Power distribution network power flow security domain power distribution reactive power reserve planning method and system

Abstract

The application belongs to the technical field of power system operation control and optimal scheduling, and provides a power transmission and distribution reactive power reserve planning method and system considering power distribution network power flow security domain, generates a power flow security domain sample set considering network reconstruction based on an improved radial iterative reconstruction method, uses a physical information graph neural network (PI-GNN) to simulate and solve the security domain in consideration of new energy uncertainty, defines an equivalent four-dimensional security domain cost model of the power distribution network on the basis of considering optimal carbon emission power flow, and deduces and forms a planning scheme of the power transmission and distribution system reactive power reserve considering the operation state under a typical scene. The implementability and economy of the planning scheme under multiple working conditions throughout the year are improved.

Description

Technical Field

[0001] This invention belongs to the field of power system operation control and optimization scheduling technology, and particularly relates to a method and system for planning reactive power reserves for power transmission and distribution considering the power flow security domain of the distribution network. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] With the high proportion of new energy sources integrated into the distribution network and the widespread deployment of controllable resources (inverters, energy storage, on-load tap-changing transformers, capacitors, flexible interconnection devices, adjustable loads, etc.), the operating status of the distribution network exhibits stronger randomness, coupling, and time-varying characteristics. Under the trend of integrated transmission and distribution operation, the upstream transmission network places higher demands on the voltage support and reactive power regulation capabilities of the distribution network. How to comprehensively plan the reactive power reserve configuration of the transmission and distribution system while taking into account operational economy, under the premise of meeting engineering constraints and ensuring safety and stability, has become an important research direction in the field of power system planning and dispatch.

[0004] Existing reactive power reserve planning and reactive power compensation configuration methods are mostly modeled and optimized from the perspective of the transmission network. They typically use AC power flow or linearized power flow constraints under typical scenarios to determine the parallel capacitor / reactor capacity configuration of candidate substation buses. However, in actual systems, the distribution network is not a simple fixed "PQ load." The active and reactive power exchange range and voltage regulation range at its point of common coupling (PCC) are jointly affected by the internal equipment capacity, operating constraints, power flow distribution, and topology of the distribution network. To characterize the feasible operating range of the distribution network under given constraints, related research has proposed concepts such as "power flow safety domain." However, many existing reactive power planning or feasible domain construction methods fail to unify the modeling of topology reconfiguration and reactive power / voltage coordinated regulation, resulting in insufficient characterization of the actual reactive power support capacity that the distribution network can provide. Furthermore, traditional analytical or simulation methods face the problem of exponential growth in computational load as the system scale increases. Summary of the Invention

[0005] To address at least one of the technical problems mentioned above, this invention provides a method and system for planning reactive power reserves in power transmission and distribution considering the power flow security domain of the distribution network, and proposes a hierarchical planning strategy for reactive power reserves in power transmission and distribution systems that is more in line with the continuous operation conditions of the system.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: A first aspect of the present invention provides a method for planning reactive power reserves in power transmission and distribution considering the power flow security domain of a power distribution network, comprising the following steps: A sample set of power flow security domain for distribution networks is constructed by improving the radial iterative reconstruction method; Construct a feasibility discriminant and repairer for power distribution network constraints, and perform feasibility discriminant and repair on boundary points in the generated sample set; A physical information graph neural network model is constructed, and the repaired sample set is used for learning to fit the mapping relationship between the distribution network state and the power flow safety domain. Based on the optimal power flow model of the distribution network that takes into account voltage level and carbon emissions, an equivalent four-dimensional security domain cost model of the distribution network is defined and constructed. By combining the output power flow security domain and the cost model of the equivalent four-dimensional power flow security domain of the distribution network, and by combining typical scenario sets, a hierarchical planning model for reactive power reserve of the power transmission and distribution system considering the operating state is constructed and solved to obtain the reactive power reserve configuration scheme.

[0007] Furthermore, the sample set for constructing the power flow security domain of the distribution network through the improved radial iterative reconstruction method includes: The definition includes at least three dimensions: active power, reactive power, and voltage, for the power flow security domain of the distribution network. Given a search direction, for each search direction d, the feasibility check subproblem determines whether there is a topology and control that makes the running point feasible. The bisection method is used to accurately search for boundary points within the feasible and infeasible envelopes until the accuracy meets the set threshold, thus obtaining accurate boundary sample points.

[0008] Furthermore, a feasibility assessment and repair mechanism for distribution network constraints was constructed. The input is the running point z, and the output is 0 or 1, indicating whether all engineering constraints are met. Each feasibility test allows for topology reconfiguration and coordinated adjustment of controllable resources; if Then, the pseudo-feasible boundary points are backed up and repaired along their generation direction: By performing a binary search on α, the point closest to the boundary and feasible is obtained, where, To revert and repair feasible boundary points, These are the original pseudo-feasible boundary points. This is the equivalent bisection interval.

[0009] Furthermore, when constructing the physical information graph neural network model, it includes learning the input-output relationship between the distribution network diagram, photovoltaic load status and boundary point sets in several directions. The loss function is injected with residual terms or penalty terms through physical equations and operational constraints, so that the output is structurally closer to the feasible power flow solution.

[0010] Furthermore, the construction of the equivalent four-dimensional security domain cost model for the distribution network includes: Solve for the optimal power flow of the distribution network, taking into account voltage levels and carbon emissions; Then, using the incremental cost of deviating from the optimal power flow reference operating point as the active and reactive power resource call cost of the distribution network, an equivalent four-dimensional power flow security domain cost model of the distribution network is established. Based on the cost model of the equivalent four-dimensional power flow security domain of the distribution network, the power flow security index of the distribution network is defined by the volume of the polyhedron of the equivalent three-dimensional power flow security domain of the PCC connection point.

[0011] Furthermore, the typical scenario set is constructed based on improved K-medoids clustering, including: Construct a standardized feature vector that includes information on load, renewable energy output, net load, ramp-up, forecast error, and electricity price; The samples were grouped by seasonal labels and stratified by load quantiles within each season; Define a weighted Euclidean distance and assign greater weight to features that are prone to triggering voltage and reactive power risk; Within each subset, the goal is to minimize the weighted distance, and real samples are selected as cluster centers to form typical scenarios.

[0012] Furthermore, the process of solving the hierarchical planning model for reactive power reserve of the power transmission and distribution system under operation includes: the upper layer approximates the annual operation of typical operating scenarios obtained by clustering, and obtains the current optimal capacitor and reactance configuration with the weighted sum of annual investment costs, power generation costs, annual network losses, and equivalent distribution network costs as the objective; then the lower layer generates and updates the set of typical scenarios of the upper layer through extreme scenario column constraints, and repeatedly solves the upper layer planning, iterating until the scenario set is no longer updated.

[0013] A second aspect of the present invention provides a power transmission and distribution reactive power reserve planning system that considers the power flow security domain of a distribution network, comprising: A sample set construction module is used to construct a sample set for the power flow security domain of the distribution network by means of an improved radial iterative reconstruction method. The boundary point repair module is used to construct a distribution network constraint feasibility discriminator and repairer, and to perform feasibility discriminator and repair on the boundary points in the generated sample set. The sample set learning module is used to construct a physical information graph neural network model, which uses the repaired sample set for learning and fits the mapping relationship between the distribution network state and the power flow safety domain. The cost model construction module is used to define and construct an equivalent four-dimensional security domain cost model for the distribution network based on the optimal power flow model of the distribution network that takes into account voltage levels and carbon emissions. The reactive power reserve planning module combines the output power flow security domain and the cost model of the equivalent four-dimensional power flow security domain of the distribution network with typical scenario sets to construct and solve a hierarchical planning model for reactive power reserve of the transmission and distribution system that considers the operating state, and obtains the reactive power reserve configuration scheme.

[0014] A third aspect of the present invention provides a computer-readable storage medium.

[0015] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the power transmission and distribution reactive power reserve planning method considering the power flow security domain of the distribution network as described above.

[0016] A fourth aspect of the present invention provides a computer device.

[0017] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps in the power transmission and distribution reactive power reserve planning method considering the power flow security domain of the distribution network as described above.

[0018] Compared with the prior art, the beneficial effects of the present invention are: This invention considers network reconstruction and provides raw data for solving the power flow safety domain based on an improved radial iterative reconstruction method. It establishes a distribution network constraint feasibility discriminant and repairer, trains a PI-GNN neural network to efficiently and quickly solve the power flow safety domain, and establishes a four-dimensional equivalent cost model of the distribution network by solving for the optimal power flow considering voltage and carbon emissions. It also derives a hierarchical planning strategy for reactive power reserves in the transmission and distribution system considering operational status. This method has the following advantages: 1) This model considers the voltage regulation space at the PCC of the distribution network and constructs a four-dimensional equivalent cost model of the distribution network in terms of "active power - reactive power - voltage - cost". Compared with the safety domain that only considers active power and reactive power, the four-dimensional safety domain model can use volume to characterize the adjustable space of power flow and voltage of the distribution network, and can better characterize the power and voltage support that the distribution network can provide from the perspective of the transmission network. 2) This model establishes a planning model that takes into account the operating state. By constructing a typical set of seasonal loads and new energy scenarios, it integrates the operating cost index of the power flow security domain of the distribution network to obtain the optimal reactive power reserve planning, making it more in line with the continuous operating conditions of the system. The architecture is clear and the versatility is better.

[0019] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0020] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0021] Figure 1 This is a flowchart of the power transmission and distribution reactive power reserve planning method considering the power flow security domain of the distribution network provided in the embodiments of the present invention; Figure 2This is a schematic diagram of the improved radial iterative reconstruction provided in an embodiment of the present invention; Figure 3 This is the normal voltage level four-dimensional power flow safety domain model provided in the embodiments of the present invention; Figure 4 This is a high-voltage level four-dimensional power flow safety domain model provided in the embodiments of the present invention. Detailed Implementation

[0022] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0023] It should be noted that the following detailed description is illustrative and intended to provide further explanation of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0024] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0025] Existing technologies still have shortcomings in the following aspects: it is difficult to quickly and accurately obtain the power flow safety domain of the distribution network under the conditions of distribution network topology reconfiguration and uncertainty; there is a lack of distribution network equivalent cost modeling that takes into account both economic efficiency and voltage level to support unified scheduling and planning on the transmission network side; and there is a lack of a transmission and distribution reactive power reserve planning process that can take into account both representativeness and extreme risk verification. Therefore, there is an urgent need for a reactive power reserve planning method oriented towards transmission and distribution coordination, which can construct and efficiently solve the power flow safety domain of the distribution network on the basis of strict feasibility constraints, form an equivalent cost model that can be used for upper-level planning, and improve the feasibility and economy of the planning scheme under multiple operating conditions throughout the year through a hierarchical strategy that combines typical and extreme scenarios.

[0026] The purpose of this invention is to generate a power flow safety domain sample set considering network reconstruction based on an improved radial iterative reconstruction method, to use a Physical Information Graph Neural Network (PI-GNN) to simulate and solve the safety domain considering the uncertainty of new energy sources, to define an equivalent four-dimensional safety domain cost model for the distribution network based on the optimal power flow considering carbon emissions, and to derive a planning scheme for reactive power reserve of the transmission and distribution system considering the operating state in typical scenarios.

[0027] Example 1 like Figure 1As shown, this embodiment provides a reactive power reserve planning method for transmission and distribution considering the power flow security domain of the distribution network. The method is used for reactive power reserve planning in a coordinated manner between the transmission and distribution networks. Based on the power flow security domain and the equivalent flexibility cost model, reactive power reserve planning is performed in typical scenarios, and then robustness verification is conducted in extreme scenarios. Dynamic updates to the reactive power reserve planning are achieved through iterative updates at upper and lower levels. The method includes the following steps: Step 1: Construct a sample set of the power flow security domain of the distribution network by improving the radial iterative reconstruction method; Specifically, the steps include the following: Step 101: Define a power flow security domain for the distribution network that includes at least three dimensions: active power, reactive power, and voltage. The power flow safety domain of a distribution network is defined as the set of all feasible operating states that a distribution network can achieve by relying on network reconfiguration and controllable resources in a selected variable space, given equipment capacity and operating constraints. It is represented by the power exchange range and voltage regulation range at the point of connection (PCC).

[0028] Specifically, let the set of distribution network nodes be... The set of branches is The set of candidate reconfigurable topologies is The node corresponding to PCC is 0, and the selected runtime variable space can be represented as follows: The power flow safety domain of a distribution network is defined as all feasible operating states that the distribution network can achieve in variable space, given equipment capacity and operational constraints, through network reconfiguration and controllable resources. The set in: , in, These are network state variables (node ​​voltage, branch power flow, etc.). These are controllable variables (inverter reactive power, energy storage charging and discharging, etc.). This represents uncertain quantities (such as photovoltaics and loads). , They represent the topology respectively. t The equality and inequality constraints are defined under the following conditions. When considering reconfiguration, the power flow safety region is the union of the safety regions under each topology. , This represents the power flow safety domain under topology t.

[0029] Step 102: Consider the power flow security domain model for distribution network reconfiguration; The controllable flexibility resources of the distribution network described in this invention include, but are not limited to: photovoltaic systems and their inverters, energy storage devices, on-load tap changer taps, capacitor switching, smart soft switches, and adjustable loads. Considering the impact of distribution network topology reconfiguration (changes in switch states) on the feasible region, for any topology... Establish a radial distribution network DistFlow power flow model for branches. The following constraints must be satisfied: (1) Power flow constraints, including: Node active power balance: , Node reactive power balance: , Voltage drop constraint: , Branch current constraints: , Branch flow upper and lower limits constraints: ; in, This indicates that the injection from node i to node j is active. This indicates the outflow of active power from node j to node k. This represents the active load at node j. This indicates the active power output of the photovoltaic system at node j. This indicates the energy storage output of node j. This indicates that node jSOP has active power output. This indicates the reactive power injected from node i to node j. This indicates the outflow of reactive power from node j to node k. This indicates the reactive load at node j. This indicates the reactive power output of the photovoltaic inverter at node j. This indicates that the parallel capacitor at node j compensates for reactive power output. This indicates that node jSOP has no reactive power output. Represents the voltage at node j. Represents the voltage at node i. and This represents the per-unit values ​​of the resistance and reactance of branch ij. Indicates the current in branch ij. This indicates the maximum current carrying capacity of branch ij. This represents the per-unit voltage value of node i. This represents the minimum voltage at node i. This represents the maximum voltage at node i.

[0030] (2) Component operation constraints, including: PV active power constraints: , PV reactive power constraint: , PV ramping constraint: , ESS power constraints: , ESS capacity constraints: , ESS climbing constraints: , SC Action Constraints: , SVG power constraints: , SVG ramping constraints: , SOP capacity constraints: , SOP ramping constraints: , OLTC Motion Constraints: , AL is transferable and meets the constraints: , AL can reduce load constraints: , In the formula, , Let represent the active power output of the photovoltaic system at node i at time t and time t-1. This indicates the rated active power of photovoltaic power. This represents the efficiency of photovoltaic power at time t. , This represents the reactive power output of the photovoltaic system at node i at time t and time t-1. Indicates the rated capacity of photovoltaic power. This represents the rated active power ramp rate of the photovoltaic system at node i. This represents the rated reactive power ramp rate of the photovoltaic system at node i. Indicates the rated power of energy storage. , This represents the exchange power of the energy stored at node i at time t and time t-1. This represents the state of charge of the energy stored at node i at time t. This represents the rated ramp rate of energy storage at node i. This indicates the switchable capacitor bank's position at time t. This indicates the switching action of the switchable capacitor bank at time t. Indicates the total number of switches for the capacitor bank. This indicates the maximum number of times a switchable capacitor bank can be switched on and off per day. This indicates the maximum reactive power compensation amount of the SVG. , This represents the reactive power compensation output of the SVG at node i at time t and time t-1. This represents the rated reactive ramp rate of the SVG on node i. This indicates that the SOP at node i has active power flow at time t. This indicates the reactive power flow at time t of the SOP on node i. This represents the apparent power of SOP at node i at time t-1. Indicates the SOP rated capacity. This represents the rated apparent power ramp rate of SOP at node i. This indicates the tap position of the on-load tap-changing transformer at time t. This indicates the tap changer operation of the on-load tap-changing transformer at time t. This indicates the maximum tap value of an on-load tap-changing transformer. This indicates the maximum number of daily operations for an on-load tap-changing transformer. Indicates load transfer delay. Indicates the maximum load transfer delay. Indicates the transfer of active power load. This represents the maximum transferable active power load. This indicates the transfer of reactive load. This represents the maximum transferable reactive load. This indicates a reduction in active power load. This indicates the maximum active power load that can be reduced. This indicates a reduction in reactive power load. This indicates the maximum amount of reactive load that can be reduced.

[0031] (3) PCC boundary conditions Security domain and power flow consistency constraints: , In the formula, This indicates the equivalent outflow of active power in the distribution network at the PCC boundary. This indicates the equivalent reactive power outflow from the distribution network at the PCC boundary. This indicates the active power flowing out of node k of the transmission network at the PCC boundary. This indicates that reactive power is flowing out of node k of the transmission network at the PCC boundary. This represents the square of the transmission network voltage at the PCC junction. This indicates the voltage at the distribution network node at the PCC junction. (4) Network Reconfiguration Constraints Network reconstruction based on topology Based on the criteria of radiality and connectivity, a "single commodity flow" model is adopted to meet engineering constraints such as maintaining radial network operation, ensuring critical loads are not interrupted, and limiting the number of switching operations.

[0032] Tree edge count constraint: , Capacity binding: , Node traffic balancing: , , Among them, the number of nodes binary switch state To ensure that every node is reachable from the root node 0, a virtual flow variable is introduced. (This represents the flow of "connectivity" from the root outwards), and for each undirected branch... Establish directed arcs in two directions. , Corresponding variables , .

[0033] Step 103: Solve the power flow security domain using the improved radial iterative reconstruction method to obtain the sample set of the power flow security domain of the distribution network; Simulation methods, such as the radial fast iterative reconstruction method, can quickly solve for the power flow security domain of distribution networks. However, as system size increases and accuracy requirements rise, fast iteration struggles to meet the required accuracy and can easily lead to overestimation of the security domain. Increasing the number of iterations results in an explosion in computational complexity and a very long solution time. Therefore, this embodiment provides an improved radial iterative reconstruction method for generating security domain boundary samples, such as... Figure 2 As shown, its core idea is: in a given direction superior( This method (using ray directions in space) utilizes feasibility checks and interval scaling to accurately search for boundary points, allowing topology reconstruction and coordinated adjustment of feasibility resources during each feasibility check. It does not prioritize speed, but ensures boundary point accuracy through binary search and rigorous feasibility checks, avoiding the overestimation of the safety region caused by traditional fast iterations.

[0034] Specifically, the steps include the following: Step 1031: Solve for the first-order safety region through radial iterative reconstruction; Direction parameterization: Define the unit direction for the selected variable space. d Rays on , ,in, Indicates the radial ray length; Step 1032: Repair the pseudo-feasible boundary points obtained in Step 1; Boundary point definition: Boundary points correspond to the maximum feasible... ; Step 1033: Use a neural network to learn this solution and construction process to accelerate the solution speed and use a discriminator to ensure the solution accuracy; Given search direction The feasibility test sub-problems are used to determine whether there are topological and control options that make it feasible. The feasibility verification subproblem is represented as: , in, This represents a feasibility assessment function, used to determine whether a given boundary point meets the constraints. It returns 1 if it meets the constraints and 0 if it does not. Step 1034: For each search direction ,expand Find a pair of feasible / infeasible envelopes Then, iterate using the binary search method. ,like , place until The final boundary sample points are represented as follows: .

[0035] Step 2: Construct a power distribution network constraint feasibility discriminator and repairer to perform feasibility discriminator and repair on the boundary points in the generated sample set; To repair "pseudo-feasible boundary points" caused by numerical errors, approximate power flow, or sparse sampling, this embodiment integrates feasibility assessment and security domain repair functions to construct a distribution network constraint feasibility assessment and repair tool. Its input is the run point. The output indicates whether all engineering constraints are met. Each feasibility test allows for topology reconfiguration and coordinated adjustment of controllable resources. like Then, the pseudo-feasible boundary point is backed up and repaired along its generation direction: Through the Binary search finds the point closest to the boundary that is feasible: ,in, To revert and repair feasible boundary points, These are the original pseudo-feasible boundary points. This is an equivalent bisection interval; Step 3: Construct a physical information graph neural network model, use the repaired sample set for learning, and fit the mapping relationship from the distribution network state to the power flow safety domain; Traditional simulation methods for solving power flow safety domains involve extensive constraint discrimination calculations, making it difficult to support intraday or real-time reactive power optimization. Traditional neural network models can quickly output power flow safety domains through offline learning, but they suffer from large errors and are prone to generating "pseudo-safe points." To address this, this embodiment constructs a PI-GNN model to learn the input-output relationships from the distribution network diagram, photovoltaic load state, and boundary point sets in several directions offline. By injecting residual or penalty terms into the loss function through physical equations and operational constraints, the output is structurally closer to the feasible power flow solution and more robust to out-of-distribution samples.

[0036] 1) Graph structure and input / output definition: Given a power distribution network diagram The goal of PI-GNN is to learn a mapping The power distribution network diagram is included. Based on node features and boundary features Together constitute; Inputs for system operation include photovoltaic output and load size at each node; For the power flow safety domain related output of the distribution network, the boundary point set in several directions is used. express, For network parameters, For the active load of node i in the distribution network, For reactive load at node i in the distribution network, For the active power output of distribution network node i, For the reactive power output of distribution network node i, Voltage of node i in the distribution network; The model output is basically consistent with the definition of the simulation method output. k The ray in each direction approximates the power flow safety domain, and the boundary point of the safety domain is defined at the maximum feasible radius.

[0037] 2) Physical equations and operational constraints: In the PI-GNN network, GNN can provide structured representation for the mapping, while PI compresses the solution space to the vicinity of the feasible region through the loss function. The loss function of the model can be expressed as a weighted sum of the data fitting term, the power flow equation residual term, and the operational constraint penalty term.

[0038] The loss function is expressed as: , , , , in, For data fitting terms; For the residual terms of the power flow equation, The operational constraint penalty item represents the penalty for voltage exceeding the limit at each node of the distribution network and current exceeding the limit at each branch. , , As weight, The equations are linear DistFlow power flow equations. This represents the per-unit value of the squared voltage at node i. This represents the lower limit of the per-unit squared voltage value at node i. This represents the upper limit of the squared per-unit value of the voltage at node i. This indicates the per-unit value of the current in branch ij.

[0039] (3) Convex hull edge filling: Connect the repaired safety domain boundary points through a function to close the points into a differentiable surface; This embodiment uses the ConvexHull function for convex hull edge filling. State-space interpolation is performed on the e-th edge of the convex hull, and a normal perturbation is added to refine the convex hull edge of the power flow safety domain. , in, These are the test points on the edge of the convex hull after perturbation. For the original safety and boundary points, It is a random disturbance term that follows a normal distribution with a mean of 0 and a standard deviation of 0.05.

[0040] (4) Perform feasibility assessment and repair on the generated curves: In this embodiment, a set of candidate boundary points in several directions is generated for a certain operating state of the distribution network. Then, based on the feasibility judgment and repair function of the distribution network constraints, the boundary points are determined. Accurate power flow determination is performed to ultimately obtain the accurate power flow safety domain of the distribution network.

[0041] Step 4: Based on the optimal power flow model of the distribution network that takes into account voltage level and carbon emissions, define and construct the equivalent four-dimensional security domain cost model of the distribution network; This invention considers the voltage level and carbon emissions of the distribution network. Based on the optimal power flow within the distribution network, it proposes an equivalent four-dimensional security domain model of the distribution network from the perspective of the transmission network, which includes "active power - reactive power - voltage - cost". This allows the transmission network side to characterize the marginal cost of utilizing the active and reactive power resources of the distribution network without needing to know a large number of equipment and network details within the distribution network. Figure 3 and Figure 4 Four-dimensional power flow security domain models are characterized for normal and high voltage levels, respectively. To balance solution accuracy and efficiency, a distribution network constraint feasibility discriminant and repair tool is used after solving the reactive power reserve planning problem. Identify and repair the running points.

[0042] Specifically, the steps include the following: Step 401: Solve for the optimal power flow of the distribution network, taking into account voltage levels and carbon emissions; Given the load and photovoltaic forecast values, a deterministic solution is used to find the optimal control within the distribution network. To balance voltage quality, network loss, and carbon emissions.

[0043] In this embodiment, the objective function is: , in, For distribution network losses, denoted as , Time-of-use pricing; The component cost includes source side (cured light cost), grid side (SOP loss), and energy storage side (energy storage charging and discharging degradation cost). This is a voltage offset penalty for the distribution network, used to guide the model to approach the rated voltage. The carbon emission cost of the distribution network is expressed as the sum of the carbon intensity of electricity purchased from the upstream grid, the carbon intensity of locally controllable power sources, and the carbon intensity of grid losses: , in, Indicates the power purchased. The carbon emission coefficient per unit of electricity generated by the upper-level power grid. This refers to the carbon emission coefficient of local power sources. The weighting coefficients representing network loss, voltage, component cost, and flexibility requirements can be adjusted. To characterize the carbon emission requirements of the transmission network on the active distribution network, This represents a time interval, which is 15 minutes in this embodiment. This indicates the power generation capacity of the local power source.

[0044] The model adopts the DistFlow power flow model, which can be transformed into a mixed integer second-order cone programming solution. The model constraints are consistent with the constraints in step 102.

[0045] Step 402: Using the incremental cost of deviating from the optimal power flow reference point as the active and reactive power resource allocation cost of the distribution network, establish an equivalent four-dimensional power flow security domain cost model for the distribution network. The power flow safety domain cost model at the PCC point of the distribution network can be equivalently represented as a convex polyhedron, which in The projection onto the plane is the power flow safety domain.

[0046] First, a reference operating point is obtained within the distribution network based on optimal power flow. Its corresponding internal operating cost is When the upper-layer scheduling requires the PCC power to deviate from this reference point, the equivalent incremental cost is defined as follows: , in, This is a function for calculating the cost of a distribution network. By calling the optimal power flow, it can calculate the lowest cost under the given output conditions after the active and reactive power resources of the distribution network are determined by the upstream transmission network.

[0047] Step 403: Based on the cost model of the equivalent four-dimensional power flow security domain of the distribution network, the power flow security index of the distribution network is defined by the volume of the polyhedron of the equivalent three-dimensional power flow security domain of the PCC connection point.

[0048] This invention proposes a "distribution network power flow security" index based on the cost model of the equivalent four-dimensional power flow security domain of the distribution network, using the PCC connection point as the equivalent three-dimensional power flow security domain polyhedron. The volume representation can be used to assess the safety and adjustability of power flow and voltage in distribution networks.

[0049] Step 5: Combining the output power flow security domain and the cost model of the equivalent four-dimensional power flow security domain of the distribution network, and combining the typical scenario set, construct and solve the hierarchical planning model of reactive power reserve of the power transmission and distribution system considering the operating state, and obtain the reactive power reserve configuration scheme.

[0050] Based on the power flow security domain and its cost model of the distribution network, this invention proposes a hierarchical planning method for reactive power reserve of the power transmission and distribution system that considers the power flow security domain of the distribution network, in order to meet the needs of integrated operation and planning of power transmission and distribution. This method assumes that the regional tie line is equivalent to the external value when the active power is fully generated. Under typical scenario sets, the reactive power reserve configuration scheme that meets the engineering constraints and operational economy is obtained by taking network loss and voltage qualification rate as comprehensive indicators.

[0051] Specifically, the steps include the following: Step 501: Construct a typical scenario set based on improved K-medoids clustering; Traditional K-means clustering centers the mean point, which is often not based on real-time data and cannot support subsequent power flow verification. This embodiment proposes a K-medoids clustering method with "seasonal stratification, distance weighting, and extreme value supplementation," and makes engineering improvements for reactive power and voltage risk characteristics to ensure that representative scenarios are reproducible, replayable, and verifiable. The clustering steps specifically include: Step 5011: Construct a standardized feature vector containing information on load, renewable energy output, net load, ramp-up, forecast error, and electricity price; Let the original candidate sample set be Ω, and each sample Construct a feature vector for each day's data: , in, For the first day of the sample set i Dimensional characteristics; the sample must contain at least the total regional load. Aggregated photovoltaic power output Aggregated wind power output Net load Net load at time t Net load at time t-1 Net load ramp-up Photovoltaic forecasting error index Wind power forecasting error index and electricity price wait.

[0052] Standardize each feature dimension to obtain a standardized feature vector. ,in, The sample mean. This represents the sample variance. Step 5012: Group the samples by seasonal label and stratify them by load quantile within each season; Group the samples according to seasonal labels: ; Stratification by load quantile within each season: ; in, This represents a small load sample set. Indicates a large sample set. For the spring sample set, For the summer sample set, For the autumn sample set, For the winter sample set, For one of the above sample sets, For another sample set, Given a sample set for any season, For the total regional load, As a sample, It is the median quantile.

[0053] Step 5013: Define the weighted Euclidean distance and assign greater weights to features that are prone to triggering voltage and reactive power risk; Define feature weight vector ,satisfy and .

[0054] For any two samples and Define weighted Euclidean distance ,in, Indicates sample The standardized feature vector, Indicates sample The standardized feature vector.

[0055] In engineering practice, greater weight can be assigned to characteristics that are more likely to trigger voltage and reactive power risks, such as net load ramp-up, prediction error width, and new energy penetration rate.

[0056] Step 5014: Within each subset, select real samples as cluster centers with the goal of minimizing weighted distance to form typical scenarios; For a certain subset (e.g., low load in summer), given the number of clusters Select within this subset A set of real samples The optimization problem is solved with the objective of minimizing the weighted distance to obtain the center median sample points of each subset. In this invention, the median median is selected as follows: =1, the problem can be simplified to: Ultimately, eight typical scenarios were obtained, including low load in spring and high load in spring.

[0057] Step 5015: Supplement the extreme value scenario set, including four types of scenarios: maximum load or minimum renewable energy output, minimum load or maximum renewable energy back-transmission, maximum prediction error width, and maximum net load ramp-up.

[0058] Clustering methods are inherently biased towards representing averages, and even weighted methods may miss safety region boundaries. Therefore, this invention focuses on representative scenario sets. Additional extreme value sample set This is used to verify reactive power reserve planning strategies, including four scenarios: maximum load or minimum renewable energy output, minimum load or maximum renewable energy backfeed, maximum prediction error width, and maximum net load ramp-up, in the following format:

[0059] in, For the extreme value sample set, As a sample, It is the 95th percentile. For the sample The total regional load, The average total load of the sample set, For the sample Total regional new energy output The average total renewable energy output of the sample set. For the sample The regional average prediction error of new energy sources The average prediction error for new energy sources in the sample set.

[0060] Step 502: Construct and solve a hierarchical planning model for reactive power reserve of the power transmission and distribution system that considers the operational state, and obtain the reactive power reserve configuration scheme; This invention proposes a hierarchical planning strategy for reactive power reserves of power transmission and distribution systems that considers the operational state, based on the power flow security domain of the distribution network. Specifically, the upper layer approximates the annual operation of typical operating scenarios obtained by clustering, and obtains the current optimal capacitor and reactance configuration with the weighted sum of annual investment costs, generation costs, annual network losses, and equivalent distribution network costs as the objective. Then, the lower layer generates and updates the set of typical scenarios of the upper layer through extreme scenario column constraints, and repeatedly solves the upper layer planning, iterating until the scenario set is no longer updated (or the termination threshold is met).

[0061] Specifically, the steps include the following: Step 5021: Upper layer: Consider the transmission and distribution coordinated reactive power reserve planning in the power flow security domain of the distribution network; In distribution networks, active and reactive power are tightly coupled, and their power flow security domain must fully consider the interaction range of active and reactive power at the PCC (Power Distribution Center). The equivalent cost of a distribution network is a function of active power, reactive power, and voltage. In contrast, the active and reactive power of a transmission network can be decoupled. Therefore, this embodiment, while ensuring the voltage security and operational feasibility of both the transmission and distribution networks, considers the joint optimization of the reactive power compensation equipment configuration strategy for the transmission network on a typical scenario set.

[0062] Among them, the decision variables are: in each candidate substation bus... When investing in parallel capacitors and parallel reactances, using the "susceptance - voltage squared" formula is more in line with physical characteristics.

[0063] in This indicates the upper limit of the susceptance of the installable capacitor. This indicates the upper limit of the installable reactance susceptance. This represents the set of busbars in a substation; Runtime variables: for each typical scenario The runtime variables are as follows: Compensation equipment susceptance

[0064] Voltage relaxation beyond limit

[0065] Active and reactive power output of the generator

[0066] Operating point at PCC of distribution network

[0067] in, This represents the susceptance of the capacitor installed on node i in scenario s. This represents the reactance and susceptance installed on node i in scenario s. This indicates a relaxation variable above the voltage upper limit. This indicates a relaxation variable as the voltage approaches its lower limit. This indicates the active power output of the generator at node j in scenario s. This represents the maximum active power output of the generator at node j. This represents the minimum reactive power output of the generator at node j. This represents the reactive power output of the generator at node j in scenario s. This represents the maximum reactive power output of the generator at node j. This represents the active power exchange at PCC connection point k in scenario s. This represents the reactive power exchange at PCC connection point k in scenario s. This represents the voltage at connection point k of PCC in scenario s. This represents the power flow safety domain of the distribution network.

[0068] Objective function: The objective function is represented by a weighted approximation of the annual operation based on typical scenarios. It is expressed as the sum of the annual value of investment cost and the annual operating cost. The voltage over-limit penalty is represented in a continuously optimizable form to avoid directly introducing 0-1 variables, which would lead to difficulty in solving the problem and facilitate the lower layer to identify the degree of default in extreme scenarios.

[0069] , in, This represents the total cost of the power transmission and distribution network. This represents the annual value of the investment cost. This represents the annual operating cost. This indicates that the capacitor susceptance of the compensation device at node i is in operation. This indicates that the compensation equipment at node i has activated its reactance and susceptance. and This is the annual cost factor per unit of charge capacity. The weights represent the values ​​for a typical scenario s. The generator power generation cost under typical scenario s. The active power loss of the transmission network under typical scenario s. The equivalent cost of the distribution network under typical scenario s, This is a penalty item for voltage exceeding the limit. , , , These are the weighting coefficients for each type of operating cost.

[0070] Constraints: For each typical scenario The following AC power flow and security constraints of the transmission network should be met, and the relevant constraints of the distribution network should be represented in an integrated manner in the power flow security domain.

[0071] Active power balance constraints: , Reactive power balance constraints: .

[0072] Voltage operating constraints: .

[0073] Branch flow constraints: , in, Let i be the active power generation of node i in scenario s. Let i be the active power load of node i in scenario s. , The node voltage amplitude, Let Y be the real part of the nodal admittance matrix. Let Y be the imaginary part of the nodal admittance matrix. Let be the voltage phase angle difference between node i and node j; Let i be the reactive power generation of node i in scenario s. Let i be the reactive load of node i in scenario s. The reactive power injected into the parallel compensation device at node i; , These are the upper and lower limits of the node voltage. , This is a voltage-limited relaxation variable; For the apparent power flow of branch l in scenario s, This represents the upper limit of the capacity of branch l.

[0074] Step 5022: Lower layer: Robustness check of extreme scenario column constraint generation; The lower layer fixes the investment allocation obtained by the upper layer. For extreme scenario sets Conduct a communication trend check one by one and identify the following two types of problems: 1) Infeasible: There is no solution under the condition that relaxation is not allowed or relaxation is the primary condition; 2) Severe over-limit: The voltage over-limit penalty index exceeds the threshold.

[0075] For each extreme scenario Solve the verification problem with the objective of minimizing the degree of limit violation. The constraints should satisfy the upper-level isomorphic AC power flow, branch power flow, generator output, and distribution network power flow security domain constraints. Calculate the default degree in extreme scenarios and define the problem scenario set: , like To enhance robustness, the problem scenario is added to the set of typical upper-level scenarios by adopting the concept of column constraint generation. The iteration termination condition is: Or it may reach the maximum number of iterations.

[0076] The invention will be further illustrated below with examples. This embodiment takes "power flow safety domain sample generation—PI-GNN fast solution of safety domain—four-dimensional equivalent cost modeling—construction of typical / extreme scenario sets—two-layer reactive power reserve planning and column constraint iteration" as the main line to verify the effectiveness of the method in balancing operational economy and voltage qualification rate.

[0077] 1. Calculation System and Parameter Settings The IEEE 39-node transmission network was selected as the reactive power reserve planning system, with a voltage constraint of ±0.05 pu and line thermal stability constraints set according to the original data. The generator reactive power output boundary was uniformly set to ±0.5 pu. The IEEE 33-node distribution network was selected as the power flow security domain simulation system, with a voltage constraint of 0.94~1.06 pu. The planning layer approximates the annual operation on a typical day, with the sum of the weights of typical scenarios being 1. The operation layer is discretized with 96 points and a time granularity of 15 minutes.

[0078] The distribution network is configured according to typical active distribution network resource allocation. Photovoltaic power is connected to nodes 18 / 25 / 30, with installed capacities of 1.5 / 1.0 / 0.8MW respectively, and inverter capacity is 1.1 times the installed capacity. Energy storage is connected to nodes 13 / 24, with a rated power of 1.0MW and a rated capacity of 2.0MWh for each. The on-load tap-changing transformers are set to -8 to 8 taps, with a maximum of 10 tap changes per day. Capacitors are connected to nodes 14 / 30, with four groups connected to each node, each group having a capacity of 0.15Mvar, and a maximum of 8 switching operations per day. Five typical tie switches are used as candidate reconfiguration branches (8-21, 9-15, 12-22, 18-33, 25-29), and 32 branches in the original radial network are used as normally closed switches and critical loads. Supply failure is not permitted. Power flow safety domain boundary samples are obtained through an improved radial iterative reconstruction method, with direction number K=360 (180 directions in the PQ plane * 2 layers in the voltage dimension), and a binary search accuracy of [missing information]. pu, stop threshold The initial envelope expansion In PI-GNN, the loss weight coefficients are set to... , , A sample library is constructed using historical curves over 365 days. Training and verification samples are generated for load and photovoltaic data on typical days using the following statistical perturbations:

[0079]

[0080] The 365-day data were grouped into eight subsets based on "four seasons + load stratification". For each subset, a weighted K-medoids clustering algorithm was used to select one representative day. Several extreme scenario days were added in addition to the representative scenarios for robustness verification. The eight typical days and their weights are shown in Table 1 below: Table 1. Eight Typical Days and Their Weights

[0081] In the joint optimization of upper-level reactive power compensation equipment investment and typical scenario operation, eight candidate buses were selected. Used to configure parallel capacitors / reactors, with compensation capacity expressed in susceptance, and an upper limit of [value missing]. Annual investment cost coefficient Ten thousand yuan, The annual penalty coefficient for voltage over-limit relaxation variables is set at 200,000 yuan. In the robustness verification and column constraint generation for lower-level extreme scenarios, the threshold for the voltage over-limit penalty index is set at... Ten thousand yuan / day.

[0082] 2. Strategy Comparison and Result Analysis To verify the advantages of this invention, three strategies were compared: Strategy A (this invention): Use PI-GNN to quickly solve the power flow security domain (including reconfiguration) of the distribution network, form a four-dimensional equivalent cost model, perform reactive power reserve planning for typical scenarios, and generate two-level column constraints for verification using extreme scenarios; Strategy B (without considering the distribution network safety domain): The distribution network is equivalent to a fixed PQ load and a fixed power factor (0.98). No feasible domain or equivalent cost is provided. Planning is only done on the transmission network side. Strategy C (considering the safety domain but not refactoring, no extreme column constraints): fix the topology in the construction of the power flow safety domain, do not perform column constraint robustness verification, and plan all at once using only 8 typical scenarios.

[0083] Table 2 presents the optimal investment allocation for the three strategies (expressed in pU volt-ampere; the converted Mvar is given in parentheses, approximated based on a 100 MVA benchmark): Table 2 Optimal Investment Allocation for Three Strategies

[0084] Table 3 shows a comparison of the economics and voltage performance of the three strategies: Table 3 Comparison of economic efficiency and voltage performance of the three strategies

[0085] It can be seen that: Although Strategy B has a slightly lower investment, it fails to characterize the "feasible operating range and voltage support capability" of the distribution network, resulting in multiple failures and more serious overruns in extreme scenarios; Strategy C considers the safety domain but ignores reconfiguration and extreme column constraints, making the scheme still not robust under a few high-risk conditions; Strategy A of this invention, by considering the safety domain of reconfiguration, four-dimensional equivalent cost and robust verification of column constraints, ensures that all extreme conditions pass while achieving the lowest annualized operating cost and the highest transmission and distribution voltage qualification rate.

[0086] Example 2 This embodiment provides a power transmission and distribution reactive power reserve planning system that considers the power flow security domain of the distribution network, including: A sample set construction module is used to construct a sample set for the power flow security domain of the distribution network by means of an improved radial iterative reconstruction method. The boundary point repair module is used to construct a distribution network constraint feasibility discriminator and repairer, and to perform feasibility discriminator and repair on the boundary points in the generated sample set. The sample set learning module is used to construct a physical information graph neural network model, which uses the repaired sample set for learning and fits the mapping relationship between the distribution network state and the power flow safety domain. The cost model construction module is used to define and construct an equivalent four-dimensional security domain cost model for the distribution network based on the optimal power flow model of the distribution network that takes into account voltage levels and carbon emissions. The reactive power reserve planning module combines the output power flow security domain and the cost model of the equivalent four-dimensional power flow security domain of the distribution network with typical scenario sets to construct and solve a hierarchical planning model for reactive power reserve of the transmission and distribution system that considers the operating state, and obtains the reactive power reserve configuration scheme.

[0087] It should be noted that the specific implementation of the power transmission and distribution reactive power reserve planning system considering the power flow security domain of the distribution network in this embodiment of the invention is similar to the specific implementation of the power transmission and distribution reactive power reserve planning method considering the power flow security domain of the distribution network in this embodiment of the invention. For details, please refer to the description in the method section. In order to reduce redundancy, it will not be repeated here.

[0088] Example 3 This embodiment provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps in the power transmission and distribution reactive power reserve planning method considering the power flow security domain of the distribution network as described above.

[0089] Example 4 This embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps in the power transmission and distribution reactive power reserve planning method considering the power flow security domain of the distribution network as described above.

[0090] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.

[0091] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0092] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0093] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0094] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0095] The above description is merely a preferred embodiment of the present invention and is not intended to limit the 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 method for planning reactive power reserves in power transmission and distribution considering the power flow security domain of a distribution network, characterized in that, include: A sample set of power flow security domain for distribution networks is constructed by improving the radial iterative reconstruction method; Construct a feasibility discriminant and repairer for power distribution network constraints, and perform feasibility discriminant and repair on boundary points in the generated sample set; A physical information graph neural network model is constructed, and the repaired sample set is used for learning to fit the mapping relationship between the distribution network state and the power flow safety domain. Based on the optimal power flow model of the distribution network that takes into account voltage level and carbon emissions, an equivalent four-dimensional security domain cost model of the distribution network is defined and constructed. By combining the output power flow security domain and the cost model of the equivalent four-dimensional power flow security domain of the distribution network, and by combining typical scenario sets, a hierarchical planning model for reactive power reserve of the power transmission and distribution system considering the operating state is constructed and solved to obtain the reactive power reserve configuration scheme.

2. The power transmission and distribution reactive power reserve planning method considering the power flow security domain of the distribution network as described in claim 1, characterized in that, The sample set for constructing the power flow security domain of the distribution network using the improved radial iterative reconstruction method includes: The definition includes at least three dimensions: active power, reactive power, and voltage, for the power flow security domain of the distribution network. Given a search direction, for each search direction d, the feasibility check subproblem determines whether there is a topology and control that makes the running point feasible. The bisection method is used to accurately search for boundary points within the feasible and infeasible envelopes until the accuracy meets the set threshold, thus obtaining accurate boundary sample points.

3. The power transmission and distribution reactive power reserve planning method considering the power flow security domain of the distribution network as described in claim 1, characterized in that, Constructed Distribution Network Constraint Feasibility Determination and Repair Tool The input is the running point z, and the output is 0 or 1, indicating whether all engineering constraints are met. Each feasibility test allows for topology reconfiguration and coordinated adjustment of controllable resources; if Then, the pseudo-feasible boundary points are backed up and repaired along their generation direction: By performing a binary search on α, the point closest to the boundary and feasible is obtained, where, To revert and repair feasible boundary points, These are the original pseudo-feasible boundary points. This is the equivalent bisection interval.

4. The power transmission and distribution reactive power reserve planning method considering the power flow security domain of the distribution network as described in claim 1, characterized in that, When constructing the physical information graph neural network model, it includes learning the input-output relationship from the distribution network diagram and photovoltaic load status to the set of boundary points in several directions. The loss function is injected with residual terms or penalty terms through physical equations and operational constraints, so that the output is structurally closer to the feasible power flow solution.

5. The power transmission and distribution reactive power reserve planning method considering the power flow security domain of the distribution network as described in claim 1, characterized in that, The construction of the equivalent four-dimensional security domain cost model for the distribution network includes: Solve for the optimal power flow of the distribution network, taking into account voltage levels and carbon emissions; Then, using the incremental cost of deviating from the optimal power flow reference operating point as the active and reactive power resource call cost of the distribution network, an equivalent four-dimensional power flow security domain cost model of the distribution network is established. Based on the cost model of the equivalent four-dimensional power flow security domain of the distribution network, the power flow security index of the distribution network is defined by the volume of the polyhedron of the equivalent three-dimensional power flow security domain of the PCC connection point.

6. The method for planning reactive power reserves in power transmission and distribution considering the power flow security domain of a distribution network as described in claim 1, characterized in that, The typical scenario set is constructed based on improved K-medoids clustering and includes: Construct a standardized feature vector that includes information on load, renewable energy output, net load, ramp-up, forecast error, and electricity price; The samples were grouped by seasonal labels and stratified by load quantiles within each season; Define a weighted Euclidean distance and assign greater weight to features that are prone to triggering voltage and reactive power risk; Within each subset, the goal is to minimize the weighted distance, and real samples are selected as cluster centers to form typical scenarios.

7. The power transmission and distribution reactive power reserve planning method considering the power flow security domain of the distribution network as described in claim 1, characterized in that, The process of solving the hierarchical planning model for reactive power reserve of the power transmission and distribution system under operation includes: the upper layer approximates the annual operation of typical operating scenarios obtained by clustering, and obtains the current optimal capacitor and reactance configuration by taking the weighted sum of annual investment cost, power generation cost, annual network loss, and distribution network equivalent cost as the objective; then the lower layer generates and updates the set of typical scenarios of the upper layer through extreme scenario column constraints, and repeatedly solves the upper layer planning, iterating until the scenario set is no longer updated.

8. A power transmission and distribution reactive power reserve planning system considering the power flow security domain of a distribution network, characterized in that, include: A sample set construction module is used to construct a sample set for the power flow security domain of the distribution network by means of an improved radial iterative reconstruction method. The boundary point repair module is used to construct a distribution network constraint feasibility discriminator and repairer, and to perform feasibility discriminator and repair on the boundary points in the generated sample set. The sample set learning module is used to construct a physical information graph neural network model, which uses the repaired sample set for learning and fits the mapping relationship between the distribution network state and the power flow safety domain. The cost model construction module is used to define and construct an equivalent four-dimensional security domain cost model for the distribution network based on the optimal power flow model of the distribution network that takes into account voltage levels and carbon emissions. The reactive power reserve planning module combines the output power flow security domain and the cost model of the equivalent four-dimensional power flow security domain of the distribution network with typical scenario sets to construct and solve a hierarchical planning model for reactive power reserve of the transmission and distribution system that considers the operating state, and obtains the reactive power reserve configuration scheme.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps in the power transmission and distribution reactive power reserve planning method considering the power flow security domain of the distribution network as described in any one of claims 1-7.

10. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps in the power transmission and distribution reactive power reserve planning method considering the power flow security domain of the distribution network as described in any one of claims 1-7.

Images