A multi-scenario power system frequency safety index adaptive partition reconstruction method

Abstract

The application discloses a multi-scene power system frequency safety index adaptive partition reconstruction method, comprising the following steps: acquiring system multi-type frequency modulation resource parameters, and establishing an analytical expression of a nonlinear frequency safety index constraint based on a system multi-machine frequency response dynamic model considering the multi-type frequency modulation resources; acquiring a system frequency modulation parameter original feasible range, sampling to generate an original frequency modulation parameter region in the system frequency modulation parameter feasible range and a frequency safety index sample surface set; acquiring all sample point subsets on the inertia boundary in the frequency modulation parameter region by traversing sample points satisfying a system specific equivalent inertia value, acquiring intersection points of sample point corresponding frequency safety indexes and constraint limits on the original frequency modulation parameter region boundary, and re-partitioning the original frequency modulation parameter region into multiple sub-regions; linearizing linear sub-regions needing linearization, outputting linearization parameters satisfying a linear error threshold requirement, and reconstructing a frequency safety index constraint after partition linearization.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and specifically to an adaptive partitioning reconstruction method for frequency security indicators of power systems in multiple scenarios. Background Technology

[0002] With the large-scale and continuous grid connection of converter-based renewable energy sources and the construction of DC transmission, the power system is gradually transforming into a high-proportion power electronic interface power system dominated by new energy units. Compared with the traditional power system, the frequency support capability of the new energy-dominated power system is continuously declining, resulting in poorer system frequency immunity. When faced with active power disturbances, if the system frequency deviation exceeds the action threshold of the system stability device, it will trigger a serious power curtailment accident.

[0003] Currently, there have been numerous power outages worldwide caused by excessive frequency drops due to insufficient frequency support capacity in renewable energy power systems. Therefore, ensuring frequency security in power systems has received widespread attention. However, the highly nonlinear nature of accurate system frequency response models makes it difficult to incorporate them into various changing operating scenarios. Effectively incorporating frequency security constraints into various operating scenarios, such as optimized operation and system planning, is crucial for ensuring the safe and stable operation of the system.

[0004] Currently, methods for incorporating frequency security constraints into power system operation scenarios mainly include data-driven frequency regulation boundary extraction methods and piecewise linear approximation methods. The former is difficult to incorporate into large-scale power systems due to its excessive computational burden. The latter, however, benefits from its acceptable computational efficiency in a single operation scenario and is widely used in day-ahead-scale unit combination research. However, the accuracy of piecewise linear approximation methods is affected by the number of segments. A large number of segments leads to the approximation model containing numerous intermediate variables and linear constraints, significantly increasing the scale of the overall system model. Therefore, the significantly increased computational burden exposed in longer-term models makes it difficult to apply to large-scale power system operation models involving multiple scenarios. Summary of the Invention

[0005] In view of this, the present invention provides an adaptive partitioning reconstruction method for frequency security indicators of power systems in multiple scenarios, which at least solves the problem that the existing methods for constructing frequency security indicators are difficult to perform with high accuracy and high efficiency in multiple operating scenarios.

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

[0007] An adaptive partitioned reconstruction method for frequency security indicators in power systems under multiple scenarios includes the following steps:

[0008] S1. Under the preset active power disturbance Below, the running times of each running scenario will be... Extreme values ​​of frequency deviation at time As a nonlinear frequency safety indicator, The frequency deviation threshold that does not exceed the maximum frequency required to ensure that the system's safety and stability devices are not triggered To constrain nonlinear frequency safety indicators; based on the system's multi-type frequency regulation resource parameters, under a unified synchronous unit response time constant. Under the given conditions, construct the system's equivalent frequency modulation parameters; based on the system's equivalent frequency modulation parameters, construct the nonlinear frequency safety index constraints and their analytical expressions. The system's equivalent frequency modulation parameters include: system equivalent inertia. Adjustment coefficient and reheat time coefficient ;

[0009] S2. Obtain , and The maximum feasible range is taken as the feasible range of the original frequency modulation parameters, and sampling is generated within the feasible range of the original frequency modulation parameters. Each sample point and its corresponding data And obtain the corresponding sample surface set. ;

[0010] S3. Obtain Feasible range of original frequency modulation parameters The corresponding sample surface, traversed and The corresponding sample surface set and The intersection points of the nonlinear frequency safety index and its constraint limit on the feasible range boundary of the original frequency modulation parameters are obtained respectively, and the droop coefficient corresponding to the obtained intersection points is used as the basis for the calculation. and The feasible range of the original frequency modulation parameters is re-partitioned to obtain the partitioned sub-regions;

[0011] S4. Linearize the sub-regions that need to be linearized and output the linearization parameters that meet the linear error threshold requirements. After reconstructing the nonlinear frequency safety index constraints, the frequency safety index constraints after partition linearization are obtained.

[0012] Preferably, the multiple types of frequency regulation resources include: synchronous generator units, converter-based new energy generator units, and pumped storage; the parameters of the multiple types of frequency regulation resources include: the installed capacity of each type of frequency regulation resource, the inertial time constant, primary frequency regulation droop coefficient, governor response time constant and turbine reheat time constant of synchronous generator units, the virtual inertial time constant and virtual droop coefficient of new energy generator units, and the inertial time constant and droop coefficient of pumped storage generator units.

[0013] Preferably, the nonlinear frequency safety index constraint is expressed as:

[0014] ;

[0015] The analytical expression for the nonlinear frequency safety index is:

[0016] ;

[0017] in, This is the load damping coefficient; , and These are intermediate variables representing the nonlinear relationship of the equivalent frequency modulation parameters of the coupled system. This represents the moment when the frequency deviation reaches its extreme value.

[0018] Preferably, the specific steps for constructing an analytical expression for the nonlinear frequency security index based on the system's equivalent frequency modulation parameters include:

[0019] S11. Establish a multi-machine frequency dynamic response model for the system. The first-order differential equation of the system's frequency dynamic response under any given operating scenario is expressed as:

[0020] ;

[0021] In the formula, Let the system's equivalent inertia be given in any given operating scenario. For system frequency deviation, , and The first The first synchronous generator unit, the first Taiwan New Energy Unit and the first The primary frequency regulation response power adjustment of the pumped storage power station; , and It is a collection of synchronous generator units, new energy generator units, and pumped storage units;

[0022] Among them, the primary frequency regulation power adjustment of the equivalent prime mover-governing device for synchronous generators and pumped storage units varies with frequency deviation. The transfer function of change is expressed as:

[0023] ;

[0024] ;

[0025] In the formula, For the synchronous generator speed governor gain, The turbine coefficient is the turbine coefficient of the turbine. The reheater time constant is... This refers to the governor droop coefficient of a pumped storage power station. and These are the response time constants of the pumped storage reversible pump turbine and the guide vane, respectively.

[0026] New energy generating units based on converters participate in frequency response through virtual droop control. Their primary frequency regulation power model is as follows:

[0027] ;

[0028] In the formula, This represents the droop factor of the virtual synchronizer. This is the response time constant of the virtual synchronizer;

[0029] S11. Response time constant of unified synchronous generator units The system's multi-machine frequency dynamic response model is then aggregated into a system equivalent frequency modulation parameter model that varies with each scenario over time.

[0030] ;

[0031] In the formula, , and These are the virtual inertial time constants for synchronous generators, pumped storage units, and new energy generators, respectively. , and This refers to the installed capacity of synchronous generating units, pumped storage units, and new energy generating units. As the baseline capacity, This is the start / stop status variable for the synchronous generator unit; 1 indicates start-up, and 0 indicates stop. This represents the state variable for pumped-storage hydropower generation; it is 1 in the power generation state and 0 in the pumping state. , and These are state variables indicating whether frequency support is provided for synchronous generators, pumped storage units, and new energy generators, respectively. The value is 1 when frequency support is provided and 0 when frequency support is not provided.

[0032] S13. Based on the system's equivalent frequency modulation parameter model, the nonlinear frequency safety index constraint in the time domain and its analytical expression are derived using the inverse Laplace transform.

[0033] Preferably, the specific content of S2 includes:

[0034] S21. Based on all current frequency modulation resource parameters of the system , and Calculate the minimum and maximum frequency support schemes to obtain the maximum feasible range of system frequency modulation parameters as the original feasible range of frequency modulation parameters. , and The feasible range of the original frequency modulation parameters is:

[0035] ;

[0036] S22. Based on Latin hypercube sampling Internally generate sample points considering all possible unit combinations and resource frequency support schemes, where the first... Each sample point is represented as , and according to Calculate the extreme values ​​of frequency deviation for all sample points. , of which The extreme values ​​of frequency deviation corresponding to each sample point are denoted as ;

[0037] S23. Adjust the frequency parameters from all sample points. Choose any sample value from the dimensions In determining Obtain the sample surface set under the following conditions: ,in, Indicates the first The extreme coordinates of the frequency deviation corresponding to each sample point It represents the three-dimensional real number space.

[0038] Preferably, the specific content of obtaining the intersection point of the nonlinear frequency safety index and its constraint limit on the feasible range boundary of the original frequency modulation parameters in S3 includes:

[0039] S31. Obtain Feasible range of original frequency modulation parameters The corresponding sample surface will and The corresponding set of sample surfaces is defined as and :

[0040] ;

[0041] ;

[0042] In the formula, and They represent and The corresponding extreme coordinates of frequency deviation and They represent and The corresponding adjustment coefficient, and They represent and The corresponding extreme value of frequency deviation;

[0043] S32. Obtain the nonlinear frequency safety index and its constraint limits. The intersection points on the boundaries of the feasible range of the original frequency modulation parameters include:

[0044] (1) For the index set Perform initialization, that is, let ,in , Indicates the first The extreme value of frequency deviation corresponding to each sample point Is the value the same as Equal binary indexes, when equal When they are not equal ;

[0045] (2) Iterate through the subset For all sample points, if the extreme value of the frequency deviation corresponding to the sample point is... If they are equal, then set the current binary index. Update the index set with this value. ,otherwise Update the index set with this value. ;

[0046] (3) Obtaining satisfaction adjustment coefficient , represented as: ,in, For the present The preceding binary indicator index, corresponding to the sample point That is, the sample points on the system inertia boundary where the nonlinear frequency safety index is equal to the frequency safety constraint limit;

[0047] (4) Obtain according to steps (1)-(3) Corresponding adjustment coefficient and the corresponding sample points .

[0048] Preferably, in step S3, the feasible range of the original frequency modulation parameters is re-partitioned based on the droop coefficient corresponding to the obtained intersection point, resulting in sub-regions. , and They are respectively:

[0049] ;

[0050] ;

[0051] ;

[0052] in, and They are respectively The minimum and maximum values, and They are respectively The minimum and maximum values, and They are respectively The minimum and maximum values.

[0053] Preferably, the specific content in S4 includes:

[0054] S41. Discard completely unsafe subregions. Preserve a completely secure sub-region For partially secure sub-regions Perform linearization:

[0055] Linear hyperplane fitting method is used for sub-regions Linearization of the frequency deviation extreme value model within:

[0056] The linear optimization model is used to fit the extreme value of the nonlinear frequency deviation to the nearest value. The linear form model of the frequency modulation parameters, sub-region Nonlinear frequency deviation extremes The sample surface is approximated as a series of linearly fitted planes; the linear optimization model is:

[0057] ;

[0058] in, For the first The approximation error corresponding to each sample point To ensure that the approximate linear result lies on the sample surface, To linearize the error threshold, The first The linear fitting coefficients of each hyperplane are the linearization parameters that satisfy the linear error threshold requirement.

[0059] The nonlinear frequency safety index constraint is then finally reconstructed as follows:

[0060] .

[0061] As can be seen from the above technical solution, compared with the prior art, the present invention discloses an adaptive partitioning reconstruction method for frequency security indicators of power systems in multiple scenarios, which has the following beneficial effects:

[0062] This invention is based on sampling to obtain a frequency modulation parameter region and a series of sample points within the original frequency modulation parameter range of the system. It acquires all sample points on the equivalent inertia boundary of the system through a traversal method and introduces a loop judgment step to adaptively partition the feasible region of the original frequency modulation parameters of the system, only applying linear approximation to the sub-regions. This invention introduces a hyperplane linearization method to achieve an adaptive partitioned reconstruction method for frequency security indicators. By adaptively partitioning and shrinking the region requiring linear approximation, this invention significantly reduces the scale of the reconstructed frequency security indicator constraint model and variables. Simultaneously, it avoids excessively large linearization parameter ranges, reducing the overall range of model input parameters and thus significantly enhancing its computational efficiency in power system operation scenario models. This invention offers higher solution efficiency and adaptability for power system operation and planning with frequency security constraints across multiple long-term operating scenarios. Related research can assist grid operators in determining reasonable frequency resource planning layouts and operation schemes, providing support for the stable operation of power systems. Attached Figure Description

[0063] 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 only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0064] Figure 1 A schematic diagram of the process for an adaptive partitioning reconstruction method for frequency security indicators in a multi-scenario power system provided by the present invention;

[0065] Figure 2 This is a test system topology diagram provided in an embodiment of the present invention;

[0066] Figure 3 This is a schematic diagram comparing the linearity results of a sample surface under the same error constraints using different methods provided in the embodiments of the present invention. Figure 3 (a) is the adaptive partitioning reconstruction method proposed in this invention; Figure 3 (b) is the traditional piecewise linear method;

[0067] Figure 4 A schematic diagram of the frequency deviation extreme value index obtained by solving the frequency security index reconstruction method proposed in this invention for the four models provided in the embodiments of this invention;

[0068] Figure 5 The system frequency simulation trajectory diagram provided in the embodiment of the present invention;

[0069] Figure 6This is a schematic diagram showing the duration proportion of each sub-region where the frequency modulation parameters obtained by solving the frequency security index partitioning reconstruction method proposed in this invention are located in different models provided in the embodiments of this invention. Figure 6 (a) is Model 1. Figure 6 (b) is Model 2. Figure 6 (c) represents model 3. Figure 6 (d) is model 4. Detailed Implementation

[0070] 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, and 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.

[0071] This invention provides an adaptive partitioned reconfiguration method for frequency security indicators in power systems under multiple scenarios, such as... Figure 1 As shown, it includes the following steps:

[0072] S1. Under the preset active power disturbance Below, the running times of each running scenario will be... Extreme values ​​of frequency deviation at time As a nonlinear frequency safety indicator, The frequency deviation threshold that does not exceed the maximum frequency required to ensure that the system's safety and stability devices are not triggered To constrain nonlinear frequency safety indicators; based on the system's multi-type frequency regulation resource parameters, under a unified synchronous unit response time constant. Under the given conditions, construct the system's equivalent frequency modulation parameters; based on the system's equivalent frequency modulation parameters, construct the nonlinear frequency safety index constraints and their analytical expressions. The system's equivalent frequency modulation parameters include: system equivalent inertia. Adjustment coefficient and reheat time coefficient ;

[0073] S2. Obtain , and The maximum feasible range is taken as the feasible range of the original frequency modulation parameters, and sampling is generated within the feasible range of the original frequency modulation parameters. Each sample point and its corresponding data And obtain the corresponding sample surface set. ;

[0074] S3. Obtain Feasible range of original frequency modulation parameters The corresponding sample surface, traversed and The corresponding sample surface set and The intersection points of the nonlinear frequency safety index and its constraint limit on the feasible range boundary of the original frequency modulation parameters are obtained respectively, and the droop coefficient corresponding to the obtained intersection points is used as the basis for the calculation. and The feasible range of the original frequency modulation parameters is re-partitioned to obtain the partitioned sub-regions;

[0075] S4. Linearize the sub-regions that need to be linearized and output the linearization parameters that meet the linear error threshold requirements. After reconstructing the nonlinear frequency safety index constraints, the frequency safety index constraints after partition linearization are obtained.

[0076] To further implement the above technical solutions, the various types of frequency regulation resources include: synchronous generator units, converter-based new energy generator units, and pumped storage. The parameters of the various types of frequency regulation resources include: the installed capacity of each type of frequency regulation resource, the inertial time constant, primary frequency regulation droop coefficient, governor response time constant, and turbine reheat time constant of synchronous generator units, the virtual inertial time constant and virtual droop coefficient of new energy generator units, and the inertial time constant and droop coefficient of pumped storage generator units.

[0077] To further implement the above technical solution, the nonlinear frequency safety index constraint is expressed as follows:

[0078] ;

[0079] The analytical expression for the nonlinear frequency safety index is:

[0080] ;

[0081] in, This is the load damping coefficient; , and These are intermediate variables representing the nonlinear relationship of the equivalent frequency modulation parameters of the coupled system. This represents the moment when the frequency deviation reaches its extreme value.

[0082] To further implement the above technical solution, the specific steps for constructing an analytical expression for the nonlinear frequency safety index based on the system's equivalent frequency modulation parameters include:

[0083] S11. Establish a multi-machine frequency dynamic response model for the system. The first-order differential equation of the system's frequency dynamic response under any given operating scenario is expressed as:

[0084] ;

[0085] In the formula, Let the system's equivalent inertia be given in any given operating scenario. For system frequency deviation, , and The first The first synchronous generator unit, the first Taiwan New Energy Unit and the first The primary frequency regulation response power adjustment of the pumped storage power station; , and It is a collection of synchronous generator units, new energy generator units, and pumped storage units;

[0086] Among them, the primary frequency regulation power adjustment of the equivalent prime mover-governing device for synchronous generators and pumped storage units varies with frequency deviation. The transfer function of change is expressed as:

[0087] ;

[0088] ;

[0089] In the formula, For the synchronous generator speed governor gain, The turbine coefficient is the turbine coefficient of the turbine. The reheater time constant is... This refers to the governor droop coefficient of a pumped storage power station. and These are the response time constants of the pumped storage reversible pump turbine and the guide vane, respectively.

[0090] New energy generating units based on converters participate in frequency response through virtual droop control. Their primary frequency regulation power model is as follows:

[0091] ;

[0092] In the formula, This represents the droop factor of the virtual synchronizer. This is the response time constant of the virtual synchronizer;

[0093] S11. Response time constant of unified synchronous generator units The system's multi-machine frequency dynamic response model is then aggregated into a system equivalent frequency modulation parameter model that varies with each scenario over time.

[0094] ;

[0095] In the formula, , and These are the virtual inertial time constants for synchronous generators, pumped storage units, and new energy generators, respectively. , and This refers to the installed capacity of synchronous generating units, pumped storage units, and new energy generating units. As the baseline capacity, This is the start / stop status variable for the synchronous generator unit; 1 indicates start-up, and 0 indicates stop. This represents the state variable for pumped-storage hydropower generation; it is 1 in the power generation state and 0 in the pumping state. , and These are state variables indicating whether frequency support is provided for synchronous generators, pumped storage units, and new energy generators, respectively. The value is 1 when frequency support is provided and 0 when frequency support is not provided.

[0096] S13. Based on the system's equivalent frequency modulation parameter model, the nonlinear frequency safety index constraint in the time domain and its analytical expression are derived using the inverse Laplace transform.

[0097] It should be noted that:

[0098] In multiple scenarios, the frequency regulation parameters of various resources depend on their installed capacity, the unit combination status of synchronous generators, the charging and discharging status of pumped storage hydroelectric power plants, and the state variables of whether various resources participate in frequency response. Considering Unit response time constant Its sensitivity is much lower than other parameters. We conservatively assume that the time constants required for various resources are the same as the response time constant of a uniform synchronous unit. The system's equivalent frequency modulation parameters include the system's equivalent inertia. Adjustment coefficient and reheat time coefficient The parameters are aggregated into time-varying parameters for each scenario through an equivalent frequency response model.

[0099] To further implement the above technical solution, the specific content of S2 includes:

[0100] S21. Based on all current frequency modulation resource parameters of the system , and Calculate the minimum and maximum frequency support schemes to obtain the maximum feasible range of system frequency modulation parameters as the original feasible range of frequency modulation parameters. , and The feasible range of the original frequency modulation parameters is:

[0101] ;

[0102] S22. Based on Latin hypercube sampling Internally generate sample points considering all possible unit combinations and resource frequency support schemes, where the first... Each sample point is represented as , and according to Calculate the extreme values ​​of frequency deviation for all sample points. , of which The extreme values ​​of frequency deviation corresponding to each sample point are denoted as ;

[0103] S23. Adjust the frequency parameters from all sample points. Choose any sample value from the dimensions In determining Obtain the sample surface set under the following conditions: ,in, Indicates the first The extreme coordinates of the frequency deviation corresponding to each sample point It represents the three-dimensional real number space.

[0104] It should be noted that:

[0105] The original feasible range of system frequency modulation parameters is obtained by calculating the minimum and maximum frequency support scheme based on all current frequency modulation resource parameters of the system, and then obtaining the maximum feasible range of system frequency modulation parameters as the original feasible range of frequency modulation parameters. This ensures that the equivalent frequency modulation parameters of the system under any operating scenario are within this range.

[0106] To further implement the above technical solution, the specific content of obtaining the intersection point of the nonlinear frequency safety index and its constraint limit on the boundary of the feasible range of the original frequency modulation parameters in S3 includes:

[0107] S31. Obtain Feasible range of original frequency modulation parameters The corresponding sample surface will and The corresponding set of sample surfaces is defined as and :

[0108] ;

[0109] ;

[0110] In the formula, and They represent and The corresponding extreme coordinates of frequency deviation and They represent and The corresponding adjustment coefficient, and They represent and The corresponding extreme value of frequency deviation;

[0111] S32. Obtain the nonlinear frequency safety index and its constraint limits. The intersection points on the boundaries of the feasible range of the original frequency modulation parameters include:

[0112] (1) For the index set Perform initialization, that is, let ,in , Indicates the first The extreme value of frequency deviation corresponding to each sample point Is the value the same as Equal binary indexes, when equal When they are not equal ;

[0113] (2) Iterate through the subset For all sample points, if the extreme value of the frequency deviation corresponding to the sample point is... If they are equal, then set the current binary index. Update the index set with this value. ,otherwise Update the index set with this value. ;

[0114] (3) Obtaining satisfaction adjustment coefficient , represented as: ,in, For the present The preceding binary indicator index, corresponding to the sample point That is, the sample points on the system inertia boundary where the nonlinear frequency safety index is equal to the frequency safety constraint limit;

[0115] (4) Obtain according to steps (1)-(3) Corresponding adjustment coefficient and the corresponding sample points .

[0116] To further implement the above technical solution, in S3, the feasible range of the original frequency modulation parameters is re-partitioned based on the droop coefficient corresponding to the obtained intersection point, resulting in sub-regions. , and They are respectively:

[0117] ;

[0118] ;

[0119] ;

[0120] in, and They are respectively The minimum and maximum values, and They are respectively The minimum and maximum values, and They are respectively The minimum and maximum values.

[0121] To further implement the above technical solution, the specific contents of S4 include:

[0122] S41. Discard completely unsafe subregions. Preserve a completely secure sub-region For partially secure sub-regions Perform linearization:

[0123] Linear hyperplane fitting method is used for sub-regions Linearization of the frequency deviation extreme value model within:

[0124] The linear optimization model is used to fit the extreme value of the nonlinear frequency deviation to the nearest value. The linear form model of the frequency modulation parameters, sub-region Nonlinear frequency deviation extremes The sample surface is approximated as a series of linearly fitted planes; the linear optimization model is:

[0125] ;

[0126] in, For the first The approximation error corresponding to each sample point To ensure that the approximate linear result lies on the sample surface, To linearize the error threshold, The first The linear fitting coefficients of each hyperplane are the linearization parameters that satisfy the linear error threshold requirement.

[0127] The nonlinear frequency safety index constraint is then finally reconstructed as follows:

[0128] .

[0129] It should be noted that:

[0130] Extreme values ​​of frequency deviation in each sub-region divided by S3 Its constraint limit In comparison, the system skew coefficient ranges from subregion for None of them can meet their limits. The unsafe sub-region of the constraint, from the perspective of frequency safety constraints, is an unreachable region of system frequency modulation parameters and can therefore be ignored; while the range with a large droop coefficient... corresponding sub-region for All of them meet their safety limits. The frequency safety sub-region represents a frequency modulation parameter range that is equivalent to the frequency safety constraint; while the droop coefficient range located between the two intersection points... Subregions within Some sample points in the middle can meet the frequency security constraint limits, only sub-regions To further convert into equivalent frequency modulation parameters that are easy to incorporate into the system. and A linear region in the form of a linear combination.

[0131] The invention will be further illustrated by the following experiments:

[0132] This will be illustrated by taking the evaluation of the annual frequency modulation amplitude index of the existing resources of the improved IEEE HRP-38 system in a future planning scenario as an example. Figure 2 To test the system topology, the total capacity of the system's synchronous units, WT, and PV was modified to 140.14GW, 60.72GW, and 101.14GW, respectively, with renewable energy accounting for over 50%. The annual source-load forecast data was used to generate eight typical days using the K-means clustering method.

[0133] 1. Validation of the effectiveness of the adaptive partitioning reconstruction method

[0134] The effectiveness of the proposed reconstruction method is verified by comparing its computational performance and accuracy with the widely used traditional piecewise linear method. Using the feasible frequency modulation parameter range of the HRP-38 system as the original parameter space, the surface linearity results of different methods under the same error constraints for a specific sample are shown below. Figure 3 As shown, where, Figure 3 (a) is the adaptive partitioning reconstruction method proposed in this invention; Figure 3 (b) is the traditional piecewise linear method. The traditional method generates 16 linear subspaces, while the method proposed in this invention requires only 2. It can be seen that the adaptive region partitioning and reconstruction method of this invention yields negligible insecure sub-regions. This significantly narrows the range of available frequency modulation parameters and, by dividing the frequency security constraints into equivalent frequency security sub-regions that do not require linearization, further enhances the effectiveness of frequency security measures. It significantly reduces the range of frequency modulation parameters that need to be approximated by linearization, and significantly reduces the number of linear models and intermediate variables representing the segmented range.

[0135] 2. Application verification in multiple operating scenarios

[0136] To further verify the advantages and adaptability of the frequency security index adaptive partitioning reconfiguration method proposed in this invention when applied to multiple operating scenarios of power systems, four multi-scenario optimized operation models were set up in the HRP-38 system to compare and analyze the solution performance of the proposed method and the traditional piecewise linear method in power system multi-operation scenarios with different frequency regulation resources and different model scales:

[0137] Model 1: Multi-scenario operation model of existing units in the HRP-38 system;

[0138] Model 2: Gas turbine units with the same total installed capacity are connected to the HRP-38 system for multi-scenario operation.

[0139] Model 3: Pumped storage units with the same total installed capacity are connected to the HRP-38 system for multi-scenario operation.

[0140] Model 4: Multi-scenario operation model where new energy sources with the same total installed capacity are connected to the HRP-38 system;

[0141] After solving the four models constrained by frequency security indicators using two different methods, the model size and solution time for all examples under the two methods are shown in Table 1:

[0142] Table 1

[0143] ;

[0144] The significantly reduced subspace of the proposed method results in a smaller scale for incorporating various multi-scenario operational models compared to traditional methods. In the four models, the number of intermediate variables and constraints in the proposed method is reduced by 13.46% / 45.87%, 9.92% / 32.58%, 9.82% / 32.61%, and 11.23% / 37.57% respectively compared to traditional methods. Therefore, the proposed method significantly enhances the solution performance of the operational models, with solution times for the four evaluation models being only 1 / 2.85, 1 / 2.59, 1 / 4.88, and 1 / 4.80 of the traditional methods, respectively. It is worth noting that the improved solution efficiency of the proposed method stems not only from the significantly smaller model scale, but also from the more compact linearization region, which avoids linearizing subregions with excessively large or small slopes, resulting in lower-order linear coefficients. This reduces the overall coefficient range of the evaluation model, further contributing to the improved solution efficiency. The maximum input parameter range of the proposed method is 10. 5 The traditional method is 109 The evaluation objectives solved by the two methods are basically the same, indicating the effectiveness of the method proposed in this invention.

[0145] The frequency deviation extreme values ​​obtained by solving the four models using the frequency security index reconstruction method proposed in this invention are as follows: Figure 4 As shown, the lowest frequency points in different models all remain at the threshold. Within. The accuracy of the proposed adaptive partitioning reconstruction method for frequency security indicators is further verified through frequency response simulation on the Simulink simulation platform for Model 1. At a certain moment, the system frequency simulation is as follows: Figure 5 As shown, the precise simulated value of the frequency deviation extreme obtained through simulation is compared with the linear analytical value obtained from the planning results. The relative error between the two is 0.0263, which is within the predefined linear error threshold of 0.05, indicating the effectiveness of the frequency security index reconstruction method proposed in this invention. Figure 6 The presentation shows the time percentage of each sub-region where the frequency modulation parameters obtained from the four models are located. It can be seen that the system frequency modulation parameters are located in a safe sub-region that does not require linear approximation for most of the time periods. This indicates that the method proposed in this invention has no approximation error for most of the time periods and has higher accuracy than traditional methods.

[0146] In summary, the method proposed in this invention narrows the feasible range of frequency modulation parameters by dividing the region, and significantly reduces the model scale and avoids high-order linearization coefficients by compressing the piecewise linear region. Compared with traditional piecewise linear methods, it has higher accuracy and computational efficiency.

[0147] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for adaptive partitioned reconstruction of frequency security indicators in multi-scenario power systems, characterized in that, Includes the following steps: S1. Under the preset active power disturbance Below, the running times of each running scenario will be... Extreme values ​​of frequency deviation at time As a nonlinear frequency safety indicator, The frequency deviation threshold that does not exceed the maximum frequency required to ensure that the system's safety and stability devices are not triggered To constrain nonlinear frequency safety indicators; based on the system's multi-type frequency regulation resource parameters, under a unified synchronous unit response time constant. Under the given conditions, construct the system's equivalent frequency modulation parameters; based on the system's equivalent frequency modulation parameters, construct the nonlinear frequency safety index constraints and their analytical expressions. The system's equivalent frequency modulation parameters include: system equivalent inertia. Adjustment coefficient and reheat time coefficient ; S2. Obtain , and The maximum feasible range is taken as the feasible range of the original frequency modulation parameters, and sampling is generated within the feasible range of the original frequency modulation parameters. Each sample point and its corresponding data And obtain the corresponding sample surface set. ; S3. Obtain Feasible range of original frequency modulation parameters The corresponding sample surface, traversed and The corresponding sample surface set and The intersection points of the nonlinear frequency safety index and its constraint limit on the feasible range boundary of the original frequency modulation parameters are obtained respectively, and the droop coefficient corresponding to the obtained intersection points is used as the basis for the calculation. and The feasible range of the original frequency modulation parameters is re-partitioned to obtain the partitioned sub-regions; S4. Linearize the sub-regions that need to be linearized and output the linearization parameters that meet the linear error threshold requirements. After reconstructing the nonlinear frequency safety index constraints, the frequency safety index constraints after partition linearization are obtained.

2. The adaptive partitioning reconstruction method for frequency security indicators of a multi-scenario power system according to claim 1, characterized in that, The various types of frequency regulation resources include: synchronous generator units, converter-based new energy generator units, and pumped storage; the parameters of the various types of frequency regulation resources include: the installed capacity of each type of frequency regulation resource, the inertial time constant, primary frequency regulation droop coefficient, governor response time constant and turbine reheat time constant of synchronous generator units, the virtual inertial time constant and virtual droop coefficient of new energy generator units, and the inertial time constant and droop coefficient of pumped storage generator units.

3. The adaptive partitioning reconstruction method for frequency security indicators of a power system in multiple scenarios according to claim 1, characterized in that, The nonlinear frequency safety index constraint in multiple scenarios is expressed as follows: ； The analytical expression for the nonlinear frequency safety index is: ； in, This is the load damping coefficient; , , and These are intermediate variables representing the nonlinear relationship of the equivalent frequency modulation parameters of the coupled system. This represents the moment when the frequency deviation reaches its extreme value.

4. The adaptive partitioning reconstruction method for frequency security indicators of a multi-scenario power system according to claim 3, characterized in that, The specific steps for constructing an analytical expression for a nonlinear frequency security index based on the system's equivalent frequency modulation parameters include: S11. Establish a multi-machine frequency dynamic response model for the system. The first-order differential equation of the system's frequency dynamic response under any given operating scenario is expressed as: ； In the formula, Let the system's equivalent inertia be given in any given operating scenario. For system frequency deviation, , and The first The first synchronous generator unit, the first Taiwan New Energy Unit and the first The primary frequency regulation response power adjustment of the pumped storage power station; , and It is a collection of synchronous generator units, new energy generator units, and pumped storage units; Among them, the primary frequency regulation power adjustment of the equivalent prime mover-governing device for synchronous generators and pumped storage units varies with frequency deviation. The transfer function of change is expressed as: ； ； In the formula, For the synchronous generator speed governor gain, The turbine coefficient is the turbine coefficient of the turbine. The reheater time constant is... For the Laplace operator, This refers to the governor droop coefficient of a pumped storage power station. and These are the response time constants of the pumped storage reversible pump turbine and the guide vanes, respectively. For superposition and Equivalent response time constants of pumped storage units after two different response time constants; New energy generating units based on converters participate in frequency response through virtual droop control. Their primary frequency regulation power model is as follows: ； In the formula, This represents the droop factor of the virtual synchronizer. This is the response time constant of the virtual synchronizer; S12. Response time constant of unified synchronous generator units Then, the multi-machine frequency dynamic response model of the system containing multiple scenarios is aggregated into a system equivalent frequency modulation parameter model that varies with each scenario over time: ； In the formula, , and These are the virtual inertial time constants for synchronous generators, pumped storage units, and new energy generators, respectively. , and This refers to the installed capacity of synchronous generating units, pumped storage units, and new energy generating units. As the baseline capacity, This is the start / stop status variable for the synchronous generator unit; 1 indicates start-up, and 0 indicates stop. This represents the state variable for pumped-storage hydropower generation; it is 1 in the power generation state and 0 in the pumping state. , and These are state variables indicating whether frequency support is provided for synchronous generators, pumped storage units, and new energy generators, respectively. The value is 1 when frequency support is provided and 0 when frequency support is not provided. S13. Based on the system's equivalent frequency modulation parameter model, the nonlinear frequency safety index constraint in the time domain and its analytical expression are derived using the inverse Laplace transform.

5. The adaptive partitioning reconstruction method for frequency security indicators of a multi-scenario power system according to claim 1, characterized in that, The specific content of S2 includes: S21. Based on all current frequency modulation resource parameters of the system , and Calculate the minimum and maximum frequency support schemes to obtain the maximum feasible range of system frequency modulation parameters as the original feasible range of frequency modulation parameters. , and The feasible range of the original frequency modulation parameters is: ； S22. Based on Latin hypercube sampling Internally generate sample points considering all possible unit combinations and resource frequency support schemes, where the first... Each sample point is represented as , and according to Calculate the extreme values ​​of frequency deviation for all sample points. , of which The extreme values ​​of frequency deviation corresponding to each sample point are denoted as ; S23. Adjust the frequency parameters from all sample points. Choose any sample value from the dimensions In determining Obtain the sample surface set under the following conditions: ,in, Indicates the first The extreme coordinates of the frequency deviation corresponding to each sample point It represents the three-dimensional real number space.

6. The adaptive partitioning reconstruction method for frequency security indicators of a power system in multiple scenarios according to claim 1, characterized in that, The specific details of obtaining the intersection point of the nonlinear frequency safety index and its constraint limit on the feasible range boundary of the original frequency modulation parameters in S3 include: S31. Obtain Feasible range of original frequency modulation parameters The corresponding sample surface will and The corresponding set of sample surfaces is defined as and : ； ； In the formula, and They represent and The corresponding extreme coordinates of frequency deviation and They represent and The corresponding adjustment coefficient, and They represent and The corresponding extreme value of frequency deviation; S32. Obtain the nonlinear frequency safety index and its constraint limits. The intersection points on the boundaries of the feasible range of the original frequency modulation parameters include: (1) For the index set Perform initialization, that is, let ,in , Indicates the first The extreme value of frequency deviation corresponding to each sample point Is the value the same as Equal binary indexes, when equal When they are not equal ; (2) Iterate through the subset For all sample points, if the extreme value of the frequency deviation corresponding to the sample point is... If they are equal, then set the current binary index. Update the index set with this value. ,otherwise Update the index set with this value. ; (3) Obtaining satisfaction adjustment coefficient , represented as: ,in, For the present The preceding binary indicator index, corresponding to the sample point That is, the sample points on the system inertia boundary where the nonlinear frequency safety index is equal to the frequency safety constraint limit; (4) Obtain according to steps (1)-(3) Corresponding adjustment coefficient and the corresponding sample points .

7. The adaptive partitioning reconstruction method for frequency security indicators of a multi-scenario power system according to claim 1, characterized in that, In S3, the feasible range of the original frequency modulation parameters is re-partitioned based on the droop coefficients corresponding to the obtained intersection points, resulting in sub-regions. , and They are respectively: ； ； ； in, and They are respectively The minimum and maximum values, and They are respectively The minimum and maximum values, and They are respectively The minimum and maximum values.

8. The adaptive partitioning reconstruction method for frequency security indicators of a power system in multiple scenarios according to claim 7, characterized in that, The specific content in S4 includes: S41. Discard completely unsafe subregions. Preserve a completely secure sub-region For partially secure sub-regions Perform linearization: Linear hyperplane fitting method is used for sub-regions Linearization of the frequency deviation extreme value model within: The linear optimization model is used to fit the extreme value of the nonlinear frequency deviation to the nearest value. The linear form model of the frequency modulation parameters, sub-region Nonlinear frequency deviation extremes The sample surface is approximated as a series of linearly fitted planes; the linear optimization model is: ； in, For the first The approximation error corresponding to each sample point To ensure that the approximate linear result lies on the sample surface, To linearize the error threshold, The first The linear fitting coefficients of each hyperplane are the linearization parameters that satisfy the linear error threshold requirement. The nonlinear frequency safety index constraint is then finally reconstructed as follows: 。

Images