Large language model low-voltage distribution network topology sensing method based on cloud edge fusion

Abstract

The invention discloses a large language model low-voltage power distribution network topology perception method based on cloud edge fusion, and relates to the technical field of power systems, and the method comprises the steps: collecting the data of a low-voltage power distribution network through an edge device, and analyzing a topological structure through a cloud model; standardizing the power distribution network data based on a large language model, and verifying topological information; and establishing an n-shaped model to realize direct-current and alternating-current power distribution network differential modeling, and based on the standardized power distribution network data, automatically generating a standard power distribution network model and generating power distribution network topology output. According to the method, low-voltage power distribution network data is collected through edge equipment, a topological structure is analyzed through a cloud model, multi-data source input is supported, topological information is verified, standardized processing and data quality control are achieved, direct-current and alternating-current power distribution network differential modeling is achieved by establishing an n-shaped model, and the modeling efficiency is improved. The calculation efficiency and precision of the model are improved, the application efficiency of topological data is improved through automatic output, and the multi-scene requirement is met.

Description

Technical Field

[0001] The present invention relates to the technical field of power systems, and in particular to a low-voltage distribution network topology perception method based on a large language model and cloud-edge fusion. Background Art

[0002] The low-voltage distribution network is an important part of the power system. Its main task is to transmit electricity from the high-voltage network to end users. Its operation quality directly affects the stability and security of power supply. However, with the rapid growth of electricity demand and the increasing complexity of the distribution network structure, the existing low-voltage distribution network management system has exposed many problems and it is difficult to meet the requirements of modern power systems for intelligence and efficiency.

[0003] At present, most low-voltage distribution network perception systems adopt a centralized data processing architecture, which relies on a central server to centrally process perception data and computing tasks. However, as the amount of data increases, the central server is prone to computing bottlenecks due to insufficient processing power, resulting in system response delays. In addition, real-time requirements make it difficult for traditional architectures to quickly respond to dynamic changes in the operating status of the distribution network, especially when load fluctuations are severe or faults occur, which may seriously affect scheduling and decision-making. At the same time, the centralized transmission of a large amount of perception data to the central server will increase bandwidth pressure and cause network congestion.

[0004] In terms of data processing and perception capabilities, the existing system shows obvious deficiencies. The equipment and topology data of the low-voltage distribution network come from various sources. Common parameters such as line length, resistance, reactance value and transformer capacity are often difficult to integrate due to non-uniform formats. In addition, manual data input is prone to errors, which affects the accuracy of scheduling and fault diagnosis. The distribution network topology is complex and the equipment types are diverse, but the existing system lacks accurate grasp of topology changes and node connection relationships, especially insufficient support for DC distribution networks. Traditional systems find it difficult to adapt to their unique equipment characteristics and operating requirements. In addition, the system has limited capabilities in data verification and reasoning, and cannot automatically detect data deviations or contradictions, resulting in problems being ignored, further affecting the safety and reliability of the distribution network. Summary of the invention

[0005] In view of the above-mentioned problems, the present invention is proposed.

[0006] Therefore, the technical problem solved by the present invention is: the existing low-voltage distribution network perception method has high limitations in data processing architecture, difficulty in integrating diverse data formats, lack of data verification capabilities, and the problem of how to achieve low-voltage distribution network perception and topology modeling through cloud-edge collaborative architecture and large language models.

[0007] To solve the above technical problems, the present invention provides the following technical solutions: a low-voltage distribution network topology perception method based on a large language model of cloud-edge fusion, comprising collecting low-voltage distribution network data through edge devices, and parsing the topological structure through a cloud model; standardizing the distribution network data based on a large language model, and verifying the topological information; establishing a Π-type model to realize differentiated modeling of DC and AC distribution networks, and automatically generating a standard distribution network model and generating a distribution network topology output based on the standardized distribution network data.

[0008] As an optimal solution of the low-voltage distribution network topology perception method based on cloud-edge fusion with a large language model described in the present invention, wherein: the collection of low-voltage distribution network data through edge devices includes receiving input from different data sources through edge devices, including natural language descriptions, electrical parameters and equipment status; automatically parsing and classifying input data, monitoring changes in the topology structure in real time, and uploading the parsed results to the cloud.

[0009] As a preferred scheme of the low-voltage distribution network topology perception method based on cloud-edge fusion with a large language model described in the present invention, wherein: the topology structure is parsed through the cloud model, including that after the large language model on the cloud receives the data transmitted from the edge device, it automatically parses the distribution network topology information in the transmitted data, including node connections, line electrical parameters, equipment type, working status, load distribution and distribution network hierarchy.

[0010] As an optimal solution of the low-voltage distribution network topology perception method based on cloud-edge fusion using a large language model described in the present invention, wherein: the standardized distribution network data based on the large language model includes using the large language model on the cloud side to analyze the received low-voltage distribution network data, automatically identifying and normalizing the input data from different data sources, and converting the input data from different data sources into a unified standard format through cloud-side model reasoning and analysis to generate a standardized distribution network topology structure; using the large language model on the cloud side to analyze the received low-voltage distribution network data also includes, when processing the input data, calculating the resistance and reactance values ​​according to the line model and line length, and deriving the power transmission characteristics according to the line electrical parameters to generate distribution network data that meets the standards.

[0011] As a preferred solution of the low-voltage distribution network topology perception method based on a large language model of cloud-edge fusion described in the present invention, wherein: the verification of topology information includes verifying the distribution network data based on the large language model on the cloud, and checking whether the received low-voltage distribution network data is consistent with the predefined model; if the data is inconsistent or deviated, the reasoning algorithm will be automatically triggered to compare with historical data or standard rules, calculate the difference between the data, generate a warning prompt and interactively feedback to the operator; when the operator receives a warning, the operator is prompted to further check the data and make corrections based on the difference prompt; if the operator fails to correct the error within the predetermined time, a higher priority reminder will be issued, or corrective measures will be automatically taken, including using standard data in historical records for replacement; if the corresponding standard data cannot be found in the historical data for replacement, the reasoning result based on the current data will be used for correction, or the missing data will be supplemented by requesting the operator to make corrections; if the operator fails to supplement the data, it will be temporarily corrected by setting a default value or inferring a reasonable value based on the existing data until manual verification.

[0012] As an optimal solution of the low-voltage distribution network topology perception method based on cloud-edge fusion of a large language model described in the present invention, wherein: the establishment of a Π-type model to realize differentiated modeling of DC and AC distribution networks includes, for the AC distribution network, generating a power flow calculation model that conforms to the Π-type model according to the line resistance and reactance parameters; for the DC distribution network, only the line resistance parameters are calculated, and the reactance part is ignored to simplify the model calculation.

[0013] As an optimal solution of the low-voltage distribution network topology perception method based on a large language model of cloud-edge fusion described in the present invention, wherein: based on standardized distribution network data, a standard distribution network model is automatically generated and the distribution network topology output is generated, including supporting the modeling of various equipment in the distribution network, including lines, transformers, capacitors, and reactive compensation devices, and automatically generating a standard distribution network model; the distribution network data after standardization and verification will be output in a standard format, including the topological structure of the distribution network, the resistance, reactance and capacitance values ​​of the lines, and the line parameters generated according to the Π-type model. The generated distribution network topology diagram is visualized through the cloud model, and an interface is provided to export the standardized data to the simulation analysis and scheduling optimization system, which is convenient for real-time monitoring and optimized scheduling.

[0014] Another object of the present invention is to provide a low-voltage distribution network topology perception system based on a large language model of cloud-edge fusion, which can standardize the distribution network data based on the large language model and verify the topology information, thereby solving the problem that the current low-voltage distribution network perception technology contains missing data verification capabilities.

[0015] As an optimal solution of the low-voltage distribution network topology perception system based on a large language model of cloud-edge fusion described in the present invention, it includes: a data acquisition and transmission module, a cloud-side parsing and standardization module, and a modeling and output module; the data acquisition and transmission module is used to collect low-voltage distribution network data through edge devices, and parse the topological structure through a cloud-side model; the cloud-side parsing and standardization module is used to standardize the distribution network data based on a large language model, and verify the topological information; the modeling and output module is used to establish a Π-type model to realize differentiated modeling of DC and AC distribution networks, and generate distribution network topology output.

[0016] A computer device includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement a step of a low-voltage distribution network topology perception method based on a large language model of cloud-edge fusion.

[0017] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of a low-voltage distribution network topology perception method based on a large language model of cloud-edge fusion.

[0018] Beneficial effects of the present invention: The low-voltage distribution network topology perception method based on a large language model of cloud-edge fusion provided by the present invention collects low-voltage distribution network data through edge devices, improves the flexibility of data collection, adapts to diverse data sources, supports unified processing of multi-source data, improves the compatibility and scalability of the system, realizes efficient data flow and real-time update, enhances the dynamic monitoring capability of the distribution network, and parses the topological structure through a cloud model. The cloud model realizes intelligent parsing of complex topological information, avoids time-consuming and labor-intensive manual analysis, and separates real-time collection and computing-intensive tasks through a cloud-edge collaborative architecture, optimizes the resource allocation of the system, not only improves the parsing accuracy and speed of topological information, but also reduces the computing burden of edge devices and improves real-time response capabilities. By standardizing distribution network data based on a large language model, standardized data facilitates subsequent calculations and analysis, solves the processing difficulties caused by the diversity of data formats, extracts and calculates the electrical parameters of the lines, and derives the power transmission characteristics, providing accurate data for distribution network modeling. According to the support, the influence of format differences of different data sources is eliminated, the data processing efficiency is improved, the topology information is verified, the data quality control is realized, the credibility of the distribution network topology model is improved, the automatic error correction and prompt functions are provided, the burden of manual intervention is reduced, and the efficiency of data verification is improved. By establishing a Π-type model, differentiated modeling of DC and AC distribution networks is realized. The differentiated modeling is optimized for the characteristics of DC and AC power grids, the calculation efficiency and accuracy of the model are improved, and the automated modeling of various types of equipment in the distribution network is supported. The redundant calculations in the traditional modeling method are reduced, and the modeling efficiency is improved. Automated modeling reduces the risk of manual modeling errors and ensures the consistency of the model. Based on standardized distribution network data, a standard distribution network model is automatically generated and the distribution network topology output is generated. The automated output improves the application efficiency of the topology data and meets the needs of multiple scenarios. The visual display and interface output enhance the operability and user experience of the system. The present invention achieves better results in terms of efficiency, flexibility and reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0020] Figure 1 An overall flow chart of a low-voltage distribution network topology perception method based on a large language model and cloud-edge fusion provided for the first embodiment of the present invention.

[0021] Figure 2A module schematic diagram of a low-voltage distribution network topology perception system based on a large language model and cloud-edge fusion provided for the third embodiment of the present invention. DETAILED DESCRIPTION

[0022] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0023] Example 1, reference Figure 1 , is an embodiment of the present invention, and provides a low-voltage distribution network topology perception method based on a large language model of cloud-edge fusion, including:

[0024] S1: Collect low-voltage distribution network data through edge devices and analyze the topology structure through cloud models.

[0025] Furthermore, collecting low-voltage distribution network data through edge devices includes receiving input from different data sources through edge devices, including natural language descriptions, electrical parameters and device status; automatically parsing and classifying input data, monitoring changes in topology structures in real time, and uploading parsing results to the cloud.

[0026] It should be noted that after receiving the data transmitted from the edge device through the cloud model parsing the topology structure, including the large language model on the cloud, the distribution network topology information in the transmitted data is automatically parsed, including node connections, line electrical parameters, equipment type, working status, load distribution and distribution network hierarchy.

[0027] It should also be noted that by receiving inputs from different data sources through edge devices, multiple forms of data input can be supported, whether it is manually entered digital information, uploaded files (such as Excel, CSV, etc.), or real-time data obtained through API interfaces, all can be uniformly formatted and converted into standardized data formats, without the need for users to input in a fixed format, supporting flexible data collection methods; automatically parsing and classifying input data, edge devices will extract useful information from different data sources (such as natural language descriptions, electrical parameters, device status, etc.). For example, the system identifies the device type or electrical parameters from the text description, or extracts key parameters such as line resistance and voltage from the numerical data, and classifies the parsed data for storage or processing by type. For example, natural language descriptions are classified as device information, numerical data are classified as electrical parameters, and data related to topology structures are classified as topology descriptions. This is done to make subsequent data processing more efficient and accurate.

[0028] It should also be noted that the large language model LLM (Large Language Model) on the cloud automatically parses the distribution network topology information in the transmitted data after receiving the data transmitted from the edge device. Specifically, the large language model on the cloud will receive data from the edge device, conduct a comprehensive analysis, and extract the topological structure information of the distribution network, including electrical parameters (such as resistance and reactance) and the connection relationship between devices, etc., and finally construct a standardized topological model of the distribution network. Through the cloud-edge collaborative architecture, computing-intensive tasks are separated from real-time data collection, which not only improves the real-time response capability of the system, but also reduces bandwidth requirements.

[0029] S2: Standardize distribution network data based on a large language model and verify topology information.

[0030] Furthermore, standardizing distribution network data based on a large language model includes using a large language model on the cloud to analyze the received low-voltage distribution network data, automatically identifying and normalizing input data from different data sources, and converting the input data from different data sources into a unified standard format through cloud-based model reasoning and analysis to generate a standardized distribution network topology structure; using a large language model on the cloud to analyze the received low-voltage distribution network data also includes calculating the resistance and reactance values ​​based on the line model and line length when processing the input data, and deriving the power transmission characteristics based on the line electrical parameters to generate distribution network data that meets the standards.

[0031] It should be noted that verifying the topology information includes verifying the distribution network data based on the large language model in the cloud to check whether the received low-voltage distribution network data is consistent with the predefined model; if the data is inconsistent or deviated, the reasoning algorithm will be automatically triggered to compare with the historical data or standard rules, calculate the difference between the data, generate a warning prompt and interactively feedback to the operator; when the operator receives a warning, he will be prompted to further check the data and make corrections based on the difference prompts; if the operator fails to correct the error within the predetermined time, a higher priority reminder will be issued, or corrective measures will be automatically taken, including using standard data in historical records for replacement; if the corresponding standard data cannot be found in the historical data for replacement, the reasoning results based on the current data will be used for correction, or the missing data will be requested to be supplemented by the operator; if the operator fails to supplement the data, it will be temporarily corrected by setting a default value or inferring a reasonable value based on the existing data until manual verification.

[0032] It should also be noted that the distribution network data can be standardized based on the large language model and the topology information can be verified. The GPT-4 large language model can be used to standardize the distribution network data and verify the topology information; the received low-voltage distribution network data can be analyzed using the large language model on the cloud. It is not just a simple analysis. The cloud model will also reason based on historical data, known rules, and input data, so as to supplement and correct incomplete or biased data. The predefined model refers to the standard model established in advance by the system, which usually includes the topology structure, equipment parameters, node connections and other benchmark data of the distribution network. These models can be generated through previous collection, historical data, expert knowledge or system preset rules. As a reference for correctness verification, its role is to ensure that the input data is consistent with these standard models to verify the accuracy and consistency of the data.

[0033] S3: Establish a Π-type model to realize differentiated modeling of DC and AC distribution networks. Based on standardized distribution network data, automatically generate a standard distribution network model and generate distribution network topology output.

[0034] Furthermore, a Π-type model is established to realize differentiated modeling of DC and AC distribution networks, including generating a power flow calculation model that conforms to the Π-type model for the AC distribution network based on the line resistance and reactance parameters; for the DC distribution network, only the line resistance parameters are calculated and the reactance part is ignored to simplify the model calculation.

[0035] It should be noted that based on the standardized distribution network data, the standard distribution network model is automatically generated and the distribution network topology output is generated, including supporting the modeling of various equipment in the distribution network, including lines, transformers, capacitors, and reactive compensation devices, and automatically generating a standard distribution network model; the distribution network data after standardization and verification will be output in a standard format, including the topological structure of the distribution network, the resistance, reactance and capacitance values ​​of the lines, and the line parameters generated according to the Π-type model. The generated distribution network topology diagram is visualized through the cloud model, and an interface is provided to export standardized data to the simulation analysis and scheduling optimization system, which is convenient for real-time monitoring and optimized scheduling.

[0036] It should also be noted that for the DC distribution network, only the line resistance is calculated, which simplifies the modeling process and improves the calculation efficiency. It can also support the modeling of various equipment in the distribution network, such as lines, transformers, capacitors, reactive compensation devices, etc., and automatically generate a distribution network model that meets the standards. Through automated modeling, the errors and inconsistencies in the manual modeling process can be greatly reduced, ensuring the accuracy and efficiency of the distribution network model. The distribution network data after standardization and verification will be output in a standard format, including the topological structure of the distribution network, the resistance, reactance and capacitance values ​​of the lines, etc., as well as the line parameters generated according to the Π-type model. These data can be directly applied to the simulation analysis, optimization scheduling and fault diagnosis of the distribution network.

[0037] Example 2 is an embodiment of the present invention, which provides a low-voltage distribution network topology perception method based on a large language model of cloud-edge fusion. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0038] This experiment aims to verify the innovation and effectiveness of a low-voltage distribution network topology perception method based on a large language model of cloud-edge fusion. The experiment uses three sets of input data sources, including formatted Excel files, random txt files and line parameters with theoretical calculation verification, for comparative testing.

[0039] Experimental equipment and environment

[0040] 1. Hardware equipment: edge devices (embedded processors with data collection and real-time processing capabilities), supporting sensors, and cloud servers (supporting the operation of large language models).

[0041] 2. Software tools: Cloud-based parsing model based on GPT-4, independently developed distribution network data acquisition and processing software, and BPA power simulation tool.

[0042] 3. Input data source:

[0043] Group 1: Standardized Excel file containing parameter data of lines and transformers.

[0044] Group 2: Unformatted txt files containing line and transformer parameters in random order.

[0045] Group 3: Theoretically verified line and transformer parameters, calculated based on line model design data.

[0046] Experimental procedures

[0047] First, data collection and preprocessing are carried out. The data in Excel files and txt files are uploaded to the edge device. The edge device automatically formats and classifies the data, and classifies natural language descriptions, electrical parameters, equipment status, etc. into standardized information. Then, cloud parsing is carried out. The edge device transmits the preprocessed data to the cloud. The cloud uses the GPT-4 model to parse the topological structure of the line and transformer, including node connection, resistance, reactance and load distribution. Then, for group 3 data, the cloud parsing model performs theoretical value calculation and input value comparison. If an abnormality is found, a prompt is returned. Then, the standardized distribution network topology data is verified to check the consistency with the predefined model. If there is any deviation, it is automatically corrected. Then, based on the Π-type model, a differentiated model of DC and AC distribution networks is constructed, and a standardized distribution network topology that meets the BPA simulation requirements is automatically generated. Finally, the final processed data is exported to a BPA-compatible format and visualized.

[0048] Refer to Table 1 for a comparative analysis of the experimental data.

[0049] Table 1 Experimental data record table

[0050]

[0051]

[0052] It can be seen from the experimental data in Table 1 that the present invention has obvious advantages in data collection and analysis, topological model establishment and standardized processing. The details are as follows:

[0053] 1. Data collection and analysis capabilities:

[0054] By comparing the results of Group 1 and Group 2, it can be found that edge devices can efficiently parse and classify data regardless of whether the data format is standardized.

[0055] Especially for the disordered data in Group 2, the system can accurately identify the parameters and complete automatic standardization processing to keep consistent with the results in Group 1.

[0056] 2. Accuracy and consistency of topological model:

[0057] From the theoretical value verification, it can be seen that the resistance and reactance values ​​of all line data in Group 3 are within the error range of ±5%, indicating that the calculation of the cloud model is highly consistent with the input data.

[0058] 3. Comparison of innovation with existing technologies:

[0059] The existing technology relies on manual input and data verification, which is prone to problems such as missing data or format inconsistency. The edge-cloud collaborative architecture of the present invention not only supports the collection of multiple data sources, but also can automatically perform formatting, greatly reducing manual intervention.

[0060] Through the LLM inference model in the cloud, the present invention can infer and correct incomplete or abnormal data to ensure the integrity and accuracy of the distribution network model.

[0061] 4. Practical application value:

[0062] The generated standardized distribution network data is directly compatible with BPA simulation and analysis tools, realizing integrated applications from data collection to simulation optimization.

[0063] In power dispatching and fault diagnosis, the rapid response capability of the present invention can effectively improve the stability and safety of power grid operation.

[0064] In summary, the present invention has obvious innovation and practical value in low-voltage distribution network topology analysis and modeling technology. Through automated data processing, model construction and standardized output, it achieves a double breakthrough in technology and application. Therefore, the present invention is creative.

[0065] Example 3, reference Figure 2 , is an embodiment of the present invention, which provides a low-voltage distribution network topology perception system based on a large language model of cloud-edge fusion, including a data acquisition and transmission module, a cloud-side parsing and standardization module, and a modeling and output module.

[0066] The data acquisition and transmission module is used to collect low-voltage distribution network data through edge devices and parse the topological structure through the cloud model; the cloud parsing and standardization module is used to standardize the distribution network data based on the large language model and verify the topological information; the modeling and output module is used to establish a Π-type model to realize differentiated modeling of DC and AC distribution networks, and generate distribution network topology output.

[0067] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0068] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0069] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering, or processing in another suitable manner as necessary, and then stored in a computer memory.

[0070] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc. It should be noted that the above embodiments are only used to illustrate the technical solution of the present invention and are not limited. Although the present invention is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention, which should be included in the scope of the claims of the present invention.

[0071] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A low-voltage distribution network topology perception method based on a large language model of cloud-edge fusion, characterized in that: include: Collect low-voltage distribution network data through edge devices and analyze the topology structure through cloud models; Standardize distribution network data based on large language models and verify topology information; A Π-type model is established to realize differentiated modeling of DC and AC distribution networks. Based on standardized distribution network data, a standard distribution network model is automatically generated and the distribution network topology output is generated.

2. The low-voltage distribution network topology perception method based on a large language model of cloud-edge fusion according to claim 1 is characterized in that: The collecting of low-voltage distribution network data through edge devices includes receiving inputs from different data sources through edge devices, including natural language descriptions, electrical parameters, and device status; Automatically parse and classify input data, monitor topological changes in real time, and upload parsing results to the cloud.

3. The low-voltage distribution network topology perception method based on large language model of cloud-edge fusion according to claim 2 is characterized by: The method of parsing the topology structure through the cloud model includes automatically parsing the distribution network topology information in the transmitted data after receiving the data transmitted from the edge device, including node connections, line electrical parameters, equipment types, working status, load distribution and distribution network hierarchy.

4. The low-voltage distribution network topology perception method based on large language model of cloud-edge fusion according to claim 3 is characterized by: The method of standardizing the distribution network data based on the large language model includes analyzing the received low-voltage distribution network data using the large language model in the cloud, automatically identifying and normalizing the input data from different data sources, converting the input data from different data sources into a unified standard format through cloud model reasoning analysis, and generating a standardized distribution network topology structure; The received low-voltage distribution network data is analyzed using the large language model in the cloud. When processing the input data, the resistance and reactance values ​​are calculated based on the line model and line length, and the power transmission characteristics are derived based on the line electrical parameters to generate distribution network data that meets the standards.

5. The low-voltage distribution network topology perception method based on large language model of cloud-edge fusion according to claim 4 is characterized by: The said verifying the topology information includes verifying the distribution network data based on the large language model on the cloud, and checking whether the received low-voltage distribution network data is consistent with the predefined model; If the data is inconsistent or deviated, the inference algorithm will be automatically triggered to compare with historical data or standard rules, calculate the difference between the data, generate warning prompts and interactively feedback to the operator; When the operator receives a warning, he is prompted to further check the data and make corrections based on the discrepancy prompts; If the operator fails to correct the error within the predetermined time, a higher priority reminder will be issued, or corrective measures will be automatically taken, including replacement with standard data in the historical record; If the corresponding standard data cannot be found in the historical data for replacement, corrections will be made using the inference results based on the current data, or by requesting the operator to supplement the missing data; If the operator fails to supplement the data, a temporary correction will be made by setting a default value or inferring a reasonable value based on the existing data until manual verification.

6. The method for low-voltage distribution network topology perception based on large language model of cloud-edge fusion according to claim 5 is characterized in that: The establishment of the Π-type model to realize the differentiated modeling of DC and AC distribution networks includes generating a power flow calculation model that conforms to the Π-type model for the AC distribution network according to line resistance and reactance parameters; For DC distribution networks, only the line resistance parameters are calculated and the reactance part is ignored to simplify the model calculation.

7. The method for low-voltage distribution network topology perception based on large language model of cloud-edge fusion according to claim 6 is characterized in that: The method of automatically generating a standard distribution network model and generating a distribution network topology output based on standardized distribution network data includes supporting modeling of various devices in the distribution network, including lines, transformers, capacitors, and reactive power compensation devices, and automatically generating a standard distribution network model; The distribution network data after standardization and verification will be output into a standard format, including the topological structure of the distribution network, the resistance, reactance and capacitance values ​​of the lines, and the line parameters generated according to the Π-type model. The generated distribution network topology diagram will be visualized through the cloud model, and an interface will be provided to export standardized data to the simulation analysis and scheduling optimization system for real-time monitoring and optimized scheduling.

8. A system using the low-voltage distribution network topology perception method based on a large language model based on cloud-edge fusion as described in any one of claims 1 to 7, characterized in that: Including data collection and transmission module, cloud analysis and standardization module, modeling and output module; The data acquisition and transmission module is used to collect low-voltage distribution network data through edge devices and analyze the topology structure through cloud models; The cloud parsing and standardization module is used to standardize the distribution network data based on the large language model and verify the topology information; The modeling and output module is used to establish a Π-type model to realize differentiated modeling of DC and AC distribution networks, and automatically generate a standard distribution network model and a distribution network topology output based on standardized distribution network data.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the low-voltage distribution network topology perception method based on a large language model based on cloud-edge fusion as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the low-voltage distribution network topology perception method based on a large language model based on cloud-edge fusion as described in any one of claims 1 to 7 are implemented.

Images