto a mutually orthogonal feature space; separating the equipment regulation capability factor representing the equipment's intrinsic capabilities and the grid demand factor representing external grid demands in the feature space to obtain independent influencing factors; and a parameter collaborative execution module is used to distribute the parameter safety adjustment range to each power generation unit, and each power generation unit performs local parameter fine-tuning within the parameter safety adjustment range based on its own unit operational health to generate the final control command.

Images