A power distribution network optimal dispatching method and system

Abstract

The application relates to the field of power distribution and discloses a power distribution network optimization scheduling method and system, wherein the method comprises the following steps: constructing a power-communication coupling topological structure; constructing a power-communication coupling topological model based on the power-communication coupling topological structure; constructing a network state map according to the power-communication coupling topological model; when detecting that the node power of the power distribution network is overloaded, acquiring a power supply path set that can safely bear the transferred load according to the power grid physical topological structure; generating a communication constraint according to the network state map; screening a power supply path subset that meets the communication constraint from the power supply path set according to the communication constraint; and evaluating the power supply path subset, and if there are concurrent instructions, performing multi-instruction optimization collaborative scheduling. Through the method and the corresponding system, the coupling relationship between the power grid and the communication network can be collaboratively modeled and dynamically perceived, and the collaborative optimization distribution of the communication resources and the power supply paths can be realized under the multi-instruction concurrent condition.

Description

Technical Field

[0001] This invention relates to the field of power distribution, and in particular to a method and system for optimizing the scheduling of power distribution networks. Background Technology

[0002] Traditional power grid optimization and dispatching methods do not fully consider the coupling relationship between the power grid and the communication network. The transmission of dispatching commands depends on the underlying communication network, and parameters such as link latency, packet loss rate, and bandwidth utilization directly affect the timeliness and reliability of control commands. Even if the electrical performance of a power supply path meets the requirements, if the quality of the communication link along it is poor, dispatching commands may still fail to be delivered on time and accurately, leading to control failures or even system risks.

[0003] Secondly, existing methods typically select paths with the goal of optimizing a single task, lacking a system-level coordination mechanism. For example, the performance difference between the optimal and suboptimal paths of the first scheduling task may be small, but this path is also strongly relied upon by the second scheduling task, whose optimal and suboptimal paths have significant performance differences. Traditional distribution network optimization scheduling methods process tasks according to their timestamp order, prioritizing the allocation of communication and power paths to the first scheduling task, which was generated earlier. The first scheduling task occupies public high-quality path resources, leading to a significant decrease in the performance of the second scheduling task. Summary of the Invention

[0004] Therefore, the purpose of this invention is to overcome the problems in the prior art, which view the power grid and communication network in isolation or only perform simple static associations, failing to construct a unified coupled topology model that can dynamically reflect the interdependence of the states of both parties, and lacking collaborative optimization of the competition relationship of communication resources under multi-task concurrent scenarios. This leads to problems such as delays or loss of dispatching command transmission, deterioration of distribution network control performance, and insufficient robustness of the overall dispatching scheme. The invention provides a distribution network optimized dispatching method and system that collaboratively models and dynamically perceives the coupling relationship between the power grid and communication network, and realizes the collaborative optimization allocation of communication resources and power paths under multi-command concurrent conditions. This improves the timeliness and overall performance of distribution network operation control while ensuring reliable transmission of dispatching commands.

[0005] To address the aforementioned technical problems, this invention provides a method for optimizing the scheduling of a power distribution network, the method comprising:

[0006] Obtain the power grid physical topology and communication network topology of the distribution network, and construct a power-communication coupled topology based on the power grid physical topology and communication network topology;

[0007] Obtain the power operation parameters and communication link parameters of each node in the power distribution network, and associate the power operation parameters and communication link parameters as attribute values ​​with the edges of the corresponding nodes in the power-communication coupled topology to obtain the power-communication coupled topology model;

[0008] Construct a network state map based on the described power-communication coupled topology model;

[0009] When a node power overload is detected in the distribution network, a set of power supply paths that can safely carry the transferred load is obtained based on the physical topology of the power grid; communication constraints are generated based on the network status map; and a subset of power supply paths that meet the communication constraints are selected from the set of power supply paths based on the communication constraints.

[0010] The power supply path subset is evaluated, and if concurrent instructions exist within a preset time window, multi-instruction optimized collaborative scheduling is performed.

[0011] Preferably, the process involves obtaining the power grid physical topology and communication network topology of the distribution network, and constructing a power-communication coupled topology based on the power grid physical topology and communication network topology, including:

[0012] Define the power entities in the distribution network as a first node set, obtain the power transmission relationship between each node in the first node set, and construct the power grid physical topology of the distribution network based on the power transmission relationship;

[0013] Define the communication entities in the power distribution network as a second node set, obtain the communication connection relationships between each node in the second node set, and construct the communication network topology of the power distribution network based on the communication connection relationships;

[0014] Obtain the mapping relationship between the first node in the first node set and the second node in the second node set, and associate the power grid physical topology and the communication network topology based on the mapping relationship to obtain the power-communication coupled topology.

[0015] Preferably, the mapping relationship is that the data of the power entities in the first node set is transmitted by the communication entities in the second node set.

[0016] Preferably, the power operation parameters include: switch position status, node voltage, and distributed resource status; the communication link parameters include: end-to-end delay, packet loss rate, and bandwidth utilization.

[0017] Preferably, constructing a network state map based on the power-communication coupled topology model includes:

[0018] Historical power operation parameters and historical communication link parameters are collected based on the coupling topology relationship in the power-communication coupled topology model.

[0019] The historical power operation parameters and historical communication link parameters are cleaned and aligned to construct a supervised learning training set;

[0020] The training set is input into a prediction model built on a long short-term memory network. The prediction model predicts the probability distribution of performance parameters of each link within a preset time window. The probability distribution of performance parameters includes the latency, packet loss rate and corresponding variance and confidence interval of the communication link.

[0021] Based on the probability distribution of each link performance parameter output by the prediction model, the probability distribution of each link performance parameter is injected as attribute value into the corresponding edge in the power-communication coupled topology model to construct the network state map.

[0022] Preferably, generating communication constraints based on the network state map includes:

[0023] For scheduling instructions, extract their communication requirements and generate requirement constraint triples. ,in, This indicates the maximum end-to-end delay that the scheduling instruction can tolerate. This indicates the maximum end-to-end packet loss rate that the scheduling instruction can tolerate. Indicates scheduling instructions The amount of data to be transmitted;

[0024] From the network status map, obtain the performance prediction values ​​of all communication links within the target time window. The performance prediction values ​​include the expected value of latency prediction, the expected value of packet loss rate prediction, and the expected value of available bandwidth prediction.

[0025] For each power supply path in the power supply path set for which the scheduling instruction is given, hard constraints are applied.

[0026] The paths in the power supply path set that satisfy all the hard constraints are collectively denoted as the feasible path set. According to each scheduling instruction and its corresponding set of feasible paths Obtain the communication constraints.

[0027] Preferably, the hard constraints include bandwidth constraints, latency constraints, and packet loss rate constraints, wherein:

[0028] The bandwidth constraint includes: the expected available bandwidth of all links on the power supply path is greater than the minimum bandwidth requirement of the scheduling instruction. The minimum bandwidth requirement of the scheduling instruction is determined as follows: the bandwidth requirement obtained by dividing the data volume of the scheduling instruction by the allowed transmission time.

[0029] The delay constraint includes: determining whether the expected value of the predicted power supply path delay is less than or equal to the maximum end-to-end delay that the scheduling command can tolerate;

[0030] The packet loss rate constraint includes: determining whether the predicted expected value of the packet loss rate of the power supply path is less than or equal to the maximum end-to-end packet loss rate that the scheduling instruction can tolerate.

[0031] Preferably, the power supply path subset is evaluated, and if concurrent instructions exist within a preset time window, multi-instruction optimized collaborative scheduling is performed, including:

[0032] The power supply path subset of the dispatch command is evaluated using the following formula:

[0033] ,

[0034] ,

[0035] ,

[0036] in, Indicates coupling score; This represents the local power performance coefficient of link ij; Represents the communication performance coefficient of link ij; This represents the average input active power of link ij; This represents the average output active power of link ij; This represents the average absolute deviation of the voltage of link ij relative to its rated value. This represents the maximum permissible voltage deviation, and the two together constitute the voltage quality coefficient. This represents the average round-trip delay of link ij. Indicates the maximum allowed delay. represents the average packet loss rate of link ij; m represents the number of link segments contained in the path of the power supply path subset.

[0037] Preferably, if concurrent instructions exist within a preset time window, multi-instruction optimized collaborative scheduling is performed, including:

[0038] For all paths of scheduling instructions within a preset time window, sort them in descending order according to the difference in coupling score between the best path and the second-best path.

[0039] Prioritize scheduling instructions with the largest coupling score difference, assign them the available path with the highest score, and mark that path as occupied, until all scheduling instructions have been assigned paths.

[0040] This invention proposes a power distribution network optimization dispatching system, the system comprising:

[0041] A power-communication coupled topology module is constructed to obtain the power grid physical topology and communication network topology of the distribution network, and to construct a power-communication coupled topology based on the power grid physical topology and communication network topology;

[0042] A power-communication coupled topology model module is constructed to obtain the power operation parameters and communication link parameters of each node in the distribution network, and associate the power operation parameters and communication link parameters as attribute values ​​with the edges of the corresponding nodes in the power-communication coupled topology to obtain the power-communication coupled topology model.

[0043] A network status map construction module is used to construct a network status map based on the power-communication coupled topology model.

[0044] A power supply path subset filtering module is used to obtain a set of power supply paths that can safely carry the transferred load based on the physical topology of the power grid when a node power overload is detected in the distribution network; generate communication constraints based on the network status map; and filter a subset of power supply paths that simultaneously meet the power underload and communication constraints from the power supply path set based on the communication constraints.

[0045] The collaborative scheduling module is used to evaluate the power supply path subset. If there are concurrent instructions within a preset time window, multi-instruction optimized collaborative scheduling is performed.

[0046] Compared with the prior art, the above-described technical solution of the present invention has the following advantages:

[0047] The present invention discloses a distribution network optimization scheduling method and system, which performs collaborative modeling and dynamic perception of the coupling relationship between the power grid and communication, and realizes the collaborative optimization allocation of communication resources and power paths under the condition of multiple concurrent commands, thereby improving the timeliness and overall performance of distribution network operation control while ensuring reliable transmission of scheduling commands.

[0048] Among them, the problem of coordination failure caused by the separate design of power and communication in the traditional scheme is solved by constructing a power-communication coupled topology. The load transfer of the distribution network depends on the transmission of power paths, while the scheduling instructions of the paths, such as starting a power supply path, need to be transmitted through the communication network. Both are indispensable.

[0049] The generated power-communication coupled topology model reflects the actual operating status and reflects in real time whether a certain power path can carry the load and whether its corresponding communication can transmit dispatch instructions;

[0050] Construct a network state map to predict the real-time network status of sending scheduling instructions, and avoid decision-making biases caused by subsequent scheduling relying on static network data;

[0051] The initial screening of the power grid physical topology ensures that all selected power supply paths can safely carry the transferred load, thus avoiding the risk of line burnout and power outage expansion due to path overload from the source; the selected power supply path subset can meet the time delay and packet loss rate requirements for dispatch command transmission, avoiding command loss or delay due to communication problems, ensuring that dispatch commands can be transmitted smoothly and quickly, greatly improving dispatch response efficiency, and avoiding the risk of command delay;

[0052] Improve resource utilization and optimize scheduling efficiency in multi-instruction scenarios by using path evaluation and multi-instruction collaborative scheduling. Attached Figure Description

[0053] To make the content of this invention easier to understand, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings, wherein...

[0054] Figure 1 This is a flowchart of a power distribution network optimization scheduling method in a preferred embodiment of the present invention;

[0055] Figure 2 This is a structural block diagram of a power distribution network optimization dispatching system according to a preferred embodiment of the present invention. Detailed Implementation

[0056] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, so that those skilled in the art can better understand and implement the present invention. However, the embodiments described are not intended to limit the present invention.

[0057] Example 1: As Figure 1 As shown in the figure, an embodiment of the present invention discloses a distribution network optimization scheduling method, the method comprising:

[0058] Obtain the power grid physical topology and communication network topology of the distribution network, and construct a power-communication coupled topology based on the power grid physical topology and communication network topology;

[0059] Obtain the power operation parameters and communication link parameters of each node in the power distribution network, and associate the power operation parameters and communication link parameters as attribute values ​​with the edges of the corresponding nodes in the power-communication coupled topology to obtain the power-communication coupled topology model;

[0060] Construct a network state map based on the described power-communication coupled topology model;

[0061] When a node power overload is detected in the distribution network, a set of power supply paths that can safely carry the transferred load is obtained based on the physical topology of the power grid; communication constraints are generated based on the network status map; and a subset of power supply paths that meet the communication constraints are selected from the set of power supply paths based on the communication constraints.

[0062] The power supply path subset is evaluated, and if concurrent instructions exist within a preset time window, multi-instruction optimized collaborative scheduling is performed.

[0063] In specific application scenarios, such as through a power distribution network GIS system, the connection relationships of various power equipment nodes within the power distribution network are collected. These power equipment nodes include feeders, sectionalizing switches, tie switches, transformers, and load nodes.

[0064] Communication network topology acquisition: Collect communication system ledgers. Substation A is equipped with an optical fiber communication terminal, which is connected to the dispatch center switch through the backbone optical fiber; the FTUs (feeder terminal units) of distribution rooms B and C are interconnected with the communication terminal of substation A through optical fiber; the FTU of distribution room D communicates with the dispatch center through a 4G base station.

[0065] Coupled topology construction: Associate power nodes (substation A, distribution room B / C / D) with corresponding communication nodes (fiber optic terminal, FTU, 4G module). For example, the power node of distribution room B is bound to its FTU communication node, forming a power-communication coupled topology structure with a one-to-one mapping of "power equipment-communication terminal".

[0066] One embodiment of the present invention involves obtaining the physical topology of a power distribution network and the communication network topology, and constructing a power-communication coupled topology based on the physical topology and the communication network topology, including:

[0067] Define the power entities in the distribution network as a first node set, obtain the power transmission relationship between each node in the first node set, and construct the power grid physical topology of the distribution network based on the power transmission relationship;

[0068] Define the communication entities in the power distribution network as a second node set, obtain the communication connection relationships between each node in the second node set, and construct the communication network topology of the power distribution network based on the communication connection relationships;

[0069] Obtain the mapping relationship between the first node in the first node set and the second node in the second node set, and associate the power grid physical topology and the communication network topology based on the mapping relationship to obtain the power-communication coupled topology.

[0070] In specific application scenarios, based on the power distribution network equipment ledger and on-site survey, the first node set can be clearly identified as containing the following power entities. The first node type is categorized and labeled according to function. The first node type includes hub power nodes, distribution nodes, distributed power nodes, and centralized load nodes. The hub power nodes include substation M; the distribution nodes include distribution rooms N1, N2, P1, and P2; the distributed power nodes include PV1-PV6; and the centralized load nodes include L1, L2, and L3. The power connection logic between each node is collected to form a "node-line" association table. The power distribution network GIS system is retrieved to obtain the 10kV main lines between substation M and the distribution rooms (M→N1: line WL1, M→N2: WL2, M→P1: WL3, M→P2: WL4). Using nodes as vertices and power lines as edges, a graph is drawn... To illustrate the physical topology, consider a substation M as the central node, radiating to N1, N2, P1, and P2 via WL1-WL4 respectively. In the branch topology, N1 connects to L1, P1 connects to PV1-PV3 and L2, and P2 connects to PV4-PV6 and L3. N2 is temporarily a backup distribution node. Based on the communication system ledger and equipment installation records, the second node set is identified, containing the following communication entities, which correspond one-to-one or many-to-one with the power entities. A communication network topology is constructed. Based on the "installation location, power source, and functional affiliation of the communication entity," a mapping table for the two types of nodes is created to ensure that the operating data of the power entity can be transmitted through the corresponding communication entity. Power lines are associated with their corresponding communication links, and data for each power line is marked as being transmitted through this communication link, thus obtaining the power-communication coupled topology.

[0071] By introducing the real-time status of the communication network on the basis of a single physical topology model, scheduling no longer relies solely on electrical parameters, but can perceive the entire link information affecting command execution, thus improving the accuracy of status assessment; scheduling decisions are more precise, and the two-network coupled topology plus communication constraint screening mechanism ensures that the selected power supply path has reliable communication support while allowing physical conditions, reducing the risk of scheduling failure; the fault handling response speed is fast, and in the event of overload, alternative paths that meet dual conditions are quickly located and dispatched immediately, shortening power restoration time and reducing the scope of power outages.

[0072] In one embodiment of the present invention, the mapping relationship is that the data of the power entities in the first node set is transmitted by the communication entities in the second node set.

[0073] In one embodiment of the present invention, the power operation parameters include: switch position status, node voltage, and distributed resource status; the communication link parameters include: end-to-end delay, packet loss rate, and bandwidth utilization.

[0074] One embodiment of the present invention involves constructing a network state map based on the power-communication coupled topology model, including:

[0075] Historical power operation parameters and historical communication link parameters are collected based on the coupling topology relationship in the power-communication coupled topology model.

[0076] The historical power operation parameters and historical communication link parameters are cleaned and aligned to construct a supervised learning training set;

[0077] The training set is input into a prediction model built on a long short-term memory network. The prediction model predicts the probability distribution of performance parameters of each link within a preset time window. The probability distribution of performance parameters includes the latency, packet loss rate and corresponding variance and confidence interval of the communication link.

[0078] Based on the probability distribution of each link performance parameter output by the prediction model, the probability distribution of each link performance parameter is injected as attribute value into the corresponding edge in the power-communication coupled topology model to construct the network state map.

[0079] In specific application scenarios, historical data is obtained based on the coupling topology relationship in the power-communication coupled topology model, ensuring that the historical data covers different load scenarios (weekdays / holidays, peak / off-peak). Historical power operation parameters include real-time voltage, load rate, and active power of each power link; historical communication link parameters include transmission delay, packet loss rate, and bandwidth utilization of the corresponding communication link. The two types of parameters are associated and stored in the time series database according to "link ID + timestamp" to form a historical dataset. The data is organized according to the supervised learning logic of input features-output labels. For example, the "historical power operation parameters and historical communication link parameters of a certain link for 120 consecutive seconds" are taken as input; the output label is the "probability distribution statistics of performance parameters for the next 15 minutes (preset time window)" of the link. The dataset is divided into training set, validation set, and test set in a 7:2:1 ratio. LSTM network is used to build a prediction model to address the long-term and short-term dependency characteristics of time series data.

[0080] One embodiment of the present invention generates communication constraints based on the network state map, including:

[0081] For scheduling instructions, extract their communication requirements and generate requirement constraint triples. ,in, This indicates the maximum end-to-end delay that the scheduling instruction can tolerate. This indicates the maximum end-to-end packet loss rate that the scheduling instruction can tolerate. Indicates scheduling instructions The amount of data to be transmitted;

[0082] From the network status map, obtain the performance prediction values ​​of all communication links within the target time window. The performance prediction values ​​include the expected value of latency prediction, the expected value of packet loss rate prediction, and the expected value of available bandwidth prediction.

[0083] For each power supply path in the power supply path set for which the scheduling instruction is given, hard constraints are applied.

[0084] The paths in the power supply path set that satisfy all the hard constraints are collectively denoted as the feasible path set. According to each scheduling instruction and its corresponding set of feasible paths Obtain the communication constraints.

[0085] In specific application scenarios, the network state map is invoked to obtain the predicted performance value of the communication link of the power supply path corresponding to the scheduling instruction; the power supply path and supporting communication link of the scheduling instruction are obtained according to the power-communication coupled topology model; the network state map outputs the predicted performance value of each communication link within the target time window; the network state map outputs the probability distribution value of the performance parameters of each communication link within the target time window; the predicted performance value is the sum of the products of the corresponding performance parameter value and its probability, which is the expected value. Here, the expected value is used for constraint judgment. For each power supply path set of the scheduling instruction, the predicted performance value of its corresponding communication link is checked one by one to see if it meets the constraint triplet of the instruction. The communication constraint of each scheduling instruction is clearly defined as "its constraint triplet + the corresponding feasible path set", which directly supports subsequent multi-instruction collaborative scheduling.

[0086] In one embodiment of the present invention, the hard constraints include bandwidth constraints, latency constraints, and packet loss rate constraints, wherein:

[0087] The bandwidth constraint includes: the expected available bandwidth of all links on the power supply path is greater than the minimum bandwidth requirement of the scheduling instruction. The minimum bandwidth requirement of the scheduling instruction is determined as follows: the bandwidth requirement obtained by dividing the data volume of the scheduling instruction by the allowed transmission time.

[0088] The delay constraint includes: determining whether the expected value of the predicted power supply path delay is less than or equal to the maximum end-to-end delay that the scheduling command can tolerate;

[0089] The packet loss rate constraint includes: determining whether the predicted expected value of the packet loss rate of the power supply path is less than or equal to the maximum end-to-end packet loss rate that the scheduling instruction can tolerate.

[0090] In specific application scenarios, bandwidth constraint determination ensures that the bandwidth of all links in the power supply path meets the standard. The reliability of the power supply path depends on the bandwidth of the weakest link in the path. It is necessary to check whether the expected available bandwidth of each link in the path is greater than the minimum bandwidth requirement of the task. Delay constraint ensures that the total delay of the path does not exceed the threshold, ensuring reliable transmission and timely execution of scheduling instructions, reducing the risk of power distribution network operation, and avoiding instruction failure due to excessive delay. Through delay constraint screening, paths with excessive delay can be excluded. For example, if a fault isolation instruction uses a path with a delay of 25ms, This can delay fault isolation and cause fault propagation; after constraint screening, only paths with latency ≤10ms are retained to ensure second-level command response; avoid command retransmission caused by excessive packet loss rate, packet loss rate constraints can eliminate paths with excessive packet loss rate, such as a path with a packet loss rate of 1.5% which will cause 20% of commands to be retransmitted, delaying load transfer; after constraint screening, only paths with a packet loss rate ≤0.5% are retained, and the one-time command transmission success rate exceeds 99.5%; narrow the path evaluation range, accurately screen feasible paths, reduce invalid path evaluation, and improve scheduling decision efficiency.

[0091] In one embodiment of the present invention, the power supply path subset is evaluated, and if concurrent instructions exist within a preset time window, multi-instruction optimized collaborative scheduling is performed, including:

[0092] The power supply path subset of the dispatch command is evaluated using the following formula:

[0093] ,

[0094] ,

[0095] ,

[0096] in, Indicates coupling score; This represents the local power performance coefficient of link ij; Represents the communication performance coefficient of link ij; This represents the average input active power of link ij; This represents the average output active power of link ij; This represents the average absolute deviation of the voltage of link ij relative to its rated value. This represents the maximum permissible voltage deviation, and the two together constitute the voltage quality coefficient. This represents the average round-trip delay of link ij. Indicates the maximum allowed delay. represents the average packet loss rate of link ij; m represents the number of link segments contained in the path of the power supply path subset.

[0097] In specific application scenarios, modern power supply paths inevitably rely on communication links because control nodes need communication guarantees to perform power transmission scheduling tasks. First, the power performance and communication performance of a single link segment are quantified separately. Then, the coordinated performance of the entire path is evaluated through coupled calculations, ultimately providing a quantifiable basis for path selection; local power performance coefficients. Focusing on the efficiency and stability of power links, the core requirements of power links are efficient transmission and voltage stability, both of which directly determine power supply reliability. Therefore, the formula comprehensively evaluates these two dimensions through their product. This indicates the power transmission efficiency. Power transmission involves losses, and the closer this ratio is to 1, the smaller the losses and the higher the transmission efficiency. This represents the voltage quality coefficient. Voltage deviation is a key indicator of a power system. The smaller the deviation, the closer the coefficient is to 1, and the more stable the voltage. and The use of multiplication instead of addition is to avoid misclassifying links with high efficiency but unstable voltage or stable voltage but low efficiency as high-quality. Only when both efficiency and voltage quality meet the standards will the result approach 1, satisfying the hard requirements of power links. The core requirement of communication links is low latency. Low latency ensures the real-time performance of dispatch commands and low packet loss, while low packet loss ensures command integrity. This represents the latency compliance coefficient; if the average latency exceeds [a certain threshold], it indicates compliance. If the latency compliance coefficient becomes negative, the communication is directly deemed unqualified. The smaller the latency, the closer the coefficient is to 1. Value range: -∞ to 1, actual qualified link is 0 to 1, coupling score To comprehensively evaluate the overall performance of the communication path, which consists of m connected links, the path performance is affected by both the weakest link and the power-communication coordination. Therefore, the formula achieves two main objectives through geometric mean and square root. and Geometric means are taken separately. Geometric means are sensitive to extreme values, so it is important to avoid good links masking poor ones. The power supply path is a series system, and the failure of one link will cause the entire path to fail. Different candidate paths need to be compared fairly. The geometric mean of power and the geometric mean of communication achieve forced coupling and coordination. The product will only be large when both power performance and communication performance are excellent, so as to avoid selecting paths that are excellent in one dimension but have poor coupling. The overall square root is taken to normalize the score. The geometric means of power and communication are both 0~1, and the product is also 0~1. After square root, it still maintains the range of 0~1. By quantifying the performance of two dimensions, eliminating the interference of path length, integrating coupling indicators, and providing a unified standard, two key goals are ultimately achieved: ensuring that the selected power supply path simultaneously meets the requirements of efficient and stable power transmission and reliable and low-latency communication commands, providing an efficient and quantifiable decision basis for multi-command concurrent scheduling, and improving the efficiency and robustness of the overall scheduling system.

[0098] In one embodiment of the present invention, if concurrent instructions exist within a preset time window, multi-instruction optimized collaborative scheduling is performed, including:

[0099] For all paths of scheduling instructions within a preset time window, sort them in descending order according to the difference in coupling score between the best path and the second-best path.

[0100] Prioritize scheduling instructions with the largest coupling score difference, assign them the available path with the highest score, and mark that path as occupied, until all scheduling instructions have been assigned paths.

[0101] In specific application scenarios, the best-performing path is assigned to the instruction that needs it most, i.e., the scheduling instruction with the largest difference in coupling score. This avoids high-quality paths being occupied by low-dependency instructions, which are paths with a small difference in coupling score between the best and second-best paths, thus preventing resource waste. By allocating high-dependency instructions first and marking their occupancy, critical paths are locked in advance, avoiding subsequent contention by low-dependency instructions and reducing the chain reaction of path conflicts → reallocation → delayed scheduling.

[0102] Example 2, as Figure 2 As shown, a power distribution network optimization dispatching system includes:

[0103] A power-communication coupled topology module is constructed to obtain the power grid physical topology and communication network topology of the distribution network, and to construct a power-communication coupled topology based on the power grid physical topology and communication network topology;

[0104] A power-communication coupled topology model module is constructed to obtain the power operation parameters and communication link parameters of each node in the distribution network, and associate the power operation parameters and communication link parameters as attribute values ​​with the edges of the corresponding nodes in the power-communication coupled topology to obtain the power-communication coupled topology model.

[0105] A network status map construction module is used to construct a network status map based on the power-communication coupled topology model.

[0106] A power supply path subset filtering module is used to obtain a set of power supply paths that can safely carry the transferred load based on the physical topology of the power grid when a node power overload is detected in the distribution network; generate communication constraints based on the network status map; and filter a subset of power supply paths that simultaneously meet the power underload and communication constraints from the power supply path set based on the communication constraints.

[0107] The collaborative scheduling module is used to evaluate the power supply path subset. If there are concurrent instructions within a preset time window, multi-instruction optimized collaborative scheduling is performed.

[0108] The embodiments of the present invention are based on the same inventive concept as Embodiment 1 and have the same technical effects, which will not be repeated here.

[0109] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application 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, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0110] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. 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... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0111] 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.

[0112] 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.

[0113] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A method for optimizing the scheduling of a power distribution network, characterized in that, The method includes: Obtain the power grid physical topology and communication network topology of the distribution network, and construct a power-communication coupled topology based on the power grid physical topology and communication network topology; Obtain the power operation parameters and communication link parameters of each node in the power distribution network, and associate the power operation parameters and communication link parameters as attribute values ​​with the edges of the corresponding nodes in the power-communication coupled topology to obtain the power-communication coupled topology model; Construct a network state map based on the described power-communication coupled topology model; When a node power overload is detected in the distribution network, a set of power supply paths that can safely carry the transferred load is obtained based on the physical topology of the power grid; communication constraints are generated based on the network status map; and a subset of power supply paths that meet the communication constraints are selected from the set of power supply paths based on the communication constraints. The power supply path subset is evaluated, and if concurrent instructions exist within a preset time window, multi-instruction optimized collaborative scheduling is performed.

2. The distribution network optimization scheduling method according to claim 1, characterized in that, Obtain the power grid physical topology and communication network topology of the distribution network, and construct a power-communication coupled topology based on the power grid physical topology and communication network topology, including: Define the power entities in the distribution network as a first node set, obtain the power transmission relationship between each node in the first node set, and construct the power grid physical topology of the distribution network based on the power transmission relationship; Define the communication entities in the power distribution network as a second node set, obtain the communication connection relationships between each node in the second node set, and construct the communication network topology of the power distribution network based on the communication connection relationships; Obtain the mapping relationship between the first node in the first node set and the second node in the second node set, and associate the power grid physical topology and the communication network topology based on the mapping relationship to obtain the power-communication coupled topology.

3. The distribution network optimization scheduling method according to claim 2, characterized in that, The mapping relationship is that the data of the power entities in the first node set is transmitted by the communication entities in the second node set.

4. The distribution network optimization scheduling method according to claim 1, characterized in that, The power operation parameters include: switch position status, node voltage, and distributed resource status; the communication link parameters include: end-to-end latency, packet loss rate, and bandwidth utilization.

5. The distribution network optimization scheduling method according to claim 1, characterized in that, Constructing a network state map based on the power-communication coupled topology model includes: Historical power operation parameters and historical communication link parameters are collected based on the coupling topology relationship in the power-communication coupled topology model. The historical power operation parameters and historical communication link parameters are cleaned and aligned to construct a supervised learning training set; The training set is input into a prediction model built on a long short-term memory network. The prediction model predicts the probability distribution of performance parameters of each link within a preset time window. The probability distribution of performance parameters includes the latency, packet loss rate and corresponding variance and confidence interval of the communication link. Based on the probability distribution of each link performance parameter output by the prediction model, the probability distribution of each link performance parameter is injected as attribute value into the corresponding edge in the power-communication coupled topology model to construct the network state map.

6. The distribution network optimization scheduling method according to claim 1, characterized in that, Based on the network state map, communication constraints are generated, including: For scheduling instructions, extract their communication requirements and generate requirement constraint triples. ,in, This indicates the maximum end-to-end delay that the scheduling instruction can tolerate. This indicates the maximum end-to-end packet loss rate that the scheduling instruction can tolerate. Indicates scheduling instructions The amount of data to be transmitted; From the network status map, obtain the performance prediction values ​​of all communication links within the target time window. The performance prediction values ​​include the expected value of latency prediction, the expected value of packet loss rate prediction, and the expected value of available bandwidth prediction. For each power supply path in the power supply path set for which the scheduling instruction is given, hard constraints are applied. The paths in the power supply path set that satisfy all the hard constraints are collectively denoted as the feasible path set. According to each scheduling instruction and its corresponding set of feasible paths Obtain the communication constraints.

7. The distribution network optimization scheduling method according to claim 6, characterized in that, The hard constraints include bandwidth constraints, latency constraints, and packet loss rate constraints, wherein: The bandwidth constraint includes: the expected available bandwidth of all links on the power supply path is greater than the minimum bandwidth requirement of the scheduling instruction. The minimum bandwidth requirement of the scheduling instruction is determined as follows: the bandwidth requirement obtained by dividing the data volume of the scheduling instruction by the allowed transmission time. The delay constraint includes: determining whether the expected value of the predicted power supply path delay is less than or equal to the maximum end-to-end delay that the scheduling command can tolerate; The packet loss rate constraint includes: determining whether the predicted expected value of the packet loss rate of the power supply path is less than or equal to the maximum end-to-end packet loss rate that the scheduling instruction can tolerate.

8. The distribution network optimization scheduling method according to claim 1, characterized in that, The subset of power supply paths is evaluated. If concurrent instructions exist within a preset time window, multi-instruction optimized collaborative scheduling is performed, including: The power supply path subset of the dispatch command is evaluated using the following formula: , , , in, Indicates coupling score; This represents the local power performance coefficient of link ij. This represents the communication performance coefficient of link ij; This represents the average input active power of link ij; This represents the average output active power of link ij; This represents the average absolute deviation of the voltage of link ij relative to its rated value. This represents the maximum permissible voltage deviation, and the two together constitute the voltage quality coefficient. This represents the average round-trip delay of link ij. Indicates the maximum allowed delay. represents the average packet loss rate of link ij; m represents the number of link segments contained in the path of the power supply path subset.

9. A distribution network optimization scheduling method according to claim 8, characterized in that, If concurrent instructions exist within the preset time window, multi-instruction optimized collaborative scheduling is performed, including: For all paths of scheduling instructions within a preset time window, sort them in descending order according to the difference in coupling score between the best path and the second-best path. Prioritize scheduling instructions with the largest coupling score difference, assign them the available path with the highest score, and mark that path as occupied, until all scheduling instructions have been assigned paths.

10. A power distribution network optimization dispatching system, characterized in that, The system includes: A power-communication coupled topology module is constructed to obtain the power grid physical topology and communication network topology of the distribution network, and to construct a power-communication coupled topology based on the power grid physical topology and communication network topology; A power-communication coupled topology model module is constructed to obtain the power operation parameters and communication link parameters of each node in the distribution network, and associate the power operation parameters and communication link parameters as attribute values ​​with the edges of the corresponding nodes in the power-communication coupled topology to obtain the power-communication coupled topology model. A network status map construction module is used to construct a network status map based on the power-communication coupled topology model. A power supply path subset filtering module is used to obtain a set of power supply paths that can safely carry the transferred load based on the physical topology of the power grid when a node power overload is detected in the distribution network; generate communication constraints based on the network status map; and filter a subset of power supply paths that simultaneously meet the power underload and communication constraints from the power supply path set based on the communication constraints. The collaborative scheduling module is used to evaluate the power supply path subset. If there are concurrent instructions within a preset time window, multi-instruction optimized collaborative scheduling is performed.

Images