A method for unit commitment of power system considering partition and frequency security constraints

Abstract

The application discloses a kind of power system unit combination methods considering partition and frequency safety constraint, the method is first the unit with same frequency response characteristic is partitioned aggregation, establishes the single-machine aggregation model of frequency response after partitioning;In unit combination problem, the lowest point of frequency is expressed into frequency stability constraint, disturbance power is distributed according to synchronous power coefficient in each sub-region, the initial frequency change rate index of each sub-region is introduced, and the analytical connection between frequency safety constraint and unit combination mode is established;The effectiveness of the model and algorithm is verified in the improved IEEE39 node system.The application adopts the above-mentioned power system unit combination method considering partition and frequency safety constraint, for the frequency safety problem of partitioned system, the distribution of disturbance power between regions is considered in the analytical expression of frequency safety constraint derivation, a partitioned power system safety constraint unit combination model is established and a solving algorithm is proposed.

Description

Technical Field

[0001] This invention relates to the field of frequency control technology for power generating units, and in particular to a method for combining power generating units in a power system that takes into account zoning and frequency security constraints. Background Technology

[0002] my country's power system is gradually developing into a new type of power system, and this new power system is taking shape. As a result, the frequency stability problem caused by low inertia and low primary frequency regulation capability has attracted widespread attention.

[0003] In the context of day-ahead dispatching, adjusting unit output and start-up methods is a crucial means of maintaining the safe and stable operation of the power grid. Considering frequency stability constraints in unit configuration has become a research hotspot.

[0004] Currently, research on power system unit combination considering frequency security constraints mainly follows this approach: constructing a time-domain analytical relationship between frequency extrema after disturbances and unit combination parameters, and embedding it into a power system unit combination optimization model. Regarding the characterization of system frequency characteristics, relevant literature has proposed the SFR model and the ASF model, providing references for future frequency dynamic analysis. Other literature derives frequency extrema, adds frequency stability constraints, and uses piecewise linearization to solve higher-order nonlinear frequency response models.

[0005] In a large power grid, as the proportion of new energy sources continues to increase, the uneven distribution of inertia in the system inevitably leads to changes in the spatiotemporal distribution characteristics of the frequency response after disturbances. Existing publicly available data considering the spatiotemporal distribution characteristics of frequency quantifies and analyzes the frequency characteristics of nodes, using inertia indices to reflect the differences in frequency characteristics among different nodes. Considering the zonal characteristics of the power system's frequency response, it is necessary to partition the system after a disturbance occurs.

[0006] Existing schemes that consider frequency security constraints in unit combination problems are mostly based on the assumption of uniform frequency across the entire grid, while ignoring the spatiotemporal distribution characteristics of grid frequency. From the perspective of considering the spatiotemporal distribution of frequency, additional increments need to be considered: the frequency security constraints of each system after partitioning and the frequency support role between subsystems.

[0007] Regarding the frequency security constraints of each system after partitioning and the frequency support between subsystems, some literature considers the dynamic frequency response (FR) of large interconnected power systems with spatiotemporal characteristics (STC), establishing a two equivalent frequency response (TEFR) model and obtaining its (CFS) closed-form solution, which can be used for analysis and rapid computational analysis, but does not consider the initial rate of change of frequency (RoCoF) index. Other literature considers the maximum rate of change of frequency at the initial stage of disturbances, adding regional inertia constraints to the unit combination problem to reduce the operating cost of the power system, but does not consider the security constraints at the frequency minimum point, thus still having limitations. Some literature proposes a method for constructing frequency security constraints considering the minimum frequency drop value and directly solves for the minimum frequency drop value through Simulnik simulation, but the model accuracy is low. Summary of the Invention

[0008] The purpose of this invention is to provide a power system unit combination method that considers zoning and frequency security constraints. For the frequency security problem of the system after zoning, the method considers the distribution of disturbance power between regions in the analytical expression of frequency security constraints, establishes a zoning power system security constraint unit combination model, and proposes a solution algorithm.

[0009] To achieve the above objectives, this invention provides a power system unit combination method considering zoning and frequency security constraints. The method first groups units with the same frequency response characteristics into zones, establishing a single-unit aggregation model of the power system's frequency response after zoning. Second, in the unit combination problem, the analytical expression for the lowest frequency point is transformed into a frequency stability constraint. Disturbance power is allocated among the sub-regions according to the synchronization power coefficient, and an initial frequency change rate index is introduced for each sub-region, establishing an analytical relationship between frequency security constraints and unit combination methods. Finally, the effectiveness of the model and algorithm is verified in the improved IEEE 39-bus system.

[0010] Preferably, the method includes the following steps: (1) Regional frequency dynamic response and inter-regional disturbance power distribution; (2) Motor system assembly; (3) Frequency security constraint linearization; (4) Model solution and algorithm flow.

[0011] Preferably, in step (1), the regional frequency dynamic response and inter-regional disturbance power allocation divide the synchronous machines and network nodes with coherent frequency dynamic response characteristics into one region, and the power system is divided into multiple regions. In each region, the spatial distribution characteristics of frequency are not considered, and a frequency response model after system disturbance is established. Based on the SFR model, the power sources in the region are aggregated into an equivalent unit, and the inertial time constant of the equivalent unit is... HAdjustment coefficient R High-pressure cylinder power ratio F H and reheat time constant T R The aggregation method is as follows: In the formula: K mg = S g / S sys For each power source, the per-unit parameters are calculated based on the system capacity. λ g Here, n is an intermediate auxiliary variable, representing the number of synchronizers in the region. For time period t, the start-up and shutdown mode of unit g, H g Let g be the inertial time constant of the unit, and R be the inertial time constant. g The droop coefficient of unit g, F Hg For the high-pressure cylinder power ratio of unit g, T Rg For the reheat time constant of unit g, S g S is the capacity of unit g. sys For the system capacity; The dynamic analysis of power system frequency in the s-domain is as follows: in, , , H The inertial time constant, R This is the adjustment coefficient. T R The reheat time constant is D The damping torque coefficient is... F H This refers to the power ratio of the high-pressure cylinder. It is the undamped natural angular frequency. For the damping ratio, K m For mechanical power gain factor, ΔP L For active disturbance; When an equivalent generator unit experiences a disturbance power, the unbalanced power is distributed to each generator node according to the synchronization power coefficient. Based on the generator rotor motion equation and the unbalanced power borne by generator i, the frequency change of the generator is obtained as follows: In the formula, For unbalanced power, Let be the synchronization power coefficient of generator i. Let be the amplitude of the internal electromotive force of generator i; To reduce the susceptance to the potential between node i and node m within the generator. Let f be the initial phase angle difference between the voltages of generator i and node m. i H is the frequency of generator i. i V is the inertial time constant of generator i. m Let be the voltage at node m.

[0012] Preferably, in step (2), the power system combination is based on the regional frequency dynamic response and the power allocation of disturbances between regions. For the power system interconnected by two regions after partitioning, a unit combination model is constructed. Under the premise of ensuring system frequency security, the objective function for minimizing the total power generation cost of the system is as follows: In the formula, T represents the total number of scheduling periods. , , These represent the number of synchronous power sources, wind farms, and photovoltaic power stations, respectively. , , These represent the power generation costs of synchronous power sources, wind farms, and photovoltaic power plants, respectively. , , These represent the active power of synchronous power sources, wind farms, and photovoltaic power plants, respectively. , These represent the costs of synchronous power-on and power-off, respectively. This indicates the start-up and shutdown mode of unit g after the first round of optimization at time t-1. This represents the start-up and shutdown mode of unit g at time t after the first round of optimization. This indicates the start-up and shutdown mode of unit g at time t after optimization.

[0013] Preferably, in step (3), the frequency safety constraint linearization is based on the frequency stability approximation method of hyperplane, which achieves the analysis of the lowest frequency point and key frequency modulation parameters by fitting V hyperplanes.

[0014] Preferably, in step (4), the algorithm for solving the model is as follows: (1) Data preprocessing: organize the relevant parameter data of the power system to be optimized, classify and generate data files of power source, load and tie line, and perform multi-machine aggregation of frequency regulation parameters; (2) Frequency stability constraint: Establish a frequency dynamic analysis model for a high-proportion new energy power system, parse the expression for the lowest frequency point, and use a frequency stability constraint approximation method based on hyperplanes to represent the analytical mapping relationship of key frequency regulation parameters in the frequency extreme point domain through V hyperplanes. (3) Optimize the unit combination method based on the analytical relationship between the conventional power supply start-up mode and the grid frequency response characteristics; (4) Divide the system into regions according to the frequency of each unit based on the node frequency similarity metric; (5) Substitute the result of the first unit combination into any sub-region after partitioning. The unit combination method of another region is still the decision variable. Repeat steps (1)-(3).

[0015] The advantages and positive effects of the power system unit combination method considering zoning and frequency security constraints described in this invention are as follows: 1. Compared with the unified frequency model of the whole network, the frequency response model based on the sub-regions can accurately describe the minimum point and maximum rate of change of the regional frequency of the power system in each sub-region, and has strong robustness.

[0016] 2. Based on the unit combination iterative solution algorithm proposed after considering the frequency security of each sub-region after partitioning, the cooperative relationship between the whole region and sub-region dispatching agencies can be used to complete the orderly verification of multiple frequency security constraints, realize the guarantee and coordination of system capacity, and better maintain the frequency security of the power system.

[0017] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0018] Figure 1 This is a diagram of the frequency response model of the system after disturbance according to the present invention; Figure 2 This is a diagram of the frequency response model of the aggregated new energy power system after disturbance, as presented in this invention. Figure 3 This is a schematic diagram showing the location where the disturbance power of the present invention occurs; Figure 4 This is a hyperplane diagram showing the analytical mapping relationship between the lowest frequency point and key frequency modulation parameters in this invention. Figure 5 This is a graph showing the frequency response curves of each node in this invention; Figure 6 This is a schematic diagram of the partitioning results of the present invention; Figure 7 This is a diagram showing the unit combination results of the present invention without considering frequency security constraints; Figure 8 This is a diagram showing the unit combination results considering frequency security constraints in this invention; Figure 9 This is a diagram showing the unit combination results after considering frequency security constraints following the partitioning of the present invention. Detailed Implementation

[0019] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0020] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0021] Example A power system unit combination method considering zoning and frequency security constraints is proposed. First, units with the same frequency response characteristics are grouped into zones, and a single-unit aggregation model of the frequency response of the power system after zoning is established. Second, in the unit combination problem, the analytical expression of the lowest frequency point is transformed into a frequency stability constraint, the disturbance power is distributed among the sub-regions according to the synchronization power coefficient, and the initial frequency change rate index of each sub-region is introduced to establish an analytical relationship between the frequency security constraint and the unit combination method. Finally, the effectiveness of the model and algorithm is verified in the improved IEEE 39-bus system.

[0022] A power system unit combination method considering zoning and frequency security constraints includes the following steps: (1) Regional frequency dynamic response and inter-regional disturbance power distribution; (2) Motor system assembly; (3) Frequency security constraint linearization; (4) Model solution and algorithm flow.

[0023] (1) Regional frequency dynamic response and inter-regional disturbance power distribution In large interconnected power systems, the differences in the rotational speeds of various synchronous machines lead to different frequencies at each node, resulting in spatial frequency distribution characteristics. However, since some synchronous machines with close electrical and geographical connections may exhibit similar swing characteristics, synchronous machines and nearby nodes will show similar frequency response characteristics. Synchronous machines and network nodes with coherent frequency response characteristics can be grouped into a region, and the power system can be divided into multiple regions. Within each region, frequency uniformity can be assumed.

[0024] Without considering the spatial distribution characteristics of frequency within each sub-region, the frequency response model of the system after disturbance is as follows: Figure 1 As shown in the figure: The system's equivalent inertial time constant is represented by ; D is the damping torque coefficient. , , These are the per-unit parameters for steam turbines, water turbines, and new energy power speed controllers, respectively.

[0025] Based on the SFR model, the power sources within the region are aggregated into an equivalent unit, and the inertial time constant of the equivalent unit is... H Adjustment coefficient R High-pressure cylinder power ratio F H and reheat time constant T R The aggregation method is as follows: In the formula: K mg = S g / S sys For each power source, the per-unit parameters are calculated based on the system capacity. λ g Here, n is an intermediate auxiliary variable, representing the number of synchronizers in the region. For time period t, the start / stop mode of unit g is either 0 or 1, H g Let g be the inertial time constant of the unit, and R be the inertial time constant. g The droop coefficient of unit g, F Hg For the high-pressure cylinder power ratio of unit g, T Rg For the reheat time constant of unit g, S g S is the capacity of unit g. sys This refers to the system's capacity.

[0026] The frequency response model of the aggregated new energy power system after disturbance is as follows: Figure 2 As shown.

[0027] based on Figure 2 The dynamic analysis of power system frequency in the s-domain is as follows: in, , , H The inertial time constant, R This is the adjustment coefficient. T R The reheat time constant is D The damping torque coefficient is... F H This refers to the power ratio of the high-pressure cylinder. It is the undamped natural angular frequency. For the damping ratio, Km For mechanical power gain factor, ΔP L This is a positive disturbance.

[0028] Transforming the above equation to the time domain, we obtain the frequency dynamic time-domain expression based on the system's inertial center coordinate system, as follows: in, , , , This refers to unbalanced power.

[0029] Due to zoning and multi-machine aggregation theory, each region is equivalent to one equivalent generator unit. Within a region, spatial distribution characteristics are ignored; between regions, they are assumed to be connected by uniformly long connecting lines. This is achieved through perturbation power... The allocation is used to characterize the frequency support and connections between regions. Assume the disturbance occurs at any point from the end of the aggregated isostat to the tie line, such as... Figure 3 As shown.

[0030] When an equivalent generator unit experiences a disturbance power, the unbalanced power is distributed to each generator node according to the synchronization power coefficient. Based on the generator rotor motion equation and the unbalanced power borne by generator i, the frequency change of the generator is obtained as follows: In the formula, For disturbance power, Let be the synchronization power coefficient of generator i. Let be the amplitude of the internal electromotive force of generator i; To reduce the susceptance to the potential between node i and node m within the generator. Let f be the initial phase angle difference between the voltages of generator i and node m. i H is the frequency of generator i. i V is the inertial time constant of generator i. m Let be the voltage at node m.

[0031] The active power change sensed in each sub-region can be considered as being through the synchronous power coefficient. Allocation. For a two-region system, this occurs at node m. After the power disturbance, the unbalanced power sensed by the two sub-regions are as follows: In the formula, and These are the synchronous power coefficients of the equivalent generators after aggregation in Region 1 and Region 2, respectively. and For the unbalanced power of region 1 and region 2, This represents the disturbance power.

[0032] Therefore, the frequency change rate of the two regions is: In the formula, and These are the inertial center frequencies of region 1 and region 2, respectively.

[0033] (2) Power system unit combination Based on regional frequency dynamic response and inter-regional disturbance power allocation, a unit combination model is constructed for a power system interconnected between two regions after partitioning. The objective function for minimizing the total system generation cost is as follows, under the premise of ensuring system frequency security: In the formula, T represents the total number of scheduling periods; , , These represent the number of synchronous power sources, wind farms, and photovoltaic power stations, respectively. , , These represent the power generation costs of synchronous power sources, wind farms, and photovoltaic power plants, respectively. , , These represent the active power of synchronous power sources, wind farms, and photovoltaic power plants, respectively. , These represent the costs of synchronous power-on and power-off, respectively. This indicates the start-up and shutdown mode of unit g after the first round of optimization at time t-1. This represents the start-up and shutdown mode of unit g at time t after the first round of optimization. This indicates the start-up and shutdown mode of unit g at time t after optimization.

[0034] Linearizing the absolute value minimizes the change in the boot method between the two optimizations. for Equivalent to Therefore, an additional set of constraints is added: .

[0035] When active power disturbances occur in a power system, to ensure that the dynamic extreme point of the system frequency does not exceed the boundary for maintaining system frequency stability after the disturbance, thus triggering low-frequency load shedding and affecting power quality, the analytical expression of the frequency minimum point based on the SFR model framework can be transformed into a frequency stability constraint, which can improve the frequency stability of the system under extreme disturbances. In the formula: This represents the maximum deviation of the system frequency.

[0036] At the same time, it is also necessary to consider that the initial rate of change of the system's frequency does not exceed the limit in the early stages of the disturbance: After considering the unit combination scheme, the original inertia of the system With the change, the above formula can be rewritten as: In the formula, This represents the unit combination method before partitioning, and RoCoF is the initial frequency change rate.

[0037] After partitioning, each region will have a minimum frequency point and an initial rate of change of frequency (RoCoF). Frequency stability under extreme disturbances must be guaranteed within each region. When considering the constraints of a single region, the unit configuration of the other region before partitioning is used as a parameter in the new constraints. For the initial rate of change of frequency (RoCoF) constraint, the worst-case scenario is that the disturbance occurs at both ends of the tie line, resulting in RoCoFs for both regions, approximating a 0 / 1 distribution of the disturbance between the two regions. The constraint is guaranteed to be satisfied even in the worst-case scenario, and satisfied in all other cases.

[0038] Taking region 1 as a column, the frequency security constraints of a single region after partitioning are listed, where equation (1) represents the RoCoF constraint and equation (2) represents the minimum frequency point constraint.

[0039] (1) (2) in, , .

[0040] In the formula, and The number of generating units in Region 1 and Region 2 respectively. This describes the startup method of the units in Zone 1 during time period t after partitioning. This refers to the startup method of the units in Zone 2 before partitioning during time period t.

[0041] The constraints of steady-state operation of the power system include: node active power balance constraints, i.e., equation (3); node phase angle constraints, line capacity constraints and power flow constraints, i.e., equations (4)-(6); active power output constraints of conventional power sources, wind power, and photovoltaic units and upward and downward reserve constraints, i.e., equations (7)-(13); and conventional power source ramping constraints, i.e., equations (14)-(15).

[0042] (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) in: , , The benchmark active power output for conventional power sources, wind power, and photovoltaic units; , , This refers to the capacity of conventional power sources, wind power, and photovoltaic units; , , This refers to the start-up methods for conventional power sources, wind power, and photovoltaic power. , For the power flow and susceptance of the line; , For node load and phase angle; , This is for the uphill and downhill limits of conventional power supplies.

[0043] (3) Frequency security constraint linearization Frequency safety constraint linearization is based on a hyperplane-based frequency stability approximation method. By fitting v hyperplanes, it achieves analytical analysis of the minimum frequency point and key frequency modulation parameters. For example... Figure 4 As shown in the figure, the blue surface represents the lowest frequency point, while the green, purple, and orange surfaces represent the hyperplanes that characterize the analytical mapping relationship between the lowest frequency point and the key frequency modulation parameters.

[0044] (4) Model solution and algorithm flow The algorithm for solving the model is as follows: (1) Data preprocessing: organize the relevant parameter data of the power system to be optimized, classify and generate data files of power source, load and tie line, and perform multi-machine aggregation of frequency regulation parameters; (2) Frequency stability constraint: Establish a frequency dynamic analysis model for a high-proportion new energy power system, parse the expression for the lowest frequency point, and use a frequency stability constraint approximation method based on hyperplanes to represent the analytical mapping relationship of key frequency regulation parameters in the frequency extreme point domain through V hyperplanes. (3) Optimize the unit combination method based on the analytical relationship between the conventional power supply start-up mode and the grid frequency response characteristics; (4) Divide the system into regions according to the frequency of each unit based on the node frequency similarity metric; (5) Substitute the result of the first unit combination into any sub-region after partitioning. The unit combination method of another region is still the decision variable. Repeat steps (1)-(3).

[0045] Verification Example Based on the IEEE 10-unit 39-bus standard test system, 500MW of photovoltaic power was connected to nodes 1 and 5 respectively, and 500MW of wind power was connected to nodes 14, 16, 18 and 28 respectively, with the proportion of renewable energy installed capacity being 29.5%. The time resolution of the unit combination was set to 1 hour, and the total duration was 24 hours. Active power disturbance was set to 10% of the load level for each time period, and the frequency stability range was 50±0.5Hz. The steady-state operating parameters and frequency regulation parameters of synchronous power and renewable energy power are shown in Table 1 and Table 2 respectively.

[0046] Table 1.Power supply steady-state parameter

[0047] Table 2.Frequency parameters of power supply

[0048] In the BPA software, at time t=2s, the active power of the load connected to node 27 increases by 50MW. Observe the frequency response curves of each node. Figure 5 As shown.

[0049] Area 1: G1, G2, G3, G8, G10 Area 2: G4.G5.G6.G7.G9 Organize the chart according to the partitioning results t Frequency data for the 2-4s time period, and the frequency response curves of corresponding nodes in Region 1 and Region 2 are shown in the figure. For nodes within the same region, their frequency response curves have similar numerical values ​​and oscillation characteristics. However, the frequency response curves of nodes in different regions differ in both numerical value and morphology, indicating that the partitioning result is reasonable. A schematic diagram of the partitioning result is shown below. Figure 6 As shown.

[0050] Verification of the unit combination scheme after partitioning: In order to analyze the impact of frequency security constraints and partitioning effects on the day-ahead scheduling of the system, the following three different schemes are compared.

[0051] Option 1: Traditional unit combination model without considering frequency security constraints.

[0052] Option 2: Add frequency security constraints to Option 1.

[0053] Option 3: Based on Option 2, further divide the area into zones, considering that the inter-zone disturbance power occurs at 1 / 4 of the zone tie line.

[0054] The unit start-up and shutdown statuses of the unit combinations obtained from the above three schemes are as follows: Figures 7-9 As shown in the diagram, gray represents generator units in operation, and white represents generator units in shutdown. In Scheme 2, the number of generator units in operation increases significantly, which can prevent frequency out-of-bounds violations and meet frequency safety constraints. Compared with Scheme 2, in Scheme 3, the disturbance power allocated to Region 1 is greater, resulting in more prominent frequency safety issues. Therefore, the number of generator units in operation in Region 1 is increased compared to Scheme 2, while the number of generator units in operation in Region 2 is decreased compared to Scheme 2.

[0055] Table 3 shows a comparison of the power generation costs for the three schemes. Scheme 3 saw the largest increase in power generation cost, while Scheme 1 saw the smallest. This is because, in order to ensure the regional frequency after the disturbance, Schemes 2 and 3 both added more generator units, leading to an increase in power generation costs.

[0056] Table 3. Comparison of operating costs by different solutions

[0057] A system frequency response simulation model was built based on the Matlab-Simulink platform. The unit combination schemes of the three schemes were substituted into the built model. The time period 20 with different start-up methods was selected for verification. The minimum frequency point and maximum frequency change rate (RoCoF) of the three schemes are shown in Table 4.

[0058] Table 4. Comparison of the frequency nadir and RoCoF by different solutions

[0059] Compared to Schemes 2 and 3, Scheme 1 does not consider frequency security constraints, resulting in the largest minimum frequency point and maximum frequency change rate. In Scheme 2, the minimum frequency point is improved, but RoCoF is affected by the distribution of disturbance power between regions. In Scheme 3, which considers regional division, the regional frequency change rate constraint and the regional minimum frequency point constraint play a good role in ensuring regional frequency security.

[0060] Therefore, this invention adopts the above-mentioned power system unit combination method considering zoning and frequency security constraints. For the frequency security problem of the system after zoning, the method considers the distribution of disturbance power between regions in the analytical expression of frequency security constraints, establishes a zoning power system security constraint unit combination model, and proposes a solution algorithm.

[0061] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for combining generating units in a power system considering zoning and frequency security constraints, characterized in that, The method involves first grouping generator units with the same frequency response characteristics into sub-regions to establish a single-unit aggregation model of the frequency response of the power system after partitioning; secondly, in the generator unit combination problem, the analytical expression of the lowest frequency point is transformed into a frequency stability constraint, the disturbance power is distributed among the sub-regions according to the synchronization power coefficient, an initial frequency change rate index is introduced for each sub-region, and an analytical relationship is established between the frequency security constraint and the generator unit combination method; finally, the effectiveness of the model and algorithm is verified in the improved IEEE 39-bus system. The method includes the following steps: (1) Regional frequency dynamic response and inter-regional disturbance power distribution; The synchronous machine and network node with the same frequency dynamic response characteristics are divided into a region, and the power system is divided into multiple regions. The spatial distribution characteristics of frequency are not considered in each region. The frequency response model after system disturbance is established. Based on the SFR model, the power sources in the region are aggregated into an equivalent unit. (2) Generator system combination: Based on the regional frequency dynamic response and inter-regional disturbance power distribution, a generator combination model is constructed for the power system interconnected by two regions after partitioning. The objective function for minimizing the total power generation cost of the system is constructed under the premise of ensuring system frequency security as follows: ； In the formula, T represents the total number of scheduling periods. , , These represent the number of synchronous power sources, wind farms, and photovoltaic power stations, respectively. , , These represent the power generation costs of synchronous power sources, wind farms, and photovoltaic power plants, respectively. , , These represent the active power of synchronous power sources, wind farms, and photovoltaic power plants, respectively. , These represent the costs of synchronous power-on and power-off, respectively. This indicates the start-up and shutdown mode of unit g after the first round of optimization at time t-1. This represents the start-up and shutdown mode of unit g at time t after the first round of optimization. This indicates the start-up and shutdown mode of unit g at time t after optimization. (3) Frequency safety constraint linearization: Based on the hyperplane frequency stability approximation method, the analysis of the minimum frequency point and key frequency modulation parameters is achieved by fitting V hyperplanes; (4) Model solution and algorithm flow: (4.1) Data preprocessing: organize the relevant parameter data of the power system to be optimized, classify and generate data files of power sources, loads and tie lines, and perform multi-machine aggregation of frequency regulation parameters; (4.2) Frequency stability constraint: Establish a frequency dynamic analysis model for a high-proportion new energy power system, parse the expression for the lowest frequency point, and use a frequency stability constraint approximation method based on hyperplanes to represent the analytical mapping relationship of key frequency regulation parameters in the frequency extreme point domain through V hyperplanes. (4.3) Based on the analytical relationship between the conventional power supply start-up mode and the grid frequency response characteristics, the unit combination mode is optimized; (4.4) Divide the system into regions according to the frequency of each unit based on the node frequency similarity metric; (4.5) Substitute the result of the first unit combination into any sub-region after partitioning. The unit combination method of another region is still the decision variable. Repeat steps (4.1)-(4.3).

2. The power system unit combination method considering zoning and frequency security constraints according to claim 1, characterized in that: In step (1), the inertial time constant of the equivalent unit H Adjustment coefficient R High-pressure cylinder power ratio F H and reheat time constant T R The aggregation method is as follows: In the formula: K mg = S g / S sys For each power source, the per-unit parameters are calculated to the system capacity. λ g Here, n is an intermediate auxiliary variable, representing the number of synchronizers in the region. For time period t, the start-up and shutdown mode of unit g, H g Let g be the inertial time constant of the unit, and R be the inertial time constant. g The droop coefficient of unit g, F Hg For the high-pressure cylinder power ratio of unit g, T Rg For the reheat time constant of unit g, S g S is the capacity of unit g. sys For the system capacity; The dynamic analysis of power system frequency in the s-domain is as follows: in, , , H The inertial time constant, R This is the adjustment coefficient. T R The reheat time constant is D The damping torque coefficient is... F H This refers to the power ratio of the high-pressure cylinder. It is the undamped natural angular frequency. For the damping ratio, K m For mechanical power gain factor, ΔP L For active disturbance; When an equivalent generator unit experiences a disturbance power, the unbalanced power is distributed to each generator node according to the synchronization power coefficient. Based on the generator rotor motion equation and the unbalanced power borne by generator i, the frequency change of the generator is obtained as follows: In the formula, For unbalanced power, Let be the synchronization power coefficient of generator i. Let be the amplitude of the internal electromotive force of generator i; To reduce the susceptance to the potential between node i and node m within the generator. Let f be the initial phase angle difference between the voltages of generator i and node m. i H is the frequency of generator i. i V is the inertial time constant of generator i. m Let be the voltage at node m.

Images