Distribution network spatial resource allocation methods, devices and computer equipment

By dividing the initial cells in the distribution network based on the electrical topology and constructing an optimization model, and combining multi-dimensional time series data to drive load forecasting, the problem of low resource allocation efficiency in the distribution network is solved, and refined zoning and efficient resource utilization are achieved.

CN123121641APending Publication Date: 2026-11-17CHINA SOUTHERN POWER GRID ARTIFICIAL INTELLIGENCE TECHNOLOGY CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611587689.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-10-14
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing technologies cannot monitor and adjust power resource allocation in real time, resulting in low power resource allocation efficiency and an inability to adapt to the complex distribution network topology after the large-scale access of diverse loads such as distributed power sources and electric vehicles.

Method used

The initial cells are determined based on the electrical topology of the distribution network. A cell optimization model is constructed to maximize the tightness of electrical connections and minimize the difference between user load and power supply curves. A power load prediction model is driven by multi-dimensional time series data to obtain predicted load sequence data and allocate spatial resources according to electrical capacity constraints.

Benefits of technology

It enables refined zoning of the distribution network spatial structure, improves the rationality of network topology and operational coordination, enhances resource utilization efficiency and power resource allocation efficiency, and provides reliable data support.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123121641A_ABST
    Figure CN123121641A_ABST
Patent Text Reader

Abstract

This application relates to a method, apparatus, and computer equipment for spatial resource allocation in a power distribution network. It pertains to the field of power resource allocation technology. The method includes: determining multiple initial cells of the power distribution network based on its electrical topology; maximizing the electrical connection density within the initial cells, minimizing the difference between user load and power supply curves, and constructing a cell optimization model in conjunction with multiple preset constraints to perform cell partitioning on the power distribution network, thereby obtaining multiple current cells; acquiring multi-dimensional time series data of each current cell over a preset historical time period, and inputting this data into a pre-constructed power load prediction model to obtain predicted load sequence data for each current cell within a preset future time period, thereby determining the matching status between each current cell and preset electrical capacity constraints, and determining a spatial resource allocation suitable for the power distribution network. This method can improve the efficiency of power resource allocation.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application relates to the field of power resource allocation technology, and in particular to a method, apparatus and computer equipment for allocating spatial resources in a power distribution network. Background Technology

[0002] As the final user-facing link in the power system, the distribution network's operational efficiency and the rationality of resource allocation directly affect power supply reliability and power quality. However, with the advancement of new power system construction, a large number of diverse loads such as distributed power sources and electric vehicles are being connected, making the distribution network's topology and operation increasingly complex. Therefore, the distribution network currently needs to possess high-level spatial resource planning and operation management capabilities.

[0003] However, the power resource allocation methods in related technologies define the operating condition stability indicators on both the supply and demand sides of the power grid, thereby setting a monitoring cycle and analyzing the allocation effect of the initial power allocation strategy in multiple dimensions. The power allocation strategy in the next monitoring cycle is then adjusted in combination with the evaluation indicators to optimize the power resource allocation of the power grid system. However, the above methods cannot monitor and adjust the allocation of power resources in real time, resulting in the technical problem of low power resource allocation efficiency. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, apparatus, computer equipment, computer-readable storage medium, and computer program product for allocating spatial resources in a power distribution network, addressing the aforementioned technical problem of low efficiency in power resource allocation.

[0005] Firstly, this application provides a method for allocating spatial resources in a power distribution network, including:

[0006] The electrical topology of the distribution network is used to determine multiple initial cells of the distribution network;

[0007] To maximize the electrical connection tightness within the multiple initial cells, minimize the difference between user load and power curves within the multiple initial cells, and combine multiple preset constraints, construct cell optimization models corresponding to the multiple initial cells.

[0008] Based on the cell optimization model, the distribution network is divided into cells to obtain multiple current cells of the distribution network;

[0009] The multi-dimensional time series data of each current cell in a preset historical time period are obtained, and the multi-dimensional time series data are input into a pre-constructed power load prediction model to obtain the predicted load sequence data of each current cell in a future preset time period.

[0010] Based on the predicted load sequence data corresponding to each current cell, the matching status between each current cell and the preset electrical capacity constraint is determined, and the spatial resource allocation adapted to the distribution network is determined based on the matching status.

[0011] In one embodiment, the step of inputting the multi-dimensional time series data into a pre-built power load prediction model to obtain the predicted load sequence data for each current cell within a preset future time period includes:

[0012] The multi-dimensional time series data is input into a pre-built power load forecasting model to obtain the predicted load sequence data for each current cell in the future preset time period, including short-term predicted load sequence data, medium-term predicted load sequence data and long-term predicted load sequence data.

[0013] Wherein, the duration span between the data tail of the short-term predicted load sequence data and the corresponding data tail of the multi-dimensional time series data is less than the duration span between the data tail of the medium-term predicted load sequence data and the corresponding data tail of the multi-dimensional time series data, and the duration span between the data tail of the medium-term predicted load sequence data and the corresponding data tail of the multi-dimensional time series data is less than the duration span between the data tail of the long-term predicted load sequence data and the corresponding data tail of the multi-dimensional time series data.

[0014] In one embodiment, obtaining multi-dimensional time series data of each current cell over a preset historical time period includes:

[0015] Based on the load time series data, power generation time series data, meteorological time series data and economic time series data of each current cell in the preset historical time period, multi-dimensional time series data of each current cell in the preset historical time period are obtained.

[0016] In one embodiment, determining the matching status between each current cell and a preset electrical capacity constraint based on the predicted load sequence data corresponding to each current cell, and determining the spatial resource allocation suitable for the distribution network based on the matching status, includes:

[0017] Based on the predicted load sequence data corresponding to each current cell, determine the maximum total load of the current cell;

[0018] The matching status between the maximum total load of the current cell and the preset electrical capacity constraint is determined. If the matching status is that the maximum total load of the current cell exceeds the threshold of the preset electrical capacity constraint, the spatial resource allocation suitable for the distribution network is determined by marking the current cell as a node to be optimized and determining the highest-level power supply node corresponding to the node to be optimized, so as to allocate the spatial resources of the distribution network based on the power supply range and capacity characteristics of the highest-level power supply node.

[0019] In one embodiment, the process of maximizing the electrical connection tightness within the plurality of initial cells, minimizing the difference between user load and power curves within the plurality of initial cells, and constructing a cell optimization model corresponding to the plurality of initial cells in combination with multiple preset constraints includes:

[0020] To maximize the electrical connection tightness within the multiple initial cells, minimize the difference between user load and power curves within the multiple initial cells, and combine preset constraints including allocation integrity constraints, spatial continuity constraints, scale appropriateness constraints, electrical capacity constraints, topology consistency constraints, and cell number constraints, an optimization model corresponding to the multiple initial cells is constructed.

[0021] In one embodiment, the distribution network-based electrical topology diagram determines a plurality of initial cells of the distribution network, including:

[0022] Based on the electrical topology diagram of the distribution network, the smallest divisible units belonging to the same upstream transformer, line or switch are divided into the same initial cell, and multiple initial cells are determined in combination with the geographical area corresponding to the distribution network; the smallest divisible unit is a meter or a building unit.

[0023] Secondly, this application also provides a distribution network spatial resource allocation device, comprising:

[0024] A cell determination module is used to determine multiple initial cells of the distribution network based on the electrical topology diagram of the distribution network;

[0025] The model building module is used to maximize the electrical connection tightness within the multiple initial cells, minimize the difference between user load and power curves within the multiple initial cells, and construct cell optimization models corresponding to the multiple initial cells in combination with multiple preset constraints.

[0026] The cell processing module is used to perform cell partitioning processing on the distribution network based on the cell optimization model to obtain multiple current cells of the distribution network.

[0027] The cell processing module is also used to acquire multi-dimensional time series data of each current cell in a preset historical time period, and input the multi-dimensional time series data into a pre-built power load prediction model to obtain the predicted load sequence data of each current cell in a future preset time period.

[0028] The scheme determination module is used to determine the matching status between each current cell and the preset electrical capacity constraint based on the predicted load sequence data corresponding to each current cell, and to determine the spatial resource allocation suitable for the distribution network based on the matching status.

[0029] Thirdly, this application also provides a computer device, the computer device including a memory and a processor, the memory storing a computer program, and the processor executing the computer program to perform the following steps:

[0030] The electrical topology of the distribution network is used to determine multiple initial cells of the distribution network;

[0031] To maximize the electrical connection tightness within the multiple initial cells, minimize the difference between user load and power curves within the multiple initial cells, and combine multiple preset constraints, construct cell optimization models corresponding to the multiple initial cells.

[0032] Based on the cell optimization model, the distribution network is divided into cells to obtain multiple current cells of the distribution network;

[0033] The multi-dimensional time series data of each current cell in a preset historical time period are obtained, and the multi-dimensional time series data are input into a pre-constructed power load prediction model to obtain the predicted load sequence data of each current cell in a future preset time period.

[0034] Based on the predicted load sequence data corresponding to each current cell, the matching status between each current cell and the preset electrical capacity constraint is determined, and the spatial resource allocation adapted to the distribution network is determined based on the matching status.

[0035] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0036] The electrical topology of the distribution network is used to determine multiple initial cells of the distribution network;

[0037] To maximize the electrical connection tightness within the multiple initial cells, minimize the difference between user load and power curves within the multiple initial cells, and combine multiple preset constraints, construct cell optimization models corresponding to the multiple initial cells.

[0038] Based on the cell optimization model, the distribution network is divided into cells to obtain multiple current cells of the distribution network;

[0039] The multi-dimensional time series data of each current cell in a preset historical time period are obtained, and the multi-dimensional time series data are input into a pre-constructed power load prediction model to obtain the predicted load sequence data of each current cell in a future preset time period.

[0040] Based on the predicted load sequence data corresponding to each current cell, the matching status between each current cell and the preset electrical capacity constraint is determined, and the spatial resource allocation adapted to the distribution network is determined based on the matching status.

[0041] Fifthly, this application also provides a computer program product, which includes a computer program that, when executed by a processor, performs the following steps:

[0042] The electrical topology of the distribution network is used to determine multiple initial cells of the distribution network;

[0043] To maximize the electrical connection tightness within the multiple initial cells, minimize the difference between user load and power curves within the multiple initial cells, and combine multiple preset constraints, construct cell optimization models corresponding to the multiple initial cells.

[0044] Based on the cell optimization model, the distribution network is divided into cells to obtain multiple current cells of the distribution network;

[0045] The multi-dimensional time series data of each current cell in a preset historical time period are obtained, and the multi-dimensional time series data are input into a pre-constructed power load prediction model to obtain the predicted load sequence data of each current cell in a future preset time period.

[0046] Based on the predicted load sequence data corresponding to each current cell, the matching status between each current cell and the preset electrical capacity constraint is determined, and the spatial resource allocation adapted to the distribution network is determined based on the matching status.

[0047] The aforementioned distribution network spatial resource allocation method, device, computer equipment, storage medium, and computer program product, in the process of distribution network spatial resource allocation, determine multiple initial cells of the distribution network based on the electrical topology diagram of the distribution network; maximize the electrical connection tightness within the multiple initial cells, minimize the difference between user load and power supply curves within the multiple initial cells, and construct cell optimization models corresponding to the multiple initial cells in combination with multiple preset constraints; perform cell partitioning processing on the distribution network based on the cell optimization models to obtain multiple current cells of the distribution network; acquire multi-dimensional time series data of each current cell in a preset historical time period, and input the multi-dimensional time series data into a pre-constructed power load prediction model to obtain the predicted load sequence data of each current cell in the future preset time period; based on the predicted load sequence data corresponding to each current cell, determine the matching status between each current cell and the preset electrical capacity constraints, and determine the spatial resource allocation suitable for the distribution network based on the matching status. By dividing the distribution network electrical topology into initial cells and constructing an optimization model with the goals of maximizing the tightness of internal electrical connections and minimizing the difference between load and power supply curves, the spatial structure of the distribution network is refined and modularized, effectively improving the rationality of network topology and operational coordination. By using multi-dimensional historical time series data to drive the power load forecasting model, the future load sequence of each cell is accurately obtained, providing reliable data support for grid operation. Furthermore, by combining electrical capacity constraints to determine the resource matching status of each cell and formulating spatial resource allocation, the resource utilization efficiency and power resource allocation efficiency of the distribution network are significantly improved. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1 This is a flowchart illustrating a distribution network spatial resource allocation method in one embodiment;

[0050] Figure 2 This is a flowchart illustrating the steps of allocating spatial resources in a power distribution network in one embodiment.

[0051] Figure 3 This is a flowchart illustrating the distribution network spatial resource allocation method in another embodiment;

[0052] Figure 4 This is a structural block diagram of a power distribution network spatial resource allocation device in one embodiment;

[0053] Figure 5 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0055] As the final user-facing link in the power system, the distribution network's operational efficiency and resource allocation rationality directly affect power supply reliability and power quality. With the advancement of new power system construction, the large-scale integration of diverse loads such as distributed power sources and electric vehicles has led to increasingly complex distribution network topologies and operating modes. The power flow has shifted from the traditional unidirectional radial pattern to multidirectional interaction, placing higher demands on the spatial resource planning and operation management of the distribution network. Related technologies involve training predictive models on both the supply and demand sides of the power grid. Specifically, this training considers the time lags on both sides and analyzes the complex coupling relationships between them, providing a reliable foundation for subsequent power resource allocation strategies and effectiveness evaluation. Specifically, by defining operational stability indicators on both the supply and demand sides of the power grid, a monitoring cycle is set, and the allocation effect of the initial power allocation strategy is analyzed from multiple dimensions. Combined with evaluation indicators, the power allocation strategy for the next monitoring cycle is adjusted. This comprehensive evaluation of the power allocation strategy from multiple dimensions optimizes the power resource allocation of the power grid system. However, it cannot achieve real-time monitoring and timely adjustment of power resources, thus affecting the efficiency of power resource allocation.

[0056] To solve the above-mentioned technical problems, in an exemplary embodiment, such as Figure 1 As shown, a method for allocating spatial resources in a distribution network is provided. Taking the application of this method to a terminal as an example, the method includes the following steps S102 to S110. Wherein:

[0057] Step S102: Determine multiple initial cells of the distribution network based on the electrical topology diagram of the distribution network.

[0058] Among them, the distribution network electrical topology diagram is a schematic diagram of the connection relationship of all electrical equipment in the distribution network, clearly showing the electrical connection path, hierarchical structure and topological relationship between equipment; the initial cell is the basic unit of the distribution network obtained by preliminary partitioning based on the distribution network electrical topology diagram. The partitioning of the initial cell can be combined with preliminary conditions such as the power grid topology, power supply area, and equipment distribution.

[0059] Step S104: Maximize the electrical connection tightness within multiple initial cells, minimize the difference between user load and power curves within multiple initial cells, and construct cell optimization models corresponding to multiple initial cells by combining multiple preset constraints.

[0060] Among them, electrical connection tightness refers to the degree of electrical connection between electrical nodes within a cell, which is usually measured by indicators such as electrical distance between nodes, power transmission efficiency, and line connection density; user load and power supply curve difference refers to the degree of deviation between the curve of user power load changing over time and the curve of power output changing over time within a cell; preset constraints are pre-set restrictions to ensure that the cell division is scientific and reasonable and meets the requirements of distribution network operation; the cell optimization model is a mathematical model constructed by combining preset constraints with the dual optimization objectives of "maximizing electrical connection tightness and minimizing source-load curve difference", which can be used to adjust and optimize the initial cells.

[0061] Step S106: Based on the cell optimization model, the distribution network is divided into cells to obtain multiple current cells of the distribution network.

[0062] Among them, the cell partitioning process is based on the constructed cell optimization model, and the process of adjusting, merging or splitting the boundary, range and internal composition of the initial cell through the solution algorithm; the current cell is the final partition unit of the distribution network obtained after solving the cell optimization model and completing the partitioning process, and the current cell has met the requirements of tight electrical connection, source-load curve matching and conformity to preset constraints.

[0063] Step S108: Obtain multi-dimensional time series data of each current cell in a preset historical time period, and input the multi-dimensional time series data into a pre-built power load prediction model to obtain the predicted load sequence data of each current cell in the future preset time period.

[0064] Among them, multi-dimensional time series data are various types of continuous time-related data reflecting the operating status of each current cell within a preset historical time period. Core dimensions may include load data, power supply data, power grid operation data, environmental data, etc. The preset historical time period is a time range set in advance to obtain effective historical data and ensure prediction accuracy. It is usually selected from time periods similar to the operating conditions of the future prediction period. The power load prediction model is a pre-built intelligent model used to predict future loads. The predicted load sequence data is the set of predicted load values ​​for each current cell at each time node within the preset future time period, output by the power load prediction model. The preset future time period is the future time range for which load prediction is required and can be set according to the needs of distribution network dispatching.

[0065] Step S110: Based on the predicted load sequence data corresponding to each current cell, determine the matching status between each current cell and the preset electrical capacity constraints, and determine the spatial resource allocation suitable for the distribution network based on the matching status.

[0066] Among them, the preset electrical capacity constraint is the maximum electrical load capacity that each cell can carry in advance to ensure the safe and stable operation of the distribution network. It is the core standard for judging the load matching degree of the cell. The matching status is the relationship between the predicted load sequence data of the current cell and the preset electrical capacity constraint. It can be divided into three situations: the predicted load is continuously lower than the capacity constraint, the predicted load is within the capacity constraint range, and the predicted load is continuously higher than the capacity constraint. The spatial resource allocation is a distribution network spatial resource adjustment and allocation scheme formulated based on the matching status of each cell and the electrical capacity constraint.

[0067] In the aforementioned method for allocating spatial resources in a distribution network, multiple initial cells of the distribution network are determined based on the electrical topology diagram of the distribution network. The electrical connection density within these initial cells is maximized, the difference between user load and power supply curves within these initial cells is minimized, and multiple preset constraints are combined to construct a cell optimization model corresponding to each initial cell. Based on the cell optimization model, the distribution network is divided into cells to obtain multiple current cells. Multi-dimensional time series data of each current cell over a preset historical time period are obtained, and this multi-dimensional time series data is input into a pre-constructed power load prediction model to obtain the predicted load sequence data for each current cell within a preset future time period. Based on the predicted load sequence data corresponding to each current cell, the matching status between each current cell and the preset electrical capacity constraints is determined, and based on the matching status, a spatial resource allocation suitable for the distribution network is determined. By dividing the distribution network electrical topology into initial cells and constructing an optimization model with the goals of maximizing the tightness of internal electrical connections and minimizing the difference between load and power supply curves, the spatial structure of the distribution network is refined and modularized, effectively improving the rationality of network topology and operational coordination. By using multi-dimensional historical time series data to drive the power load forecasting model, the future load sequence of each cell is accurately obtained, providing reliable data support for grid operation. Furthermore, by combining electrical capacity constraints to determine the resource matching status of each cell and formulating spatial resource allocation, the resource utilization efficiency and power resource allocation efficiency of the distribution network are significantly improved.

[0068] In an exemplary embodiment, multi-dimensional time series data is input into a pre-built power load forecasting model to obtain the predicted load sequence data for each current cell within a preset future time period. This includes: inputting multi-dimensional time series data into a pre-built power load forecasting model to obtain the predicted load sequence data for each current cell within a preset future time period, including short-term predicted load sequence data, medium-term predicted load sequence data, and long-term predicted load sequence data; wherein the duration span between the data tail of the short-term predicted load sequence data and the data tail of the corresponding multi-dimensional time series data is less than the duration span between the data tail of the medium-term predicted load sequence data and the data tail of the corresponding multi-dimensional time series data, and the duration span between the data tail of the medium-term predicted load sequence data and the data tail of the corresponding multi-dimensional time series data is less than the duration span between the data tail of the long-term predicted load sequence data and the data tail of the corresponding multi-dimensional time series data.

[0069] Among them, short-term forecast load sequence data is sequence data formed by predicting the load in a relatively short period of time in the future, starting from the end of the historical multi-dimensional time series data of the current cell. It has the smallest forecast duration and the highest requirements for data real-time performance and accuracy. Medium-term forecast load sequence data is sequence data obtained by predicting load changes in a medium-term time span in the future based on the end of historical data. It is often used for weekly and monthly distribution network operation planning. Its duration is longer than short-term but shorter than long-term. Long-term forecast load sequence data is sequence data generated by forecasting load for a longer period of time in the future. It has the largest duration. The data tail refers to the latest and closest end data position in the time dimension of the time series data. It is the benchmark reference point for dividing the duration of short-term, medium-term and long-term forecasts. The duration span is the length of time from the end of the historical multi-dimensional time series data to the end of the corresponding forecast load sequence data. It is used to quantitatively distinguish the time range differences between short-term, medium-term and long-term forecasts.

[0070] In this embodiment, the power load forecasting model outputs short-, medium-, and long-term load sequence data. By combining the differences in time span, it achieves layered coverage of load forecasting, adapting to the needs of different distribution network scenarios. Short-term forecasts can support intraday and day-ahead fine-grained scheduling, ensuring real-time power supply stability. Medium-term forecasts provide a basis for weekly and monthly operation plans. Long-term forecasts assist in distribution network planning and resource allocation. At the same time, the time span is divided based on the tail end of historical multi-dimensional time series data to improve forecast accuracy, provide comprehensive data support for distribution network cell resource allocation, and enhance the scientific nature of distribution network operation.

[0071] In an exemplary embodiment, the above-mentioned step of obtaining multi-dimensional time series data of each current cell in a preset historical time period includes: obtaining multi-dimensional time series data of each current cell in a preset historical time period based on load time series data, power generation time series data, meteorological time series data and economic time series data of each current cell in the preset historical time period.

[0072] The specific start and end times corresponding to the preset historical time period can be selected based on needs, and this application does not impose any limitations on this. In the embodiments provided in this application, the "multi-dimensional" in "multi-dimensional time series data" can involve at least four dimensions: load, power generation, meteorology, and economy, but other related data can also be added based on needs to improve the richness of the multi-dimensional time series data.

[0073] In this embodiment, by acquiring the time series data of the current cell in four dimensions—load, power generation, meteorology, and economy—within a preset historical time period, the amount of multi-dimensional time series data obtained is richer, which helps to improve the accuracy of the predicted load sequence data corresponding to each current cell in the future preset time period.

[0074] In one embodiment, such as Figure 2 As shown, based on the predicted load sequence data corresponding to each current cell, the matching status between each current cell and the preset electrical capacity constraint is determined, and the spatial resource allocation suitable for the distribution network is determined based on the matching status, including: step S202, determining the maximum total load of the current cell based on the predicted load sequence data corresponding to each current cell; step S204, determining the matching status between the maximum total load of the current cell and the preset electrical capacity constraint, and if the matching status is that the maximum total load of the current cell exceeds the threshold of the preset electrical capacity constraint, determining the spatial resource allocation suitable for the distribution network is to mark the current cell as a node to be optimized, and determine the highest-level power supply node corresponding to the node to be optimized, so as to allocate the spatial resources of the distribution network based on the power supply range and capacity characteristics of the highest-level power supply node.

[0075] Among them, the maximum total load refers to the peak summation value of all predicted load sequence data of the current cell in the future prediction period; the threshold is a critical value set in the electrical capacity constraint to determine whether the cell has an overload risk. Exceeding the threshold is judged as power supply shortage or overload; the node to be optimized is the current cell whose maximum total load exceeds the capacity constraint threshold and needs to be reallocated power supply resources; the highest-level power supply node is the core node in the distribution network topology with a higher voltage level, stronger power supply capacity, and wider power supply range, such as main transformers, busbars, and switching stations; the power supply range and capacity characteristics are the comprehensive operating attributes of the highest-level power supply node, such as the power supply area that can be stably covered, rated power supply capacity, allowable load rate, and transmission constraints.

[0076] In this embodiment, by calculating the sum of the maximum load of each cell, the matching status with the electrical capacity constraint can be accurately determined, and overloaded cells can be quickly located and marked as nodes to be optimized, thus improving the targeting of resource allocation. Taking the highest-level power supply node as the core and combining its power supply range and capacity characteristics to formulate spatial resource allocation, the supporting capacity of backbone power supply equipment can be fully utilized, the power flow can be optimized, the cell overload problem can be effectively alleviated, the resource utilization rate and power supply reliability of the distribution network can be improved, and the foundation for the safe and stable operation of the distribution network can be laid.

[0077] In one embodiment, maximizing the electrical connection tightness within multiple initial cells, minimizing the difference between user load and power supply curves within multiple initial cells, and combining multiple preset constraints, a cell optimization model corresponding to multiple initial cells is constructed. This includes: maximizing the electrical connection tightness within multiple initial cells, minimizing the difference between user load and power supply curves within multiple initial cells, and combining preset constraints including allocation integrity constraints, spatial continuity constraints, scale appropriateness constraints, electrical capacity constraints, topology consistency constraints, and cell number constraints to construct an optimization model corresponding to multiple initial cells.

[0078] Among these constraints, the integrity constraint ensures that all loads, lines, and equipment within the distribution network are fully covered during cell partitioning, without omissions, gaps, or duplicate assignments; the spatial continuity constraint requires that the partitioned cells remain continuous in geographical space, avoiding regional fragmentation and enclave-like distribution; the scale appropriateness constraint limits the load size, number of devices, and power supply range of a single cell to a reasonable range; the topology consistency constraint requires that cell boundaries be consistent with the actual electrical topology of the distribution network, without disrupting the original line connections and power supply logic, ensuring the normal operation of the power grid; and the cell quantity constraint limits the total number of cells partitioned into the entire distribution network, ensuring a moderate number of partitions for easy scheduling, monitoring, and resource coordination.

[0079] In this embodiment, an optimization model is constructed with the objectives of maximizing the tightness of electrical connections and minimizing the difference between source and load curves. Combined with multiple constraints, the cell partitioning can better reflect the actual operating rules of the distribution network. The combined effect of multiple constraints ensures both the continuity and appropriate size of the cell space, as well as the integrity and reasonable number of the power grid topology, effectively improving the rationality of the partitioning. At the same time, by standardizing the cell boundaries and number, it is convenient for subsequent load forecasting and resource scheduling, providing a reliable foundation for the refined and modular management and control of the distribution network.

[0080] In an exemplary embodiment, determining multiple initial cells of the distribution network based on the electrical topology diagram of the distribution network includes: dividing the smallest divisible units belonging to the same upstream transformer, line or switch into the same initial cell based on the electrical topology diagram of the distribution network, and determining multiple initial cells in combination with the geographical area corresponding to the distribution network; the smallest divisible unit is a meter or a building unit.

[0081] Among them, the same upstream transformer, line or switch refers to the same distribution transformer, the same main line or the same control switch that provides power supply support to downstream loads in the distribution network, and is the basis for electrical attribution of the initial cells; the smallest divisible unit is the basic load unit that cannot be further divided in the cell division of the distribution network; the electricity meter is the terminal equipment used to measure the electricity consumption of users; the building unit is an independent building, building or user block as the basic unit, taking into account the main body of electricity consumption and geographical attributes, which facilitates regional management and operation and maintenance; the geographical area is the actual geographical space range corresponding to the distribution network. Combining geographical continuity to divide cells can avoid fragmentation of zones and facilitate on-site operation and maintenance and fault handling.

[0082] In this embodiment, by using electricity meters or building units as the smallest divisible units, initial cells are divided according to the affiliation of the same upstream power supply equipment, and optimized in combination with geographical regions, so that the cells have both electrical correlation and spatial rationality. This ensures that the electrical connection within the cells is close, and facilitates subsequent source-load balance analysis and curve difference calculation. This lays a reliable foundation for cell splitting, load forecasting and resource allocation, and also improves the scientific nature of the distribution network zoning.

[0083] This application provides a method for allocating spatial resources in a distribution network. To better understand the process of the above-mentioned method for allocating spatial resources in a distribution network, combined with... Figure 3 As shown below, the specific process of a distribution network spatial resource allocation method according to this application is described in detail, including the following steps:

[0084] Step S302: Divide the distribution network into different cells based on the electrical topology diagram, abstracting the power grid as an undirected graph. Step S304: Integrate multiple practical constraints to form a complete cell optimization model, and use this model to complete the distribution network cell partitioning. Step S306: Input a series of time-series data from the cells and import them into the power load forecasting model, outputting predicted load sequence data. Step S308: Execute different resource allocation schemes based on the predicted load sequence data results, completing the identification of nodes for optimal allocation of spatial resources in the distribution network.

[0085] Specifically, steps S302-S308 can be executed as follows: Based on the electrical topology diagram, the distribution network is divided into different cells, the power grid is abstracted into an undirected graph, the electrical connection density within the cells is maximized and the differences between user load or power supply curves within the cells are minimized, these objectives are integrated with multiple practical constraints such as spatial continuity, appropriate scale, and electrical capacity to form a complete optimization model, and the distribution network cells are divided through the model; a series of time series data of the cells are input and imported into the power load prediction model, and the predicted load series data of the future short-term, medium-term and long-term for each cell are output; the predicted load series data are used to determine whether the cells meet the new electrical capacity constraints, and different schemes are executed based on the judgment results to complete the identification of nodes to be optimized in the spatial resource allocation of the distribution network.

[0086] For example, the above step of "dividing the distribution network into different cells based on the electrical topology diagram" can be specifically executed as follows: based on the electrical topology diagram, units belonging to the same upstream transformer, line or switch are divided into the same cell, and the geographical area covered by the distribution network is initially divided into K cells {C1,C2,...,Ck}. A cell refers to the smallest divisible unit, such as a single meter or a building unit.

[0087] For example, the above step of "abstracting the power grid into an undirected graph" can be specifically executed as follows: the distribution network is abstracted into an undirected graph G=(V,E,W), where node V is the smallest divisible unit, and edge E is where two units are directly connected on the physical power grid and belong to the same upstream transformer / line / switch. There is an edge between them, and the edge weight W is defined as the reciprocal of the electrical distance or the reciprocal of the impedance. The smaller the impedance, the stronger the electrical correlation, and the larger the weight.

[0088] For example, the steps described above, which involve "maximizing the tightness of electrical connections within a cell and minimizing the differences between user loads or power supply curves within a cell, integrating these objectives with multiple practical constraints such as spatial continuity, appropriate scale, and electrical capacity to form a complete optimization model, and using the model to complete the cell partitioning of the distribution network," can be specifically executed as follows:

[0089] The quantization function is maximized as the objective function for electrical correlation (tightness of electrical connection): F_electric=(1 / (2m))*ΣΣ[w(i,j)-(k_i*k_j) / (2m)]*δ(L(i),L(j)), where F_electric=Q modularity, which measures electrical correlation.

[0090] In the electrical correlation objective function, m is the sum of the weights of all edges, 1 / (2m) is the normalization factor, and the summation result is divided by the total weight of the network, 2m, to make the final F_electric a standardized index, normalized to [-1,1]. This allows for comparison of the partitioning results between distribution networks of different sizes and connection densities. w(i,j) is the edge weight, the weight of the connection between node i and node j, k_i is the sum of the weights of all edges connected to node i, and k_j is the sum of the weights of all edges connected to node j. w(i,j)-(k_ The term `i*k_j) / (2m)` represents the actual connection strength minus the expected strength under random conditions. If the result is greater than 0, it means that the actual connection between nodes i and j is tighter than under random conditions, and they should be grouped into the same cell. If the result is less than 0, it means that the actual connection between nodes i and j is weaker than under random conditions, and they should not be grouped into the same cell. `δ(L(i),L(j))` is the Kronecker delta function. If nodes i and j belong to the same cell, then `δ(L(i),L(j))` is 1; otherwise, it is 0. This ensures that the summation symbol `ΣΣ` is only applied to node pairs belonging to the same cell. Node pairs from different cells do not contribute to the modularity calculation.

[0091] The quantization function is minimized as the objective function for load / power characteristic homogeneity (differences between user load or power curves), F_homo = Σ(for all cells Ck)[Σ(for all cells i,j within a cell)Distance(T_i,T_j)]. The average profile coefficient measures homogeneity. Here, T_i and T_j are the time-series feature vectors of the units, representing the historical behavior data of each smallest unit. Typically, this is a historical daily load curve, for example, active power data at 96 points (15-minute intervals) per day for the past 30 days, which is a vector with dimensions (30*96). Distance(T_i, T_j) is a difference measure function between two time series, used to quantify the dissimilarity of the behavioral patterns of the two units. The larger the Distance value, the greater the behavioral difference between unit i and unit j. For example, DTW Distance (Dynamic Time Warping Distance) can be used to handle curves of different shapes but similar trends. DTW can find the optimal alignment between two time series and then calculate the distance.

[0092] Σ(for all units i,j in a cell) Distance(T_i,T_j) is the inner summation, which calculates the total difference between all pairs of units in a specific cell C_k. It iterates through each pair of units (i,j) in cell C_k, calculates their distances and sums them up. If there are n units in a cell, then n*(n-1) / 2 distances need to be calculated and summed. This value measures the total amount of "disharmony" in cell C_k. The larger the value, the more diverse the user behaviors in this cell are and the worse the homogeneity. Σ(for all cells C_k)[] is the outer summation, which sums up the internal differences of all cells to obtain the global total difference F_homo of the entire partitioning scheme. It is an overall index that evaluates the homogeneity of the entire cell partitioning scheme. The optimization goal is to minimize this F_homo value.

[0093] Furthermore, the electrical relevance objective and the load / power supply characteristic homogeneity objective are integrated and constraints are added to obtain a complete cellular optimization model: Maximize{F=ω1*F_electric+ω2*F_homo} or Minimize{F=ω1*(-F_electric)+ω2*F_homo}. Here, ω1 and ω2 are dynamic adaptive weights used to adjust the relative importance of the two objectives. The constraints specifically involve allocation integrity constraints, spatial continuity constraints, scale appropriateness constraints, electrical capacity constraints, topology consistency constraints, and cell number constraints.

[0094] Allocation integrity constraints: Where i∈{1,2,…,N}, and N is the number of indivisible smallest units in the entire distribution network. i and j are indices of the smallest units, used to traverse and identify the subscript variables of each smallest unit. k is the index of the cell, used to traverse and identify the subscript variables of each cell, k∈{1,2,3,...,K}, and K is the total number of cells finally divided. It is the total number of cells into which the entire region is divided after the optimization algorithm is executed. K is not a pre-set fixed value, but the result after algorithm optimization. The value range is controlled by the constraints. It ensures that each smallest unit must be assigned to one cell and can only be assigned to one cell to avoid unit omission or duplicate calculation. During the algorithm iteration process, when adjusting the affiliation of a unit, it must be removed from the original cell and added to the new cell at the same time.

[0095] Spatial continuity constraints: Where k∈{1,2,3,...,K}, Connectivity(Ck) is a function that returns 1 when all cells in a cell form a continuous region in space, otherwise returns 0 to prevent the creation of "enclaves" and ensure that each cell is a continuous geographical region.

[0096] Size Appropriateness Constraints: ;

[0097] in, This represents the summation of the size metrics of all cells in cell k. `x_ik` is an indicator variable, either 0 or 1. If `x_ik = 1`, it means cell i belongs to cell k; if `x_ik = 0`, it means cell i does not belong to cell k. If cell i is in cell k, the product equals `Size_i`; if cell i is not in cell k, the product equals 0. `Size_i` is the size metric of cell i, representing the power consumption or importance of each smallest cell. The calculation method can be either `Size_i = P_i,max` (the historical maximum load of cell i) or `Size_i = P_i,avg` (the historical average load of cell i). S_min = α * AvgLoad, where α is the lower limit coefficient, usually taken as 0.1-0.3. The minimum allowable cell size is 10%-30% of the average size. AvgLoad is the average cell load of the system = (total system load) / (expected number of cells). If the total system load is 100MW and it is expected to be divided into 200 cells, then AvgLoad = 100MW / 200 = 0.5MW. S_max = β * S_transformer, where β is the safety factor, typically taken as 0.6-0.8, providing a 20%-40% safety margin for the transformer / line / switch. S_transformer is the rated capacity of the upstream transformer / line / switch. Let N be the number of users in cell k, Nmin be the minimum number of cells (5-10), and Nmax be the maximum number of cells (100-200). The lower bound ensures that cells have sufficient statistical samples for reliable prediction, avoiding unstable prediction results due to too few cells, while the upper bound prevents individual cells from being too large, affecting local prediction accuracy, and ensures that the impact of faults is controllable.

[0098] Electrical capacity constraints: ;in, For the historical maximum load of unit i, The rated capacity of the line or switch supplying power to cell k, where γ is a safety factor, typically 0.8-0.9, to ensure that the maximum possible load / total power supply of all units within the cell does not exceed the capacity limit of the line or equipment supplying power to that cell;

[0099] Topological consistency constraints: If there is a normally open switch or a disconnect point between two units, they cannot be assigned to the same cell to ensure that the cell division is consistent with the actual operation of the power grid. The specific implementation method is to obtain the switch status from the SCADA (Supervisory Control and Data Acquisition) system and preprocess a list of "prohibited merging" cell pairs before division.

[0100] Cell number constraint: Among them, K min To ensure a minimum number of cells for spatially differentiated predictions, K... max To minimize the number of cells allowed, and to avoid excessive computational complexity due to an excessive number of cells;

[0101] Furthermore, iterate through all cells. If a cell has an extremely poor F_homo value (exceeding the preset value), that is, a huge difference in the internal unit curves, it is marked as a cell to be split even if its electrical connections are tight. Temporal clustering algorithms, such as K-MeanswithDTW, are used to cluster the units within this cell, splitting it into multiple more homogeneous sub-cells. Size constraints are also applied to ensure that the sub-cells are not too small. The internal unit curve difference refers to the degree of deviation between the time series curves of each sub-unit within a single current cell. The preset value is a pre-defined threshold for judging the internal unit curve difference, determined by the distribution network operation requirements and cell optimization objectives; it is the core criterion for determining whether the current cell needs to be split. Cells to be split are those whose internal unit curve differences exceed the preset value and do not meet the cell optimization requirements after traversal and judgment. Temporal clustering algorithms are clustering analysis algorithms for time series data. Their core function is to group similar sequence data into one category based on the time series' changing trends, fluctuation characteristics, and numerical patterns. Splitting is the process of using a temporal clustering algorithm to divide the cell to be split into multiple new cells.

[0102] Furthermore, after cell splitting is complete, all adjacent cell pairs are traversed. If two cells have very low F_electric values, meaning their electrical connection is weak, but their time-series curves F_homo are highly similar (e.g., two adjacent residential areas powered by the same line), they are considered to be merged into one cell to simplify the model and impose size constraints to ensure that the merged cell is not too large. For cells located on cell boundaries, the impact on the overall objective function F if they are assigned to adjacent cells is calculated. If F increases, their assignment is changed. When the growth of the objective function F falls below a certain threshold or the maximum number of iterations is reached, the algorithm stops, and the distribution network cell partitioning is complete.

[0103] In this embodiment, by traversing the current cell, marking and splitting cells whose internal unit curve differences exceed the standard, the internal structure of the cell can be optimized, the operational differences between units can be reduced, the source-load coordination within the cell can be improved, and load prediction deviations caused by imbalances within the cell can be avoided. By combining the splitting of cells with the time-series clustering algorithm, it is ensured that the new cells after splitting meet the optimization requirements, improving the relevance and accuracy of the prediction data, and providing reliable data support for the allocation of spatial resources and operation scheduling of the distribution network.

[0104] For example, the steps described above, namely "inputting a series of time-series data of the input cell into the power load forecasting model, outputting the predicted load sequence data of each cell in the short, medium and long term, determining whether the cell meets the new electrical capacity constraints based on the predicted load sequence data, and executing different schemes based on the determination results to complete the identification of nodes to be optimized in the distribution network spatial resource allocation", can be specifically executed as follows:

[0105] Each cell is input with its historical load, power generation, meteorological, and economic time-series data, which are imported into the power load forecasting model. The model outputs the predicted load sequences for the short, medium, and long term for each cell. It then determines whether the predicted load sequences for each cell meet the new electrical capacity constraints. If any do not meet the new electrical capacity constraints, the cell node is marked as needing optimization. The new electrical capacity constraints are: ,in, For the historical maximum load of unit i, The rated capacity of the line or switch supplying power to cell k; for cells that meet the new electrical capacity constraints, find the upper-level line / transformer / switch according to the electrical topology diagram, integrate all cell data belonging to the upper-level node to determine whether the new electrical capacity constraints of the upper-level node are met, if there are any that do not meet the new electrical capacity constraints, mark the upper-level node as to be optimized, repeat until the highest node is reached, the algorithm stops, and the identification of nodes to be optimized for distribution network space resource allocation is completed.

[0106] In the above embodiments, the distribution network is divided into different cells based on the electrical topology diagram, and the power grid is abstracted as an undirected graph. This maximizes the electrical connectivity within each cell and minimizes the differences between user load or power supply curves within the cell. These objectives are integrated with multiple practical constraints such as spatial continuity, appropriate scale, and electrical capacity to form a complete optimization model. By using this model to divide the distribution network into cells, the dual objectives of electrical connectivity and load homogeneity can be dynamically optimized. The resulting cells not only have tight electrical connections but also highly similar user behavior patterns, making the analysis, prediction, and control of the entire cell more accurate and effective. The dynamic splitting, merging, and boundary optimization strategies employed allow the division process to automatically adapt to the complexities of different regional power grids. The model eliminates complex characteristics, yielding reasonable and stable partitioning results and reducing manual intervention. By inputting a series of time-series data from cells into the power load forecasting model, it outputs the predicted load sequence data for the short, medium, and long term for each cell. Based on the predicted load sequence data, it determines whether the cell meets the new electrical capacity constraints. Different schemes are executed based on the judgment results, completing the identification of nodes to be optimized in the distribution network spatial resource allocation. This elevates traditional load forecasting to the level of spatial resource optimization and allocation, allowing for advance planning of capacity expansion and renovation, deployment of distributed resources, or adjustment of operating modes. It transforms passive response into proactive optimization, employing a bottom-up, step-by-step aggregation verification logic from cells to upper-level nodes, comprehensively improving the rationality and reliability of distribution network resource allocation.

[0107] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0108] Based on the same inventive concept, this application also provides a distribution network spatial resource allocation device for implementing the above-mentioned distribution network spatial resource allocation method. The solution provided by this device is similar to the implementation described in the above method; therefore, the specific limitations in one or more embodiments of the distribution network spatial resource allocation device provided below can be found in the limitations of the distribution network spatial resource allocation method described above, and will not be repeated here.

[0109] In one exemplary embodiment, such as Figure 4 As shown, a power distribution network spatial resource allocation device is provided, comprising: a cell determination module 401, a model construction module 402, a cell processing module 403, and a scheme determination module 404, wherein:

[0110] Cell determination module 401 is used to determine multiple initial cells of the distribution network based on the electrical topology diagram of the distribution network;

[0111] The model building module 402 is used to maximize the electrical connection tightness within multiple initial cells, minimize the difference between user load and power curves within multiple initial cells, and construct cell optimization models corresponding to multiple initial cells in combination with multiple preset constraints.

[0112] The cell processing module 403 is used to perform cell partitioning processing on the distribution network based on the cell optimization model to obtain multiple current cells of the distribution network.

[0113] The cell processing module 403 is also used to acquire multi-dimensional time series data of each current cell in a preset historical time period, and input the multi-dimensional time series data into a pre-built power load prediction model to obtain the predicted load sequence data of each current cell in the future preset time period.

[0114] The scheme determination module 404 is used to determine the matching status between each current cell and the preset electrical capacity constraints based on the predicted load sequence data corresponding to each current cell, and to determine the spatial resource allocation suitable for the distribution network based on the matching status.

[0115] Furthermore, in one embodiment, the cell processing module 403 is also used to input multi-dimensional time series data into a pre-built power load forecasting model to obtain the predicted load sequence data for each current cell within a preset future time period, including short-term predicted load sequence data, medium-term predicted load sequence data, and long-term predicted load sequence data; wherein the duration span between the data tail of the short-term predicted load sequence data and the data tail of the corresponding multi-dimensional time series data is less than the duration span between the data tail of the medium-term predicted load sequence data and the data tail of the corresponding multi-dimensional time series data, and the duration span between the data tail of the medium-term predicted load sequence data and the data tail of the corresponding multi-dimensional time series data is less than the duration span between the data tail of the long-term predicted load sequence data and the data tail of the corresponding multi-dimensional time series data.

[0116] Furthermore, in one embodiment, the cell processing module 403 is also used to obtain multi-dimensional time series data of each current cell in the preset historical time period based on the load time series data, power generation time series data, meteorological time series data and economic time series data of each current cell in the preset historical time period.

[0117] Furthermore, in one embodiment, the scheme determination module 404 is also used to determine the maximum total load of the current cell based on the predicted load sequence data corresponding to each current cell; determine the matching situation between the maximum total load of the current cell and the preset electrical capacity constraint; and if the matching situation is that the maximum total load of the current cell exceeds the threshold of the preset electrical capacity constraint, determine that the spatial resource allocation suitable for the distribution network is to mark the current cell as a node to be optimized and determine the highest-level power supply node corresponding to the node to be optimized, so as to allocate the spatial resources of the distribution network based on the power supply range and capacity characteristics of the highest-level power supply node.

[0118] Furthermore, in one embodiment, the model building module 402 is also used to maximize the electrical connection tightness within the multiple initial cells, minimize the difference between user load and power curves within the multiple initial cells, and combine preset constraints including allocation integrity constraints, spatial continuity constraints, scale appropriateness constraints, electrical capacity constraints, topology consistency constraints, and cell number constraints to build an optimization model corresponding to the multiple initial cells.

[0119] Furthermore, in one embodiment, the cell determination module 401 is also used to divide the smallest divisible units belonging to the same upstream transformer, line or switch into the same initial cell based on the electrical topology diagram of the distribution network, and to determine multiple initial cells in combination with the geographical area corresponding to the distribution network; the smallest divisible unit is a meter or a building unit.

[0120] Each module in the aforementioned power distribution network spatial resource allocation device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0121] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 5As shown, the computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores data on the allocation of spatial resources in the power distribution network. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a method for allocating spatial resources in a power distribution network.

[0122] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0123] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0124] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above method embodiments.

[0125] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0126] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0127] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0128] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0129] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for allocating spatial resources in a power distribution network, characterized in that, The method includes: The electrical topology of the distribution network is used to determine multiple initial cells of the distribution network; To maximize the electrical connection tightness within the multiple initial cells, minimize the difference between user load and power curves within the multiple initial cells, and combine multiple preset constraints, construct cell optimization models corresponding to the multiple initial cells. Based on the cell optimization model, the distribution network is divided into cells to obtain multiple current cells of the distribution network; The multi-dimensional time series data of each current cell in a preset historical time period are obtained, and the multi-dimensional time series data are input into a pre-constructed power load prediction model to obtain the predicted load sequence data of each current cell in a future preset time period. Based on the predicted load sequence data corresponding to each current cell, the matching status between each current cell and the preset electrical capacity constraint is determined, and the spatial resource allocation adapted to the distribution network is determined based on the matching status.

2. The method according to claim 1, characterized in that, The step of inputting the multi-dimensional time series data into a pre-built power load prediction model to obtain the predicted load sequence data for each current cell within a preset future time period includes: The multi-dimensional time series data is input into a pre-built power load forecasting model to obtain the predicted load sequence data for each current cell in the future preset time period, including short-term predicted load sequence data, medium-term predicted load sequence data and long-term predicted load sequence data. Wherein, the duration span between the data tail of the short-term predicted load sequence data and the corresponding data tail of the multi-dimensional time series data is less than the duration span between the data tail of the medium-term predicted load sequence data and the corresponding data tail of the multi-dimensional time series data, and the duration span between the data tail of the medium-term predicted load sequence data and the corresponding data tail of the multi-dimensional time series data is less than the duration span between the data tail of the long-term predicted load sequence data and the corresponding data tail of the multi-dimensional time series data.

3. The method according to claim 1 or 2, characterized in that, The step of obtaining multi-dimensional time series data of each current cell within a preset historical time period includes: Based on the load time series data, power generation time series data, meteorological time series data and economic time series data of each current cell in the preset historical time period, multi-dimensional time series data of each current cell in the preset historical time period are obtained.

4. The method according to claim 1, characterized in that, The step of determining the matching status between each current cell and the preset electrical capacity constraint based on the predicted load sequence data corresponding to each current cell, and determining the spatial resource allocation suitable for the distribution network based on the matching status, includes: Based on the predicted load sequence data corresponding to each current cell, determine the maximum total load of the current cell; The matching status between the maximum total load of the current cell and the preset electrical capacity constraint is determined. If the matching status is that the maximum total load of the current cell exceeds the threshold of the preset electrical capacity constraint, the spatial resource allocation suitable for the distribution network is determined by marking the current cell as a node to be optimized and determining the highest-level power supply node corresponding to the node to be optimized, so as to allocate the spatial resources of the distribution network based on the power supply range and capacity characteristics of the highest-level power supply node.

5. The method according to claim 1, characterized in that, The process involves maximizing the electrical connection tightness within the multiple initial cells, minimizing the differences between user load and power supply curves within the multiple initial cells, and combining multiple preset constraints to construct a cell optimization model corresponding to the multiple initial cells, including: To maximize the electrical connection tightness within the multiple initial cells, minimize the difference between user load and power curves within the multiple initial cells, and combine preset constraints including allocation integrity constraints, spatial continuity constraints, scale appropriateness constraints, electrical capacity constraints, topology consistency constraints, and cell number constraints, an optimization model corresponding to the multiple initial cells is constructed.

6. The method according to claim 1, characterized in that, The electrical topology diagram based on the distribution network determines multiple initial cells of the distribution network, including: Based on the electrical topology diagram of the distribution network, the smallest divisible units belonging to the same upstream transformer, line or switch are divided into the same initial cell, and multiple initial cells are determined in combination with the geographical area corresponding to the distribution network; the smallest divisible unit is a meter or a building unit.

7. A spatial resource allocation device for a power distribution network, characterized in that, The device includes: A cell determination module is used to determine multiple initial cells of the distribution network based on the electrical topology diagram of the distribution network; The model building module is used to maximize the electrical connection tightness within the multiple initial cells, minimize the difference between user load and power curves within the multiple initial cells, and construct cell optimization models corresponding to the multiple initial cells in combination with multiple preset constraints. The cell processing module is used to perform cell partitioning processing on the distribution network based on the cell optimization model to obtain multiple current cells of the distribution network. The cell processing module is also used to acquire multi-dimensional time series data of each current cell in a preset historical time period, and input the multi-dimensional time series data into a pre-built power load prediction model to obtain the predicted load sequence data of each current cell in a future preset time period. The scheme determination module is used to determine the matching status between each current cell and the preset electrical capacity constraint based on the predicted load sequence data corresponding to each current cell, and to determine the spatial resource allocation suitable for the distribution network based on the matching status.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.