A system and method for generating power system similar topology flow samples

Abstract

The present invention discloses a system and method for generating similar topology flow samples of a power system, belonging to the field of power system operation analysis and calculation technology. A basic flow sample includes pre-set nodes and a full-network CIM connection model; wherein a part of the nodes are designated as load nodes and another part of the nodes are designated as generator nodes; and the load adjustment of each load node is set to the maximum value; a power grid topology generation module generates a power grid topology structure and generates multiple similar power grid topologies by changing the in-station topology state or the state of the in-station line; the power grid topology generation module includes an in-station topology processing unit and a plant-station line topology unit; the plant-station line topology unit changes the plant-station line topology structure by removing the plant-station line; a power grid state quantity generation module is used to generate multiple load scenarios based on the maximum load value; and a power grid sample storage and management module stores and manages the characteristic values ​​and identifiers in the flow calculation sample.

Description

Technical Field

[0001] The present invention belongs to the technical field of power system operation analysis and calculation, and particularly relates to a system and method for generating power system similar topology power flow samples. Background Art

[0002] Samples are the fundamental data objects in data-driven methods, which are applied to studying power system operation. The most fundamental issue is generating large-scale power flow sample datasets. Data-driven methods place a high demand on power flow samples, and these datasets must fully reflect the power system's operating space. Because actual operating data suffers from data bias and cannot fully reflect the power system's operating space, the generated samples need to be supplemented. This problem is known as the power flow sample generation problem. To address this issue, power flow sample generation methods must be able to quickly generate a large number of samples, and the power flow samples must have a good data distribution.

[0003] Because the topology of the power system changes in real time under the control of power system dispatchers, the grid topology does not remain constant. Power system status information is also derived from the power flow state of the grid through analysis of switchgear within a complete CIM model. Considering the potential feasible topologies of the grid, sample generation should also consider the generation of similar topologies.

[0004] There are two main methods for generating power system flow samples under the determined grid topology: the planned flow method and the inverse function method.

[0005] The planned power flow method uses power flow calculation technology. First, given the power system's network topology, connection parameters, and some node parameters, the problem is transformed into a nonlinear algebraic equation based on simple circuit constraints in physics. Solving the equations yields all the properties of all nodes that meet the constraints. The main principles are as follows:

[0006] The nodes in the power system are divided into PV nodes, PQ nodes and balancing machine nodes. The attributes of a single node mainly include the active and reactive power injected into the node, the amplitude and phase angle of the voltage, four attributes. The PV node is a node with given injected active power and voltage amplitude, usually a generator; the PQ node is a node with given injected active and reactive power, usually a load; the balancing machine node is a node with given voltage and phase angle. The flow sample is defined as an operating point of the power system under a certain steady state, that is, the combination of the injected power of all nodes and the node voltage when the power system is in a steady state. Its expression is: {{P i ,Q i ,V i ,θ i}|i=0,1,…,N}, where P i is the injected active power of node i, Q iis the injected reactive power of node i, V i is the voltage amplitude at node i, θ i is the voltage phase angle of node i;

[0007] For each node i in the power system, a set of real equations constrains its four variables {P i ,Q i ,V i ,θ i}, the real number equation is as follows:

[0008]

[0009] Among them, δ ij is the voltage phase angle of node i ahead of node j, G ij and B ij are the real and imaginary parts of the mutual admittance of node i and node j, respectively. i is the node number, node 0 is the default balanced node, n is the total number of nodes, V j is the voltage amplitude at node j. Although direct flow calculation can clearly define a certain state of the power grid, it often fails to converge when dealing with large systems.

[0010] When using an inverse function to generate power flow samples, all node voltages can be directly set to generate node injection power to determine the grid state. This has the advantages of fast generation speed and guaranteed convergence. However, the generated samples are highly random and lack a strong sense of purpose, which has a greater impact on the reactive output of the generated samples and a smaller impact on the active output. It also affects the load scenario and cannot guarantee coverage of the complete sample space. Summary of the Invention

[0011] In response to the problems in the related art, the present invention proposes a system and method for generating power system flow samples with similar topologies to overcome the above-mentioned technical problems in the existing related art. The present invention mainly considers generating similar topologies of flow samples based on plant and station information or network-wide CIM connection model information. Secondly, the method of exchanging partial output is adopted for unit output. While ensuring active power balance, the unit output combination is generated according to the specified difference to improve the coverage of the sample space. For non-converged flow scenarios, samples with similar active power distribution characteristics are searched as the initial value of the flow calculation to improve convergence. For the samples, the node voltage phase angle is defined as the storage unique identifier, the voltage phase angle of the generator and load nodes is the characteristic identifier, and the load value of the load node is the management identifier. The relative output load ratio is calculated as the active power distribution characteristic.

[0012] The technical solution of the present invention is implemented as follows: a system for generating power system similar topology flow samples, comprising:

[0013] A basic power flow sample includes pre-defined nodes and a network-wide CIM connection model; wherein a portion of the nodes are designated as load nodes and another portion of the nodes are designated as generator nodes; and the load of each of the load nodes is set to a maximum value;

[0014] A power grid topology generation module includes an intra-station topology processing unit and a plant-station line topology unit; the intra-station topology processing unit is used to generate a power grid topology structure and obtain multiple similar power grid topologies reflected by the basic power flow samples by changing the intra-station topology state or the state of the intra-station lines;

[0015] The power grid topology structure includes an off-station topology structure and an on-station topology structure, wherein the off-station topology structure includes a off-station related nodes and multiple off-station lines, and the on-station topology structure includes b on-station independent nodes, d on-station lines, and multiple on-station connection nodes, wherein a, b, and d are all positive integers; each of the off-station related nodes is connected to an on-station connection node via the off-station line, and the on-station connection node is connected to an on-station independent node via the on-station line;

[0016] The in-station topology status includes the status of each independent node in the station, the status of each line in the station and the status of each communication node in the station;

[0017] The plant-station line topology unit modifies the plant-station line topology structure by removing the plant-station line to adapt to the situation where the power system is shut down for line maintenance;

[0018] The grid state quantity generation module is used to generate multiple load scenarios based on the maximum load value. It evenly distributes the output of the generators to generate basic samples for a single topology and a single load scenario. It adjusts the output of the units by exchanging outputs. It uses the Newton-Ray method to calculate the power flow and obtain power flow calculation samples.

[0019] The power grid sample storage and management module stores and manages the characteristic values ​​and identifiers in the power flow calculation samples.

[0020] In the present invention, the data such as the independent nodes within the station, the lines within the station, the communication nodes within the station and the related nodes outside the station are mainly used to describe the power grid topology within the plant station; wherein, the lines within the station include the connection equipment within the station, and the connection equipment within the station includes the transformer.

[0021] Furthermore, the independent node within the station is a node that is not connected to related nodes outside the station, and includes the generator node and the load node; wherein the number of the communication nodes within the station is no more than d.

[0022] Furthermore, in the power grid topology, the state of the switching device is used as a node of the topology, wherein the state of the switching device includes a connected state and a disconnected state;

[0023] Furthermore, the present invention uses the states of all switchgear as nodes in the topology, which undoubtedly includes the intermediate process of switching between different power flow states, along with some intermediate states of the entire network topology connection. Pre-calculating and then storing the data in the station is acceptable because the changes within a single power plant are relatively slow.

[0024] In the present invention, if the CIM model information in the station is known, the status of two types of switch components, circuit breakers and disconnectors, are mainly traversed, and the normal operating status topology of the plant station is statistically analyzed and pre-stored. Only when new equipment is added to the plant station, the possible status is re-traversed and stored.

[0025] Furthermore, the topology of the power grid is modified by splitting or merging the intra-station contact nodes; merging two intra-station contact nodes of the same voltage level into one intra-station contact node, and there is one topological connection situation; for an intra-station topological contact node that is connected to x intra-station lines and a out-station nodes, it is split into two intra-station contact nodes, and there are two connection situations. species, where x is a positive integer and x≤d.

[0026] Furthermore, in the plant-station line topology unit, for a power network with m plant-station lines, if one line is removed, there are m situations, and if n lines are removed, there are C m n The topology of the entire network can be determined by removing the selected n lines and then performing a topological analysis in conjunction with the in-station topology to remove in-station contact nodes and independent nodes that have no electrical connection with nodes outside the station due to line failure.

[0027] Furthermore, in the grid state quantity generation module, for a single load scenario, different unit output combinations are generated by exchanging values, and then the output is modified according to a predetermined difference while ensuring active power balance.

[0028] Furthermore, in the grid state quantity generation module, for a network with N nodes under a single topology, a certain generator node is designated as a balancing node. Given the voltage amplitudes of the N nodes and the voltage phase angle differences between the N-1 unbalanced nodes and the balancing node, a power flow calculation sample can be determined, i.e., 2N-1 characteristic values ​​are used as the unique identifier of a single power flow calculation sample.

[0029] Furthermore, for each power flow calculation sample, the voltage phase angle of the node is obtained as a unique identifier, and the management identifier and the characteristic identifier are determined to calculate the active power distribution characteristics.

[0030] A method for generating power system similar topology flow samples is applied to the above-mentioned power system similar topology flow sample generation system. The generation method is as follows:

[0031] Step S1: pre-set a basic power flow sample; set the load of the given load node to the maximum value, and set the CIM connection model of the entire network; wherein the load node and the generator node are independent nodes within the station and are not directly connected to related nodes outside the station;

[0032] Step S2: First, different power grid topologies are generated, the intra-station contact nodes in the power grid topologies are modified, and then the line states in the plant-station line topologies are modified to obtain multiple similar power grid topologies reflected by basic power flow samples;

[0033] Step S3: First, based on the maximum load value, multiple load scenarios are generated; the output of the generators is evenly distributed to generate basic samples under a single topology and single load scenario; the output of the units is adjusted by using the output exchange method; the power flow calculation is performed using the Newton method to obtain power flow calculation samples;

[0034] Step S4: determining whether the power flow of the power flow calculation sample has converged;

[0035] Step S5: If the power flow calculation sample converges, the characteristic values ​​and identifiers in the power flow calculation sample are extracted to the power grid sample storage and management module for storage and management.

[0036] Furthermore, in step S4, if the flow of the flow calculation sample does not converge, the voltage and amplitude of the flow sample with similar active power distribution characteristics in the generated flow sample library are searched as the initial value of the Newton method flow calculation, and then the Newton method flow calculation is re-applied to the initial value to obtain the flow calculation sample, and then it is re-judged whether the flow of the newly generated flow calculation sample converges. If it is determined to converge, step S5 is entered to record the number of times the flow calculation is recalculated after the search. If the number of times exceeds the specified number and still fails to converge, the search is stopped, and the failed flow calculation sample is saved for manual analysis.

[0037] Beneficial effects of the present invention:

[0038] (1) First, the present invention generates similar topologies of flow samples based on plant information or grid CIM model information, effectively utilizing the existing grid structure and information, making the generated flow samples closer to the actual grid conditions, and improving the practicality and accuracy of the samples. In terms of unit output, the present invention adopts a method of exchanging partial outputs, and generates unit output combinations according to specified differences while ensuring active power balance. This method can generate a variety of unit output combinations, thereby significantly improving the coverage of the sample space and providing a richer and more comprehensive data foundation for subsequent flow analysis, grid optimization, etc.

[0039] (2) Secondly, for non-converged power flow scenarios, the present invention effectively utilizes the information of existing samples by searching for samples with similar active power distribution characteristics as the initial values ​​for power flow calculation, reduces the difficulty of power flow calculation convergence, and improves calculation efficiency and stability. Moreover, the present invention defines the node voltage phase angle as a storage unique identifier, the voltage phase angle of the generator node and the load node as a feature identifier, and the load value of the load node as a management identifier. This identification method makes the management of sample data clearer and more orderly, and facilitates subsequent feature extraction and analysis.

[0040] (3) Finally, by calculating the relative output load ratio as the active power distribution feature, the present invention provides a concise and intuitive way to describe the distribution of active power in the power grid, which helps to quickly identify the operating status and potential problems of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a schematic structural diagram of the dual-mother main wiring of the present invention;

[0042] Figure 2 A schematic diagram of the structure of the intra-station contact node splitting and merging of the present invention;

[0043] Figure 3 The present invention is a flowchart of a method for generating power system similar topology flow samples. DETAILED DESCRIPTION

[0044] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0045] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like to indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0046] This embodiment provides a system for generating power system similar topology power flow samples, including:

[0047] A basic power flow sample includes pre-defined nodes and a network-wide CIM connection model; wherein a portion of the nodes are designated as load nodes and another portion of the nodes are designated as generator nodes; and the load of each of the load nodes is set to a maximum value;

[0048] The power grid topology generation module includes an intra-station topology processing unit and a plant-station line topology unit;

[0049] The in-station topology processing unit is configured to generate a power grid topology structure and obtain a plurality of similar power grid topologies reflected by the basic power flow samples by changing the in-station topology state or the state of the in-station lines;

[0050] The power grid topology structure includes an off-station topology structure and an on-station topology structure, wherein the off-station topology structure includes a off-station related nodes and multiple off-station lines, and the on-station topology structure includes b on-station independent nodes, d on-station lines, and multiple on-station connection nodes, wherein a, b, and d are all positive integers; each of the off-station related nodes is connected to an on-station connection node via the off-station line, and the on-station connection node is connected to an on-station independent node via the on-station line;

[0051] The in-station topology status includes the status of each independent node in the station, the status of each line in the station and the status of each communication node in the station;

[0052] The plant-station line topology unit modifies the plant-station line topology structure by removing the plant-station line to adapt to the situation where the power system is shut down for line maintenance;

[0053] The grid state quantity generation module is used to generate multiple load scenarios based on the maximum load value. It evenly distributes the output of the generators to generate basic samples for a single topology and a single load scenario. It adjusts the output of the units by exchanging outputs. It uses the Newton-Ray method to calculate the power flow and obtain power flow calculation samples.

[0054] The power grid sample storage and management module stores and manages the characteristic values ​​and identifiers in the power flow calculation samples.

[0055] In this embodiment, the data such as the independent nodes within the station, the lines within the station, the communication nodes within the station and the related nodes outside the station are mainly used to describe the power grid topology within the plant station; wherein, the lines within the station include the connection equipment within the station, and the connection equipment within the station includes the transformer.

[0056] Specifically, the independent node within the station is a node that is not connected to related nodes outside the station, and includes the generator node and the load node; wherein the number of the communication nodes within the station is no more than d.

[0057] Specifically, in the power grid topology, the state of the switching device is used as a node of the topology, wherein the state of the switching device includes a connected state and a disconnected state;

[0058] More specifically, this embodiment uses the states of all switchgear as topological nodes, which undoubtedly includes the intermediate process of switching between different power flow states, along with some intermediate states of the entire network topology connection. Pre-calculating and then storing the data within the station is acceptable due to the slow changes within a single power plant.

[0059] In this embodiment, if the CIM model information of the station is known, the status of two types of switch components, circuit breakers and disconnectors, is traversed. The normal operation status topology of the station is statistically analyzed and pre-stored. Only when new equipment is added to the station is the possible status traversed and stored again.

[0060] like Figure 1 As shown in the figure, taking the dual-mother main connection of a power plant or substation as an example, it includes 6 high-voltage lines (WL1, WL2, WL3, WL4, WL5 and WL6) and 2 transformers as incoming lines. The specific connection equipment of the two transformers depends on the type of power plant or substation. For power plants, the transformers are connected to generators, while for substations, they are connected to other power plants and substations. The 6 outgoing lines are connected to multiple different power plants and substations.

[0061] Among them, 6 high-voltage lines are connected to at most 6 different power plants and stations;

[0062] The disconnectors or circuit breakers in the substation are allowed to present a topological state of either high-voltage or low-voltage sections in the power flow sample. The traversal method is used to set the states of each switch, eliminating unreasonable operating states, and then performing a topological analysis to obtain all possible connection states in the station.

[0063] Specifically, in this embodiment, if a random small-scale topology modification / uncertain plant-specific CIM model is adopted, the grid topology structure can be modified by splitting or merging the intra-station contact nodes; merging two intra-station contact nodes of the same voltage level into one intra-station contact node, there is one topological connection situation; for an intra-station contact node connected to x intra-station lines and a external nodes, it is split into two intra-station contact nodes, and the connection situation is species, where x is a positive integer and x≤d.

[0064] like Figure 2As shown, the intra-station communication nodes connected to multiple devices are split or merged. In the case of 6 off-station related nodes, 2 independent intra-station nodes, and 2 intra-station lines, there may be 1 or 2 intra-station communication nodes. The number of intra-station communication nodes after splitting is no more than the number of intra-station lines. There are 1 possible situation for a single intra-station communication node and 2 possible topological connection situations for two intra-station communication nodes. It should be noted that the main electrical connection type of the plant will affect the topological state, and some states cannot be achieved due to the number of switchgear.

[0065] Specifically, in the plant-station line topology unit, for a power network with m plant-station lines, if one line is removed, there are m situations, and if n lines are removed, there are The topology of the entire network is determined by removing the selected n lines and then performing a topological analysis in conjunction with the in-station topology to remove in-station contact nodes and independent nodes that have no electrical connection with nodes outside the station due to line failure.

[0066] Specifically, in the grid state quantity generation module, for a single load scenario, different unit output combinations are generated by exchanging values, and then the output is modified according to a predetermined difference while ensuring active power balance.

[0067] Specifically, in the grid state quantity generation module, for a network with N nodes under a single topology, a certain generator node is designated as a balanced node. Given the voltage amplitudes of the N nodes and the voltage phase angle differences between the N-1 unbalanced nodes and the balanced node, a power flow calculation sample can be determined, that is, 2N-1 characteristic values ​​are used as the unique identifier of a single power flow calculation sample.

[0068] Considering the issue of similar topologies, current methods for generating similar topologies affect the total number of nodes but do not affect the number of load and generator nodes. Using only the voltage amplitudes and voltage phase angles of load and generator nodes can be used as characteristic identifiers to describe the grid state, but due to issues such as multiple power flow solutions, this cannot uniquely determine the grid state. For example, for a network with M load and generator nodes, the characteristic values ​​are M voltage amplitudes and M-1 node voltage phase angle differences, for a total of 2M-1 characteristic values ​​that can serve as characteristic identifiers.

[0069] Based on the output of a specific generator, calculate the relative values ​​of each generator's output and load as the active power distribution characteristics. For example, if generator e is the base, its output is 100MW, generator f is 50MW, and load g is 90MW, then the active power distribution characteristics are arranged as [1, 0.5, 0.9] for e, f, and g;

[0070] For the power grid, the injected active power of the load node is more valuable as a reference, allowing for faster matching to similar scenarios and being considered as belonging to the same power flow problem. Gridding is used to split power flow samples, with a single grid considered a load scenario. The load value interval of the load node serves as a management identifier.

[0071] For example, for each load node within its possible range, a step size of 100MW is used as the benchmark demand. For example, if the possible load range of any load node is 0-900MW, there are 10 benchmark demands. Taking half of the step size as the basis for division, for a flow sample with 4 load nodes, if its load demand is [480, 710, 890, 770], it is classified as a load demand scenario of [500, 700, 900, 800].

[0072] More specifically, for each power flow calculation sample, the voltage phase angle of the node is obtained as a unique identifier, and the management identifier and the characteristic identifier are determined to calculate the active power distribution characteristics.

[0073] like Figure 3 As shown, this embodiment also provides a method for generating power system similar topology flow samples, which is applied to the above-mentioned power system similar topology flow sample generation system. The generation method is as follows:

[0074] Step S1: pre-set a basic power flow sample; set the load of the given load node to the maximum value, and set the CIM connection model of the entire network; wherein the load node and the generator node are independent nodes within the station and are not directly connected to related nodes outside the station;

[0075] Step S2: First, different power grid topologies are generated, the intra-station contact nodes in the power grid topologies are modified, and then the line states in the plant-station line topologies are modified to obtain multiple similar power grid topologies reflected by basic power flow samples;

[0076] Step S3: First, based on the maximum load value, multiple load scenarios are generated; the output of the generators is evenly distributed to generate basic samples under a single topology and single load scenario; the output of the units is adjusted by using the output exchange method; the power flow calculation is performed using the Newton method to obtain power flow calculation samples;

[0077] Step S4: determining whether the power flow of the power flow calculation sample has converged;

[0078] Step S5: If the power flow calculation sample converges, the characteristic values ​​and identifiers in the power flow calculation sample are extracted to the power grid sample storage and management module for storage and management.

[0079] Specifically, in step S4, if the flow of the flow calculation sample does not converge, the voltage and amplitude of the flow sample with similar active power distribution characteristics in the generated flow sample library are searched as the initial value of the Newton method flow calculation, and then the Newton method flow calculation is used again for the initial value to obtain the flow calculation sample, and then it is re-judged whether the flow of the newly generated flow calculation sample converges. If it is determined to converge, it goes to step S5, records the number of times the flow calculation is recalculated after the search, and stops searching if it fails to converge after exceeding the specified number of times, and saves the failed flow calculation sample for manual analysis.

[0080] Given a basic sample adjusted by experts, the maximum load at a given load node, and the CIM connection model for the entire network, load nodes and generator nodes are usually independent nodes within the station and not directly connected to related nodes outside the station.

[0081] The specific process of a method for generating power system similar topology power flow samples is as follows:

[0082] First, different power grid topologies are generated, the intra-station contact nodes are modified, and the line states in the plant-station line topology are modified to obtain other possible topological states reflected by the basic power flow samples.

[0083] Secondly, for a single fixed topology, the three steps are load determination, unit output determination, and random voltage and reactive power adjustment. For a given load node's demand range, the total load is calculated, and the unit output combination is adjusted based on the base sample output combination. First, the generator output is set as evenly as possible. Two generators are randomly selected to exchange partial outputs, dP. One generator's output increases by dP, while the other decreases by dP. The magnitude of dP is related to the upper and lower limits of the generator output and is set to an integer multiple of a fixed value, p, to ensure a significant difference in output after adjustment. For a single generator, if its upper limit is Pmax and its lower limit is Pmin, the uniform output under a certain load scenario is Pout. The adjustable range is (Pmax - Pout) / p + (Pout - Pmin) / p = w. First, the total number of load changes is specified, selecting the number of positive and negative changes.

[0084] For example, a single generator can adjust up to 5 steps and down to 5 steps. Therefore, the sum of the adjustment steps for all generators should be 0. The sum of positive adjustments should equal the sum of negative adjustments. Define the total number of positive adjustments as the load. For example, if there are 10 generators and the total load adjustment is 10 steps, break 10 into multiple positive integer combinations, such as {1, 5, 1, 3}, where the number of adjustments does not exceed the generator's adjustable capacity. Then, select generators to increase output and select some combinations to decrease output. The generators adjusted up and down must not overlap.

[0085] A single power flow sample is obtained by calculating the Newton-Raphson method. Because the power flow method may not converge, among the generated power flow samples, samples with similar active power output distribution are found. Their voltage phase angles are used as the initial values ​​of the Newton-Raphson method, and recalculation is performed to improve convergence.

[0086] For each power flow sample, the node voltage phase angle is obtained as a unique identifier, and the management identifier and feature identifier are determined to calculate the active power distribution characteristics.

[0087] First, this embodiment generates similar topologies of flow samples based on plant information or grid CIM (Common Information Model) model information, effectively utilizing the existing grid structure and information, making the generated flow samples closer to the actual grid conditions and improving the practicality and accuracy of the samples.

[0088] Secondly, regarding unit output, this embodiment employs a method of exchanging partial outputs. While ensuring active power balance, unit output combinations are generated based on specified differences. This method can generate diverse unit output combinations, significantly improving the coverage of the sample space and providing a richer and more comprehensive data foundation for subsequent power flow analysis and grid optimization.

[0089] Again, for the unconverged power flow scenario, this embodiment searches for samples with similar active power distribution characteristics as the initial values ​​for the power flow calculation, effectively utilizing the information of existing samples, reducing the difficulty of power flow calculation convergence, and improving calculation efficiency and stability.

[0090] Furthermore, this embodiment defines the node voltage phase angle as a unique storage identifier for samples, the voltage phase angles of generator and load nodes as feature identifiers, and the load value of the load node as a management identifier. This identification method makes sample data management clearer and more organized, while also facilitating subsequent feature extraction and analysis.

[0091] Finally, by calculating the relative output load ratio as the active power distribution feature, this embodiment provides a concise and intuitive way to describe the distribution of active power in the power grid, which helps to quickly identify the operating status and potential problems of the power grid.

[0092] Based on the disclosure and teachings of the above description, those skilled in the art may also make changes and modifications to the above embodiments. Therefore, the present invention is not limited to the specific embodiments disclosed and described above, and modifications and variations of the present invention should also fall within the scope of protection of the claims of the present invention. In addition, although certain specific terms are used in this description, these terms are only for convenience of description and do not constitute any limitation to the present invention.

Claims

1. A system for generating power system topology-similar power flow samples, characterized in that: include: A basic power flow sample includes pre-defined nodes and a network-wide CIM connection model; wherein a portion of the nodes are designated as load nodes and another portion of the nodes are designated as generator nodes; and the load of each of the load nodes is set to a maximum value; A power grid topology generation module includes an intra-station topology processing unit and a plant-station line topology unit; the intra-station topology processing unit is used to generate a power grid topology structure and obtain multiple similar power grid topologies reflected by the basic power flow samples by changing the intra-station topology state or the state of the intra-station lines; The power grid topology structure includes an off-station topology structure and an on-station topology structure. The off-station topology structure includes a off-site related nodes and multiple off-site lines, the on-site topology includes b Independent nodes within the station, d The said intra-station lines and multiple intra-station contact nodes, wherein, a 、 b 、 d are all positive integers; each of the off-site related nodes is connected to the on-site contact node via the off-site line, and the on-site contact node is connected to the on-site independent node via the on-site line; The in-station topology status includes the status of each independent node in the station, the status of each in-station line and the status of each in-station contact node; The plant-station line topology unit modifies the plant-station line topology structure by removing the plant-station line to adapt to the situation where the power system is shut down for line maintenance; The grid state quantity generation module is used to generate multiple load scenarios based on the maximum load value. It evenly distributes the output of the generators to generate basic samples for a single topology and a single load scenario. It adjusts the output of the units by exchanging outputs. It uses the Newton-Ray method to calculate the power flow and obtain power flow calculation samples. A power grid sample storage and management module stores and manages the characteristic values ​​and identifiers in the power flow calculation samples; The method for generating the power system similar topology flow sample generation system is as follows: Step S1: pre-set a basic power flow sample; set the load of the given load node to the maximum value, and set the CIM connection model of the entire network; wherein the load node and the generator node are independent nodes within the station and are not directly connected to related nodes outside the station; Step S2: First, different power grid topologies are generated, the intra-station contact nodes in the power grid topologies are modified, and then the line states in the plant-station line topologies are modified to obtain multiple similar power grid topologies reflected by basic power flow samples; Step S3: First, based on the maximum load value, multiple load scenarios are generated; the output of the generators is evenly distributed to generate basic samples under a single topology and single load scenario; the output of the units is adjusted by using the output exchange method; the power flow calculation is performed using the Newton method to obtain power flow calculation samples; Step S4: determining whether the power flow of the power flow calculation sample has converged; Step S5: If the power flow calculation sample converges, the characteristic values ​​and identifiers in the power flow calculation sample are extracted to the power grid sample storage and management module for storage and management.

2. A system for generating power system similar topology flow samples according to claim 1, characterized in that: The independent node in the station is a node that is not connected to the related nodes outside the station, including the generator node and the load node; wherein the number of the communication nodes in the station is not more than d indivual.

3. A system for generating power system similar topology flow samples according to claim 1, characterized in that: In the power grid topology, the state of the switching device is used as a node of the topology, wherein the state of the switching device includes a connected state and a disconnected state.

4. A system for generating power system similar topology flow samples according to claim 1, characterized in that: For the grid topology, the topology modification is performed by splitting or merging the intra-station contact nodes; merging two intra-station contact nodes of the same voltage level into one intra-station contact node, and there is one topological connection situation; for the connected x In-station lines, a The topological contact node of an external node in the station is split into two internal contact nodes. The connection status is species, among which x is a positive integer, and .

5. The system for generating power system similar topology power flow samples according to claim 1, characterized in that: In the plant line topology unit, for m The power network of the plant-station lines, excluding one line, has m In this case, eliminate n Lines, there are situations, among which m、n are all positive integers; by taking the selected n After removing the lines, a topology analysis is performed in conjunction with the in-station topology to remove in-station contact nodes and independent nodes that have no electrical connection with nodes outside the station due to line failure. The resulting in-station topology status can be used to determine the topology of the entire network.

6. A system for generating power system similar topology power flow samples according to claim 1, characterized in that: In the grid state quantity generation module, for a single load scenario, different unit output combinations are generated by exchanging values, and then the output is modified according to a predetermined difference while ensuring active power balance.

7. The system for generating power system similar topology flow samples according to claim 1, characterized in that: In the grid state quantity generation module, for a network with N nodes under a single topology, a certain generator node is designated as a balanced node. Given the voltage amplitudes of the N nodes and the voltage phase angle differences between the N-1 unbalanced nodes and the balanced node, a power flow calculation sample can be determined, that is, 2N-1 characteristic values ​​are used as the unique identifier of a single power flow calculation sample.

8. A system for generating power system similar topology power flow samples according to claim 7, characterized in that: For each power flow calculation sample, the voltage phase angle of the node is obtained as a unique identifier, and the management identifier and feature identifier are determined to calculate the active power distribution characteristics.

9. A method for generating power system similar topology flow samples, characterized in that: Applied to a system for generating power system similar topology flow samples according to any one of claims 1 to 8, the generation method is as follows: Step S1: pre-set a basic power flow sample; set the load of the given load node to the maximum value, and set the CIM connection model of the entire network; wherein the load node and the generator node are independent nodes within the station and are not directly connected to related nodes outside the station; Step S2: First, different power grid topologies are generated, the intra-station contact nodes in the power grid topologies are modified, and then the line states in the plant-station line topologies are modified to obtain multiple similar power grid topologies reflected by basic power flow samples; Step S3: First, based on the maximum load value, multiple load scenarios are generated; the output of the generators is evenly distributed to generate basic samples under a single topology and single load scenario; the output of the units is adjusted by using the output exchange method; the power flow calculation is performed using the Newton method to obtain power flow calculation samples; Step S4: determining whether the power flow of the power flow calculation sample has converged; Step S5: If the power flow calculation sample converges, the characteristic values ​​and identifiers in the power flow calculation sample are extracted to the power grid sample storage and management module for storage and management.

10. The generation method according to claim 9, characterized in that In step S4, if the flow of the flow calculation sample does not converge, the voltage and amplitude of the flow sample with similar active power distribution characteristics in the generated flow sample library are searched as the initial value of the Newton method flow calculation, and then the Newton method flow calculation is re-applied to the initial value to obtain the flow calculation sample, and then the flow of the newly generated flow calculation sample is re-judged whether it converges. If it is determined to converge, it goes to step S5.

Images