A frequency-voltage coupling and hierarchical zoning support analysis method considering main and auxiliary coordination

Abstract

The application belongs to the technical field of power system analysis, and discloses a frequency-voltage coupling and layered partition support analysis method considering main distribution coordination, which is applied to a power system containing a main network and a distribution network, and comprises the following steps: a main network frequency-voltage coupling mechanism model is constructed; the distribution network is divided into multiple partitions equivalent to virtual power plants, and equivalent parameters of each partition are extracted; an optimization model for analyzing the support capacity of the distribution network is constructed, with the goal of maximizing the active and reactive power support of each partition to the main network and the constraint of safe operation of the system; and the optimization model is solved by using an external point penalty function combined with a particle swarm algorithm to obtain the quantitative support capacity of each partition to the main network. The method systematically depicts the interaction mechanism of the main and distribution networks, reduces the calculation complexity through layered partition equivalence, converts the support capacity analysis into an optimization problem with constraints, and solves the optimization problem by using an external point penalty function combined with a particle swarm algorithm, so that the support capacity is accurately quantified, and the accuracy of power grid stability analysis and the feasibility of dispatching decision are improved.

Description

Technical Field

[0001] This invention belongs to the field of power system analysis technology, specifically relating to a frequency-voltage coupling and hierarchical partitioning support analysis method that considers primary and secondary coordination. Background Technology

[0002] Existing power grid stability analysis frameworks have significant shortcomings. First, at the modeling level, traditional methods typically ignore the coupling effects of frequency and voltage, or simply equate the distribution network to a load with constant impedance / power, failing to reflect the active support characteristics and coupling interaction mechanisms of distributed generation within the distribution network when frequency or voltage disturbances occur in the power grid. Second, in terms of analysis scope, the analysis of the main grid and distribution network has long been fragmented, lacking effective collaborative modeling methods and failing to quantitatively assess the distribution network's overall support capacity limit for the main grid. Furthermore, from a computational efficiency perspective, performing integrated detailed modeling and simulation of a transmission and distribution network containing tens of thousands of nodes is computationally unsustainable and cannot meet the needs of online analysis and real-time decision-making.

[0003] From a mathematical perspective, the stability analysis of a system under primary-distribution coordination is a complex dynamic system analysis problem with high dimensionality, strong nonlinearity, and multiple time scales. The frequency-voltage coupling relationship is reflected in the differential-algebraic equations of the distributed power controller, while the support capacity analysis requires solving an optimization problem with a maximum capacity boundary while satisfying the power flow constraints and equipment operation constraints within the distribution network. This demands that the model accurately describe the system's physical mechanisms and achieve computational feasibility through effective simplification methods.

[0004] Currently, research in this field largely focuses on a single technical direction, either considering only frequency stability or voltage stability, and is often limited to the main grid or distribution network level. Although some studies have attempted joint simulation of transmission and distribution networks, most are computationally intensive and fail to provide clear quantitative indicators of "support capabilities," making them difficult to directly apply to online decision support in dispatch centers. Existing methods have limitations in terms of the completeness of mechanistic models, the synergy of analytical frameworks, and the practicality of computational results.

[0005] Therefore, there is an urgent need to establish an analytical method that can accurately characterize the frequency-voltage coupling interaction mechanism, connect the main and distribution network levels, and take into account both computational efficiency and engineering practicality. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a method, equipment, and medium for analyzing the support capacity of distribution networks. The method systematically characterizes the frequency-voltage coupling interaction mechanism between the main and distribution networks, breaking through the limitations of traditional decoupling analysis. Simultaneously, it simplifies the complex distribution network into a virtual power plant through hierarchical and partitioned equivalence, significantly reducing computational complexity. The support capacity analysis is transformed into a constrained optimization problem and solved using an external penalty function combined with a particle swarm optimization algorithm. This enables precise quantification of the distribution network's support capacity to the main grid, providing core data support for coordinated control of the main and distribution networks, and improving the accuracy of grid stability analysis and the feasibility of dispatching decisions.

[0007] This invention provides the following technical solution: The primary objective of this invention is to provide a frequency-voltage coupling and hierarchical zoning support analysis method considering main grid and distribution network coordination, applicable to power systems comprising a main grid and a distribution network, including: Construct a main grid frequency-voltage coupling mechanism model; The distribution network is divided into multiple zones that are equivalent to virtual power plants, and the equivalent parameters of each zone are extracted. With the goal of maximizing the active and reactive power support of each zone to the main grid and with the constraint of safe system operation, a distribution network support capacity analysis and optimization model is constructed. The constraints in the optimization model are handled by using the external point penalty function, and the optimization model is solved by combining the particle swarm optimization algorithm to obtain the quantified support capability of each partition for the main network.

[0008] As a further improvement of the present invention, the construction of the main grid frequency-voltage coupling mechanism model includes: The relationship between the active power output of the distributed power source and the system frequency deviation, as well as the relationship between the reactive power output and the voltage deviation at the grid connection point, are constructed using the droop control principle. By introducing distributed generation operation constraint equations that limit the active and reactive power outputs of distributed generation sources, a main grid frequency-voltage coupling mechanism model is obtained to characterize the dynamic coupling and mutual constraint relationship between the active and reactive power outputs of distributed generation sources under system disturbances.

[0009] By establishing a droop control model for distributed power sources and introducing operating capacity constraints, the dynamic coupling and mutual constraint relationship between active and reactive power outputs under system disturbances is systematically described. This overcomes the prediction bias caused by forced decoupling in traditional methods and provides a theoretical basis that is closer to the actual physical characteristics of high-proportion renewable energy power grids.

[0010] As a further improvement of the present invention, the distribution network division includes: An impedance model is constructed based on data from distribution network nodes, branches, loads, and distributed power sources. The electrical correlation strength between nodes in the distribution network is analyzed based on the impedance model, and an electrical distance matrix is ​​constructed. The power distribution network is divided into multiple zones based on the electrical distance and scale criteria between nodes, ensuring that each zone is tightly coupled and the inter-zone connections are sparse.

[0011] As a further improvement of the present invention, the equivalent parameters include: equivalent impedance, maximum active power support capacity, maximum reactive power support capacity, equivalent inertia time constant, and overall response delay time.

[0012] By adopting the equivalent modeling approach of hierarchical partitioning, the complex distribution network is equivalent to a "virtual power plant" with clear external characteristics. The frequency and voltage support capabilities that the power plant can provide to the upper-level main grid under different operating conditions are accurately quantified. This provides core data support and theoretical basis for dispatching departments to perceive the system stability boundary, stimulate the regulation potential of distributed resources, and formulate collaborative control strategies.

[0013] As a further improvement of the present invention, the objective function of the optimization model is to maximize the weighted sum of the incremental active power support and the incremental reactive power support of the partition, and the priority of the incremental active power support and the incremental reactive power support of the partition is adjusted by the weight coefficient.

[0014] As a further improvement of the present invention, the constraints include: nodal power flow equations, linearized AC power flow equations and branch power transmission equations constructed based on the nodal admittance matrix, injected power balance constraints, nodal voltage offset range constraints, and connection point voltage and frequency boundary condition constraints given by the main network layer.

[0015] As a further improvement of the present invention, the step of using an exterior penalty function to handle constraints in the optimization model and combining it with a particle swarm optimization algorithm to solve the optimization model includes: Individual particles are constructed based on column vectors of active power reference value, reactive power reference value, voltage reference value, frequency reference value, active power change, and reactive power change; A particle population is constructed based on the upper and lower bound matrices of the preset particle position vector components and velocity components. The constraint is transformed into a penalty term by using an external penalty function and added to the objective function to construct the fitness function; Based on the fitness function, the individual extreme value and the global extreme value are used to guide the particle to update its position, and the process is iterated until all particles have found the optimal position.

[0016] The exterior point penalty function method is cleverly used to handle a large number of nonlinear constraints. Combined with the global search capability of the particle swarm optimization algorithm, it effectively solves the problem of solving high-dimensional and strongly constrained optimization problems, ensuring the feasibility and convergence speed of the calculation results and enhancing the engineering practicality of complex models.

[0017] Solving the optimization model yields a standardized equivalent parameter vector. The main grid dispatch center can then quickly construct a simplified model of the entire system and perform dynamic simulations, thereby formulating control strategies with higher coordination and lower risk, significantly enhancing the grid's resilience to disturbances. The final output is clear and quantifiable, directly applicable to the dispatch center's decision support system, the planning department's adaptability assessment, and operational mode formulation, providing a universal solution for the coordinated planning, operation, and control of the main and distribution networks under the new power system context.

[0018] As a further improvement of the present invention, when the particle position vector component and / or velocity component exceed the preset upper and lower bound matrices during the search process, a correction function based on uniformly distributed random numbers is used to reset them back to the preset upper and lower bound matrices.

[0019] A second objective of this invention is to provide a computer device comprising at least one processing unit and at least one storage unit, wherein the storage unit stores a computer program, and when the program is executed by the processing unit, the processing unit performs the aforementioned method.

[0020] A third objective of this invention is to provide a computer-readable storage medium storing a computer program executable by an electronic device, which, when run on the electronic device, causes the electronic device to perform the above-described method.

[0021] Compared with the prior art, the beneficial effects of the present invention are as follows: The method systematically characterizes the frequency-voltage coupling interaction mechanism between the main and distribution networks, breaking through the limitations of traditional decoupling analysis. At the same time, it simplifies the complex distribution network into a virtual power plant through hierarchical and partitioned equivalence, significantly reducing computational complexity. It transforms the support capacity analysis into a constrained optimization problem and uses an external penalty function combined with a particle swarm optimization algorithm to solve it, realizing the accurate quantification of the distribution network's support capacity for the main network. This provides core data support for the coordinated control of the main and distribution networks, improving the accuracy of power grid stability analysis and the feasibility of dispatching decisions.

[0022] By establishing a droop control model for distributed power sources and introducing their operating capacity constraints, the dynamic coupling relationship between active and reactive power outputs is accurately described, overcoming the limitations of traditional decoupling analysis methods and providing a more realistic theoretical basis for the stability analysis of high-proportion renewable energy power grids.

[0023] By dividing the distribution network into multiple electrically coupled zones and simplifying them by equivalent means, the complex "source-grid-load" system is mapped into a small number of "virtual power plants" with distinct characteristics. This greatly reduces the computational complexity of global stability analysis in the main grid dispatch center, making it suitable for online analysis and real-time decision-making.

[0024] The calculation of support capacity is transformed into an optimization problem with complex constraints. After solving it, the solution can output clear quantitative indicators such as the maximum active and reactive power support increment that each partition can provide under the current operating state. This enables the scheduling department to accurately perceive the system stability boundary and effectively utilize distributed resources.

[0025] The exterior point penalty function method is cleverly used to handle a large number of nonlinear constraints. Combined with the global search capability of the particle swarm optimization algorithm, it effectively solves the problem of solving high-dimensional and strongly constrained optimization problems, ensuring the feasibility and convergence speed of the calculation results and enhancing the engineering practicality of complex models. Attached Figure Description

[0026] Figure 1 A flowchart of the method provided by the present invention; Figure 2 The flowchart shows how to solve the optimization model using the particle swarm optimization algorithm. Figure 3 This is the main distribution coordination power grid structure diagram used in the experiment; Figure 4 The diagram shows the simulation results of the PCC current of each model obtained from the experiment. Detailed Implementation

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

[0028] The present invention will now be described in further detail with reference to the accompanying drawings: like Figure 1 As shown, this embodiment provides a frequency-voltage coupling and hierarchical zoning support analysis method considering main grid and distribution network coordination, applicable to power systems including main grid and distribution network, including: Construct a main grid frequency-voltage coupling mechanism model; The distribution network is divided into multiple zones that are equivalent to virtual power plants, and the equivalent parameters of each zone are extracted. With the goal of maximizing the active and reactive power support of each zone to the main grid and with the constraint of safe system operation, a distribution network support capacity analysis and optimization model is constructed. The constraints in the optimization model are handled by using the external point penalty function, and the optimization model is solved by combining the particle swarm optimization algorithm to obtain the quantified support capability of each partition for the main network.

[0029] The construction of the main grid frequency-voltage coupling mechanism model includes: Construct the droop control equation using the droop control principle: ; ; In the formula, Indicates the first The actual output active power of a distributed power source; Indicates the first Reference value of active power for each distributed power source; Indicates the first Active-frequency droop factor of a distributed power source; Indicates the first The rated frequency of a distributed power source; Indicates the first The measured system frequency of a distributed power source; Indicates the first The actual output reactive power of a distributed power source; Indicates the first Reference value of reactive power for a distributed power source; Indicates the first The reactive power-voltage droop factor of a distributed power source; Indicates the first The rated voltage of a distributed power source; Indicates the first Measured grid connection voltage of a distributed power source; Establish the operating constraint equations for distributed power sources: ; In the formula, This indicates the apparent power capacity limit of distributed power sources.

[0030] The model can accurately depict the dynamic coupling and mutual constraint relationship between the active and reactive power outputs of distributed power sources under system disturbances, laying a theoretical foundation for the analysis of interaction mechanisms. Based on the distribution network topology and electrical coupling relationship, the distribution network is divided into Each partition contains several nodes, lines, loads, and distributed power sources.

[0031] The specific partitioning steps are as follows: collect data on all nodes, branches, loads, and power sources in the distribution network, and construct an impedance-based network model; analyze the electrical correlation strength between nodes in the distribution network based on the impedance model to form an electrical distance matrix; divide the network into multiple tightly coupled partitions based on the distance and scale criteria; verify the topological connectivity of each partition and adjust the boundary nodes to ensure sparse connections between partitions; finally, output a complete partitioning scheme containing node affiliation, boundary information, and connection points.

[0032] For each distribution network zone Define its equivalent parameter vector : ; In the formula, and These represent the distribution network zones. The resistive and reactive components of the equivalent impedance; and These represent the distribution network zones. The maximum active and reactive power support capacity that it can provide to the outside world; Indicates distribution network partition The equivalent inertial time constant; Indicates distribution network partition The overall response delay time.

[0033] The construction of the distribution network support capacity analysis and optimization model includes: Define the objective function of the optimization model: ; ; ; In the formula, This indicates the incremental total active power support of the partition; This indicates the increment of total reactive power support in the partition; These are weighting coefficients used to adjust the priority of active and reactive power support; and Representing nodes respectively The changes in injected active and reactive power; Establish the nodal power flow equations of the distribution network based on the nodal admittance matrix: The nodal admittance matrix is ​​expressed as: ; In the formula, This represents the node admittance matrix of the distribution network; The number of nodes in the mainnet; and Representing nodes respectively and nodes The nodal conductance and nodal susceptance between them; The nodal power flow equations of a distribution network are expressed as: ; ; ; In the formula, and Representing nodes respectively The injected active power and injected reactive power; and Representing nodes respectively The magnitude and phase angle of the node voltage; The linearized AC power flow equations of the distribution network are expressed as: ; ; In the formula, and Representing nodes respectively The changes in voltage phase angle and amplitude; Establish a system branch power transmission model: ; ; In the formula, Represents a node and nodes The corresponding active power transmitted on the line; Represents a node and nodes The corresponding reactive power transmitted on the line.

[0034] Establish system operation constraints, including injected power balance constraints and node voltage offset range constraints, expressed as follows: ; ; ; In the formula, Represents a node The rated voltage at the location; and They are respectively at the node Active and reactive loads at the location; Establish boundary condition constraints: ; ; In the formula, Indicates distribution network partition Voltage at the connection point with the main grid; Indicates distribution network partition Frequency of connection to the main network; and These represent the boundary conditions for voltage and frequency given by the main network layer, respectively.

[0035] like Figure 2 As shown, the optimization model solution includes: Particle individual structure: ; In the formula, It represents the continuous decision variable in the optimization model, and also the particle population used for iteration in the optimization algorithm; The bolded letters in the middle represent a column vector consisting of the values ​​of all the scalars involved in the corresponding physical quantity; Particle population construction: ; ; ; ; In the formula, Indicates the first In the next iteration, the population of the first The position vector of each particle; for The corresponding particle velocity; , These are the upper and lower bound matrices, respectively, composed of the upper and lower bounds corresponding to each component of the position vector. , These are the upper and lower bound matrices composed of the upper and lower bounds corresponding to each velocity component. Indicates the first The position vector matrix obtained in the next iteration Indicates the first The velocity matrix obtained from the next iteration; Fitness function construction: ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; In the formula, Represents the fitness function; and They represent the first The first penalty factor and the first One penalty function; This represents the total number of distributed power sources. Particle position update: Individual extreme values The global extremum represents the best position found in the history of a particle. This represents the optimal position found in the history of all particles in the entire particle swarm.

[0036] Furthermore, during the iteration process, the particle's position vector and velocity are updated according to the following criteria: ; ; in, , For learning factors; , Let them be two random vectors, each with components in the range [0,1]. This is the weight vector; The adaptive expression for the weight vector is: ; In the formula, , They are respectively The upper and lower bounds, This represents the maximum number of iterations. Boundary condition handling: For position vector components and velocity components that exceed the corresponding components of the boundary matrix during the search process, the following correction method is used: ; ; In the formula, , They represent the first Upper and lower bounds of the unit vector components. , They represent the first Upper and lower bounds of each velocity component It is a random number that is uniformly distributed between the interval [0,1].

[0037] Solving the optimization model yields a standardized equivalent parameter vector. The main grid dispatch center can then quickly construct a simplified model of the entire system and perform dynamic simulations, thereby formulating control strategies with higher coordination and lower risk, significantly enhancing the grid's resilience to disturbances. The final output is clear and quantifiable, directly applicable to the dispatch center's decision support system, the planning department's adaptability assessment, and operational mode formulation, providing a universal solution for the coordinated planning, operation, and control of the main and distribution networks under the new power system context.

[0038] To verify the effectiveness of the distribution network support capacity analysis method provided in this embodiment in equivalent modeling and support capacity assessment, a comparative simulation analysis was conducted in a typical main-distribution coordinated system with a high proportion of distributed generation. The main-distribution coordinated power grid structure used in the experiment is as follows: Figure 3 As shown, based on the constructed main grid frequency-voltage coupling mechanism model and the distribution network hierarchical partitioning equivalent method, the distributed power sources and loads within the distribution network are jointly clustered according to electrical distance and dynamic characteristics, resulting in three equivalent partitions. Parameters such as equivalent impedance, maximum active power support capacity, maximum reactive power support capacity, equivalent inertia time constant, and overall response delay time are extracted for each partition, forming a simplified virtual power plant model oriented towards the main grid dispatch center. In the simulation, a three-phase short-circuit fault is set at the point of common coupling (PCC), with the fault occurring at 1 second and lasting for 0.5 seconds. The voltage and current transient response curves at PCC of the detailed physical model and the equivalent model proposed in this embodiment are recorded, and compared with the equivalent model A obtained by considering only electrical distance clustering, the equivalent model B obtained by considering only dynamic characteristics clustering, and the traditional single-unit power summation equivalent model.

[0039] Table 1 Comparison of simulation accuracy of various equivalent models

[0040] Regarding the verification of the accuracy of equivalent modeling, the simulation accuracy comparison of each model is shown in Table 1. As shown in Table 1, the equivalent model proposed in this embodiment exhibits good consistency with the detailed physical model in both voltage and current response. Specifically, the voltage simulation accuracy error of the equivalent model proposed in this embodiment is 0.0011, lower than 0.0044 for the single-machine equivalent model, 0.0016 for equivalent model A, and 0.0037 for equivalent model B; in terms of current response, the simulation accuracy error of the proposed equivalent model is 0.0121, lower than 0.0882 for the single-machine equivalent model, 0.0581 for equivalent model A, and 0.0421 for equivalent model B. These results demonstrate that the main grid frequency-voltage coupling mechanism model and the distribution network hierarchical and zonal equivalent method established in this invention can effectively characterize the dynamic support characteristics of the distribution network and maintain high analytical accuracy while reducing model size. The PCC response results of each model are as follows: Figure 4 As shown.

[0041] To further verify the applicability of the method provided in this embodiment under different distributed power generation spatial layout scenarios, four distributed power generation configuration schemes were set up respectively, namely upstream concentration, downstream concentration, uniform distribution, and single-area extreme concentration, while maintaining a consistent total penetration rate. The main grid frequency-voltage coupling mechanism model and distribution network support capacity analysis and optimization model constructed in this embodiment were applied to calculate the dynamic response indicators of the main grid and distribution network under each scheme. The calculation results under different distributed power generation spatial layouts are shown in Table 2.

[0042] Table 2 System response indicators under different distributed power supply spatial layout schemes

[0043] As shown in Table 2, under the condition of the same total distributed power penetration capacity, the lowest grid-side frequency corresponding to the uniform distribution scheme is 49.720Hz, which is higher than 49.690Hz for the upstream centralized scheme, 49.660Hz for the single-area extreme centralized scheme, and 49.630Hz for the downstream centralized scheme. Its grid-side frequency change rate RoCoF is 0.230Hz / s, which is lower than 0.260Hz / s for the upstream centralized scheme, 0.290Hz / s for the single-area extreme centralized scheme, and 0.310Hz / s for the downstream centralized scheme. From the perspective of distribution network response, the local maximum voltage deviation of the uniform distribution scheme is 0.028pu, which is lower than 0.037pu for the upstream centralized scheme, 0.055pu for the single-area extreme centralized scheme, and 0.061pu for the downstream centralized scheme. Its lowest voltage on the distribution network side is 0.439pu, the main-distribution interaction support response delay is 0.160s, and the active power support peak reaches 18.7MW. In contrast, while the downstream centralized scheme has a certain active power support capability, its local maximum voltage deviation and support response delay are both relatively large; the peak active power support of the single-zone extreme centralized scheme reaches 19.4MW, but its local voltage deviation is still higher than that of the uniform distribution scheme. Therefore, under the operating conditions set in this embodiment, the uniform distribution scheme performs better in terms of main grid frequency support, distribution network voltage stability, and overall performance of main grid-distribution coordination.

[0044] To verify the application effect of the method of the present invention in the scenario of main grid-distribution network coordinated control, four operating modes were further set up: main grid support only, distribution network local support only, main grid-distribution network coordinated support, and main grid-distribution network coordinated hierarchical and partitioned equivalent support. A simplified simulation environment for the main grid dispatch center was constructed using the equivalent parameter vector obtained from the optimization model of the present invention. The system operating indicators under different control modes are shown in Table 3.

[0045] Table 3 Comparison of system operation indicators under different control methods

[0046] As shown in Table 3, compared with the main grid-only support method, the adoption of the combined main and distribution network-based hierarchical and zoned equivalent support method increases the minimum frequency on the main grid side from 49.570Hz to 49.760Hz, an increase of 0.190Hz; the main grid side frequency change rate RoCoF decreases from 0.300Hz / s to 0.180Hz / s. The minimum voltage on the distribution network side increases from 0.880pu to 0.960pu, the voltage recovery time is shortened from 5.8s to 3.4s, and the number of voltage over-limit nodes decreases from 8 to 2. From the perspective of main-distribution network interaction and economic indicators, the tie-line impact power decreases from 21.4MW to 12.8MW, and the comprehensive control cost decreases from 236,000 yuan to 169,000 yuan.

[0047] Further comparison shows that the local support method alone can improve the voltage recovery of the distribution network to a certain extent, increasing the minimum voltage to 0.910 pu, shortening the voltage recovery time to 5.0 s, and reducing the number of voltage over-limit nodes to 6. However, its minimum frequency on the main grid side is 49.610 Hz, RoCoF is 0.270 Hz / s, the tie-line impact power is 18.9 MW, and the comprehensive control cost is 208,000 yuan. The main grid-distribution coordinated support method further improves the system operation indicators, increasing the minimum frequency on the main grid side to 49.690 Hz, reducing RoCoF to 0.220 Hz / s, increasing the minimum voltage on the distribution network side to 0.940 pu, shortening the voltage recovery time to 4.1 s, reducing the number of voltage over-limit nodes to 4, reducing the tie-line impact power to 15.7 MW, and reducing the comprehensive control cost to 187,000 yuan. The above results indicate that the method provided in this embodiment can provide quantitative distribution network support capability information to the main grid dispatching side, thereby providing an analytical basis for main grid-distribution coordinated control.

[0048] The method provided in this embodiment enables quantitative analysis of the distribution network's support capacity in a main grid-distribution network coordination scenario. By introducing a main grid frequency-voltage coupling mechanism model, a hierarchical partitioning equivalent method, and a solution strategy combining external point penalty functions with particle swarm optimization, it can balance model accuracy and computational efficiency, and provide support for main grid dispatching side to conduct main grid-distribution network coordination analysis and control. By establishing a distributed generation droop control model and introducing operating capacity constraints, the dynamic interaction relationship between the main grid and the distribution network can be modeled; through the hierarchical partitioning equivalent method, complex distribution networks can be simplified into equivalent partitioning models oriented towards dispatching analysis; through optimized solutions, the quantitative results of the distribution network's support capacity for the main grid can be obtained, thereby improving the engineering application feasibility of main grid-distribution network coordination operation analysis.

[0049] In summary, compared with existing technologies, the method provided in this embodiment achieves significant breakthroughs in main grid-distribution network collaborative modeling and power grid stability analysis. Addressing issues such as deep frequency and voltage coupling under high-proportion renewable energy grid integration, complex interaction mechanisms between the main grid and distribution network, and difficulties in global analysis and calculation, an integrated analysis framework is constructed that incorporates frequency-voltage coupling characteristics, hierarchical and partitioned equivalence, and quantification of support capabilities. This framework accurately describes the dynamic interaction and support relationship between the main grid and distribution network. By establishing a distributed generation droop control model and introducing operational capacity constraints, the limitations of traditional decoupling analysis methods are overcome. Furthermore, the use of a partitioned equivalence method simplifies the complex distribution network into a well-defined "virtual power plant," significantly improving computational efficiency and analytical feasibility.

[0050] Furthermore, the method presented in this embodiment innovatively proposes an intelligent solution strategy based on an exterior point penalty function and a particle swarm optimization algorithm. Compared to traditional optimization methods, this strategy effectively handles a large number of nonlinear operational constraints through a penalty function mechanism, and combined with the global search capability of the particle swarm optimization algorithm, ensures efficient solution to complex optimization problems and the engineering applicability of the results. Through a hierarchical partitioning architecture and standardized equivalent parameter output, unified analysis and collaborative simulation at the main and distribution network levels are achieved, significantly enhancing the overall resilience of the power grid to disturbances.

[0051] The method provided in this embodiment systematically solves the problems of unclear interaction mechanism between main and distribution networks, difficulty in quantifying support capabilities, and complexity of analysis and calculation in the prior art through the synergistic innovation of coupled mechanism modeling, hierarchical equivalent framework and efficient optimization algorithm. It provides a theoretical basis and practical tools for the coordinated operation and control of main and distribution networks under the background of new power system, and has important value for improving the safety and stability level of power grid and the renewable energy absorption capacity.

[0052] This embodiment provides a computer device, including at least one processing unit and at least one storage unit, wherein the storage unit stores a computer program, and when the program is executed by the processing unit, the processing unit performs the above-described method.

[0053] This embodiment provides a computer-readable storage medium storing a computer program executable by an electronic device, which, when run on the electronic device, causes the electronic device to perform the above-described method.

[0054] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A frequency-voltage coupling and hierarchical zoned support analysis method considering main coordination, applied to a power system comprising a main grid and a distribution grid, characterized in that, include: Construct a main grid frequency-voltage coupling mechanism model; The distribution network is divided into multiple zones that are equivalent to virtual power plants, and the equivalent parameters of each zone are extracted. The equivalent parameters include: equivalent impedance, maximum active power support capacity, maximum reactive power support capacity, equivalent inertia time constant, and overall response delay time. With the goal of maximizing the active and reactive power support of each zone to the main grid and with the constraint of safe system operation, a distribution network support capacity analysis and optimization model is constructed. The constraints in the optimization model are handled by using the external point penalty function, and the optimization model is solved by combining the particle swarm optimization algorithm to obtain the quantified support capability of each partition for the main network.

2. The method according to claim 1, characterized in that, The construction of the main network frequency-voltage coupling mechanism model includes: The relationship between the active power output of the distributed power source and the system frequency deviation, as well as the relationship between the reactive power output and the voltage deviation at the grid connection point, are constructed using the droop control principle. By introducing distributed generation operation constraint equations that limit the active and reactive power outputs of distributed generation sources, a main grid frequency-voltage coupling mechanism model is obtained to characterize the dynamic coupling and mutual constraint relationship between the active and reactive power outputs of distributed generation sources under system disturbances.

3. The method according to claim 1, characterized in that, Distribution network division includes: An impedance model is constructed based on data from distribution network nodes, branches, loads, and distributed power sources. The electrical correlation strength between nodes in the distribution network is analyzed based on the impedance model, and an electrical distance matrix is ​​constructed. The power distribution network is divided into multiple zones based on the electrical distance and scale criteria between nodes, ensuring that each zone is tightly coupled and the inter-zone connections are sparse.

4. The method according to claim 1, characterized in that, The objective function of the optimization model is to maximize the weighted sum of the increment of total active power support and the increment of total reactive power support in each partition, and the priority of the increment of total active power support and the increment of total reactive power support in each partition is adjusted by weighting coefficients.

5. The method according to claim 1, characterized in that, The constraints include: nodal power flow equations, linearized AC power flow equations, and branch power transmission equations constructed based on the nodal admittance matrix; injected power balance constraints; nodal voltage offset range constraints; and connection point voltage and frequency boundary condition constraints given by the main network layer.

6. The method according to claim 1, characterized in that, The method of using exterior penalty functions to handle constraints in the optimization model and combining it with particle swarm optimization to solve the optimization model includes: Individual particles are constructed based on column vectors of active power reference value, reactive power reference value, voltage reference value, frequency reference value, active power change, and reactive power change; A particle population is constructed based on the upper and lower bound matrices of the preset particle position vector components and velocity components. The constraint is transformed into a penalty term by using an external penalty function and added to the objective function to construct the fitness function; Based on the fitness function, the individual extreme value and the global extreme value are used to guide the particle to update its position, and the process is iterated until all particles have found the optimal position.

7. The method according to claim 6, characterized in that, When the particle position vector component and / or velocity component exceed the preset upper and lower bound matrices during the search process, a correction function based on uniformly distributed random numbers is used to reset them back to the preset upper and lower bound matrices.

8. A computer device, characterized in that, It includes at least one processing unit and at least one storage unit, wherein the storage unit stores a computer program that, when executed by the processing unit, causes the processing unit to perform the method as described in any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that, It stores a computer program executable by an electronic device, which, when run on the electronic device, causes the electronic device to perform the method as described in any one of claims 1 to 7.

Images