Method for ascertaining the overload probability of one or more network nodes of a power network
Patent Information
- Application Number
- EP2024721877
- Authority / Receiving Office
- EP · EP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-04-20
- Filing Date
- 2024-04-12
- Publication Date
- 2026-01-28
AI Technical Summary
Power network planners face challenges in designing effective network expansion measures due to the lack of trustworthy and up-to-date network models, particularly in decentralized renewable energy and electric vehicle integration scenarios, where existing models are often outdated or not digitally available, leading to incomplete knowledge of network topology.
A method using a random growth algorithm to generate multiple plausible network topologies based on a basic topology, assigning probabilities to each, and determining overload probabilities for network nodes, allowing for uncertainty-aware reconstruction and robust design decisions.
Enables accurate determination of overload probabilities for power network nodes, considering multiple possible topologies and their likelihood, providing a coherent and scalable approach for network expansion planning despite incomplete knowledge of the actual topology.
Smart Images
Figure EP2024059998_24102024_PF_FP_ABST
Abstract
Description
[0001] Description
[0002] Method for determining the probability of overload of one or more network nodes of a power grid
[0003] The invention relates to a method according to the preamble of patent claim 1, a control unit for controlling or regulating a power grid according to the preamble of patent claim 12 and a computer program product according to the preamble of patent claim 13.
[0004] Power grids are undergoing significant change due to the increasing spread of decentralized renewable energy, electric vehicle (EV) charging stations, and heat pumps. Energy suppliers must increasingly implement grid expansion measures to integrate these decarbonization measures without violating the system's operating limits. However, to design suitable grid expansion measures, grid planners require the most adequate grid model of the existing infrastructure possible.
[0005] However, trustworthy and easy-to-use grid models are often unavailable. Distribution grids were usually planned and built decades ago, and in some cases, grid models no longer exist. Sometimes grid models are available only on paper and not digitally. Sometimes even digital grid models exist, but these are unreliable because they have been incorrectly or not continuously updated following the numerous grid modifications over the years. As a result, grid planners often have to reconstruct corresponding models of the existing distribution grids, relying on sparse information, particularly regarding connection topology.
[0006] For this purpose, an approach for uncertainty-aware reconstruction of the grid topology is advantageous in order to enable appropriate subsequent grid analyses and to make more robust design decisions for grid expansion measures. The present invention is based on the object of providing an improved method for determining an overload probability for a power grid when the grid topology is not completely known.
[0007] The object is achieved by a method having the features of independent patent claim 1, by a control unit having the features of independent patent claim 12, and by a computer program product having the features of independent patent claim 13. Advantageous embodiments and further developments of the invention are specified in the dependent patent claims.
[0008] The method according to the invention for determining an overload probability of one or more network nodes of a power grid , wherein the power grid has at least one local network station as starting node and several end consumers as end nodes , is characterized by the following steps :
[0009] - Determining a basic network topology for the power grid, which includes at least the start nodes and the end nodes;
[0010] - Creating multiple possible network topologies for the
[0011] Power grid using a random growth algorithm in which one or more lines of the power grid between its network nodes are randomly generated based on the basic network topology;
[0012] - Determine a probability pt for each of the generated network topologies G^; and
[0013] - Determining the overload probability for one or more network nodes based on the generated ensemble of network topologies Gi and the probability pi associated with the respective network topology Gi.
[0014] The order of the steps of the procedure does not imply a chronological order of the steps. In particular, these
[0015] - as far as possible - be carried out in parallel. In other words, one or more steps of the method are preferably carried out in parallel. The method according to the invention and / or one or more functions, features and / or steps of the method according to the invention and / or one of its embodiments can be computer-aided. In particular, a control unit provided for controlling or regulating the power grid comprises one or more computing units which are designed and configured to carry out the method and to determine the distributions, for example by means of commands.
[0016] The power grid is an electrical distribution network, in particular a low-voltage network and / or a medium-voltage network. The power grid has several network nodes. Furthermore, the power grid typically has several lines that extend from one of the network nodes to another of the network nodes. The topology of the power grid can comprise strands and / or rings.
[0017] In a first step of the method, a basic topology of the power grid is determined. This basic topology includes the known starting nodes (local network substations) and end nodes (end users). The basic topology can be determined, in particular, based on the route of a road (road layout) in the area under consideration. This is because power grid power lines typically follow the route of the road. This allows the basic topology, which is a starting point for the method, to be determined.
[0018] In a second step of the process, several possible network topologies for the power grid using a random growth algorithm, in which growth algorithm one or more lines of the power grid between its network nodes are randomly generated based on the basic network topology.
[0019] A network topology typically defines radial connectivity. This is particularly advantageous for low-voltage or medium-voltage networks, as these are essentially operated radially. Ring circuits, particularly in medium-voltage networks, are typically only activated in the event of a fault. Thus, the power grid between end users and distribution substations can be mathematically described as an acyclic graph radially =(V>E), where V denotes the set of network nodes, for example local network stations, end users and intermediate nodes connecting the end users to the local network stations, and E denotes the set of edges, for example electrical lines and / or connections to the end users.
[0020] The graph P ra the a l is typically a collection of K disjoint trees (K corresponds to the number of available local network stations) and each tree T k includes one of the local network stations k as a so-called root node (initial node), a subset of the nodes V Tk c V connected to a subset of edges Ey k c E.
[0021] The procedure for determining Pradial , i.e. a network topology , can now be constructed as follows :
[0022] - Starting from the basic topology hase which describes the network region, a random growth model generates a random, but technically realistic, and in particular coherent, allocation between available local network stations and end users.
[0023] - According to the assignment, Phase is then segmented into K subgraphs. Finally, each subgraph is converted into a radial topology T k derived, for example, using a Steiner tree problem formulation.
[0024] - The growth model (growth algorithm) together with the derivation of a radial network topology is iterated to generate an ensemble of radial network topologies.
[0025] Furthermore, the network topologies are weighted in the third step of the process. For this purpose, each generated network topology is assigned a probability, i.e. a weighting. The weighting is therefore based on an estimate of how likely one of the generated network topologies corresponds to the actual network topology. This can be achieved, for example, by evaluating the extent to which common design and / or operating criteria, such as the costs of the respective topology, are met in a generated network topology, and / or by analyzing and evaluating measured data in the context of the generated topology and / or its aggregated line length.
[0026] In summary, after the above steps, an ensemble of radial network topologies is available that meet common design and / or operational criteria and are weighted with probability weights. The topologies thus generated and provided can be used according to the present invention to determine the overload probability.
[0027] In the fourth step of the procedure, the overload probability for one or more network nodes is determined based on the generated ensemble of network topologies Gi and the probability pi associated with the respective network topology Gi.
[0028] The invention has one or more of the following advantages:
[0029] - Coherent method combining heterogeneous data sources, such as available data on the electricity distribution system, such as location of secondary substations, available real measurements of end-users and the like, and open source data, such as road layouts;
[0030] - Creating and considering an ensemble of several possible plausible network topologies instead of just a single most probable network topology when only imperfect knowledge of the actual network topology is available;
[0031] - Simple and intuitive approach for generating possible network topologies that are easily scalable across regions while reflecting the underlying geographical characteristics;
[0032] - Separation between the model for generating network topologies and the model for state estimation in order to maximize the exploration of different plausible network topologies of the system, including rare events;
[0033] - Use of available network topology data, for example from measurements, to derive the probability distribution over the ensemble.
[0034] The control unit according to the invention for controlling or regulating a power grid with a plurality of network nodes and lines, comprising a computing unit, is characterized in that the computing unit is designed and configured to carry out a method according to one of the preceding claims, wherein the control unit is designed to carry out grid-friendly and / or system-friendly control measures depending on the overload probability determined by the computing unit.
[0035] Similar, equivalent and equivalent advantages and / or configurations of the control unit according to the invention result from the method according to the invention.
[0036] The computer program product according to the invention is characterized in that it comprises instructions which, when the program is executed by a computing unit, in particular a computer, cause the latter to carry out a method and / or steps of the method according to one of claims 1 to 11.
[0037] Similar, equivalent and equivalent advantages and / or embodiments of the computer program product according to the invention result from the method according to the invention.
[0038] According to an advantageous embodiment of the invention, the overload probability of the local network station is determined. This advantageously allows for the determination of which of the local network stations is likely to be overloaded.
[0039] In an advantageous development of the invention, the basic network topology is determined by means of a road layout (road course) which is associated with the power grid.
[0040] This allows an advantageous base topology to be provided that is as close as possible to the actual network topology.
[0041] According to an advantageous embodiment of the invention, data provided for the power grid about its network topology and / or real-time measurement data relating to the power grid are used in determining the basic network topology.
[0042] In an advantageous development of the invention, a Monte Carlo simulation is used for the growth algorithm.
[0043] According to an advantageous embodiment of the invention, the network topologies Gi are generated in such a way that they each form an acyclic connected graph which extends from the starting node to the end nodes.
[0044] In an advantageous development of the invention, the probability pi of each network topology Gi is calculated using a Boltzmann distribution with respect to the aggregated line length L G . the respective network topology Gi is determined.
[0045] For example, these are given by pi = where A is the dependence of the probability on the aggregated line length L G . the respective network topology Gi is modeled .
[0046] According to an advantageous embodiment of the invention, the overload probability of the local network station is determined according to
[0047] Pt ■ max t XP t >p0calculated , where P ttime-dependent electrical power aggregated at the local network station, the indicator function and P o a specified performance threshold .
[0048] Advantageously, the resulting overload probability takes into account the generated possible network topologies and their respective probabilities. This allows for a more accurate prediction of the probability of an overload.
[0049] In an advantageous further development, the power grid has a plurality of local network stations, wherein the power grid is divided into a plurality of sub-grids, each of which comprises exactly one of the local network stations, and for each of the sub-grids a respective overload probability of one or more of its network nodes is determined according to a method according to one of claims 1 to 7.
[0050] This advantageously results in segmentation into subnetworks.
[0051] According to an advantageous embodiment of the invention, the respective overload probability is determined for each of the local network stations.
[0052] In an advantageous development of the invention, the power grid comprises a medium-voltage grid and / or low-voltage grid.
[0053] This is an advantage because the network topology is typically unknown or not fully known in medium-voltage networks and / or low-voltage networks.
[0054] Further advantages, features, and details of the invention will become apparent from the exemplary embodiments described below and from the drawings. The drawings schematically show: Figure 1 is a flow diagram of a method according to one embodiment of the present invention;
[0055] Figure 2 shows a region with a power grid with several local network stations; and
[0056] Figure 3 is a diagram showing the probability of overload of a local network station.
[0057] Elements of the same type, value or effect may be provided with the same reference symbols in one or more of the figures.
[0058] Figure 1 shows a flowchart of a method according to one embodiment of the present invention. The start of the method is marked with S. The end of the method is marked with E.
[0059] In a first step S1 of the method, a basic topology is determined. This is done, in particular, based on a known spatial distribution of local network substations, end users, and / or a road layout in the region under consideration. The use of the aforementioned input data to generate the basic topology is indicated by reference numeral 21.
[0060] In a second step S2 of the process, several additional possible network topologies are determined from the base topology using a random growth model. This is preferably done using a Monte Carlo method. This is repeated until N topologies have been generated.
[0061] In a third step S3 of the method, a probability is determined for each generated topology. In other words, the generated topologies are weighted. For this purpose, additional input data 22, for example, domain knowledge and / or additional parameters and / or line lengths, are used.
[0062] In a fourth step S4 of the method, the overload probabilities of the local network stations are determined based on the generated ensemble of topologies.
[0063] Figure 2 shows an exemplary power grid 1 with several local network stations 11 and several end consumers 12. For clarity, not all end consumers 12 are provided with a reference symbol.
[0064] Furthermore, the depicted region includes a road layout 13 associated with the power grid 1 (road layout). The depicted road layout 13 as well as the spatial positions of the local network substations 11 and the end users 13 can be used to determine the base topology. It is assumed that the actual topology essentially follows the road layout 1. Thus, the base topology essentially corresponds to the road layout.
[0065] Figure 3 shows an exemplary diagram of the overload probability of a local network station in a power grid.
[0066] The electrical load in kilowatt hours is plotted on the abscissa 100 of the diagram. The probability for each load is plotted on the ordinate 101 of the diagram. A power threshold is marked with the reference symbol 32. The most probable grid topology and its associated electrical load are marked with the reference symbol 31. If only the most probable grid topology is considered, as in the prior art, no overload situation arises. However, it can be seen from the diagram and the graph that for certain grid topologies within the generated ensemble, power levels above the power threshold 32 can occur. This must be taken into account when controlling the power grid and in future expansions of the power grid.Advantageously, the present invention makes this possible since not only the most probable network topology but an ensemble of possible network topologies is taken into account in the method.
[0067] Although the invention has been illustrated and described in detail by the preferred embodiments, the invention is not limited by the disclosed examples and other variations can be derived therefrom by those skilled in the art without departing from the scope of the invention.
[0068] Reference symbol list
[0069] S Start
[0070] S1 first step S2 second step
[0071] 53 third step
[0072] 54 fourth step
[0073] E End
[0074] 1 power grid 11 local network station
[0075] 12 end users
[0076] 13 Street layout
[0077] 21 Input data
[0078] 22 Input data 31 Most likely topology
[0079] 32 Performance threshold
[0080] 100 Abs zisse
[0081] 101 Ordinate
Claims
Patent claims 1. A method for determining an overload probability of one or more network nodes (11, 12) of a power grid (1), wherein the power grid (1) has at least one local network station as a starting node (11) and several end consumers as end nodes (12), characterized by the following steps: - (Sl) determining a basic network topology for the power grid (1) which comprises at least the start nodes (11) and the end nodes (12); - (S2) Generating several possible network topologies for the power grid (1) by means of a random growth algorithm, in which one or more lines of the power grid (1) are randomly generated between its network nodes (11, 12) based on the basic network topology; - (S3) Determining a probability Pt for each of the generated network topologies G^; and - (S4) determining the overload probability (42) for one or more network nodes (11, 12) based on the generated ensemble of network topologies Gi and the probability pi associated with the respective network topology Gi.
2. Method according to claim 1, characterized in that the overload probability of the local network station (11) is determined.
3. Method according to claim 1 or 2, characterized in that the basic network topology is determined by means of a road layout (13) which is associated with the power grid (1).
4. Method according to one of the preceding claims, characterized in that data provided for the power grid (1) about its network topology and / or real-time measurement data relating to the power grid (1) are used in determining the basic network topology.
5. Method according to one of the preceding claims, characterized in that a Monte Carlo simulation is used for the growth algorithm.
6. Method according to one of the preceding claims, characterized in that the network topologies Gi are generated in such a way that they each form an acyclic connected graph which extends from the start node (11) to the end nodes (12).
7. Method according to one of the preceding claims, characterized in that the probability pi of each network topology Gi is calculated using a Boltzmann distribution with respect to the aggregated line length L G . of the respective network topology Gi is determined.
8. Method according to one of the preceding claims, characterized in that the overload probability of the local network station (11) according to Y^=iPi ' max tXp t >p0is calculated, where P ttime-dependent electrical power aggregated at the local network station (11), the indicator function and P o a specified power threshold (32).
9. Method according to one of the preceding claims, characterized in that the power grid (1) has a plurality of local network stations (11), wherein the power grid (1) is divided into a plurality of sub-grids, each of which comprises exactly one of the local network stations (11), and for each of the sub-grids a respective overload probability of one or more of its network nodes (11, 12) is determined according to a method according to one of claims 1 to 7.
10. Method according to claim 9, characterized in that the respective overload probability of the local network stations (11) is determined.
11. Method according to one of the preceding claims, characterized in that the power grid (1) comprises a medium-voltage grid and / or a low-voltage grid.
12. Control unit for controlling or regulating a power grid (1) with a plurality of network nodes (11, 12) and lines, comprising a computing unit, characterized in that the computing unit is designed and configured to carry out a method according to one of the preceding claims, wherein the control unit is designed to carry out grid-friendly and / or system-friendly control measures depending on the overload probability determined by the computing unit.
13. Computer program product, comprising instructions which, when the program is executed by a computing unit, in particular a computer, cause the latter to carry out a method and / or steps of the method according to one of claims 1 to 11.