Power grid design model optimization method and device, computer equipment, storage medium and computer program product

Abstract

The invention relates to a power grid design model optimization method and device, computer equipment, a storage medium and a computer program product. Relates to the power system field. The method comprises the following steps: inputting target condition data for a to-be-optimized power grid into a preset power grid design model to obtain power grid design target data output by the power grid design model; the power grid design model processes the target condition data based on a built-in preset electrical specification and a preset constraint optimization algorithm; the target data comprises a main wiring scheme, a line trend, substation site selection and basic structure design parameters of the to-be-optimized power grid; calling an automatic test engine to verify and evaluate an initial optimization power grid formed based on the target data; and feeding back a verification and evaluation result to the power grid design model to optimize the power grid design model. By adopting the method, the power grid design efficiency of the existing power grid design model can be improved.

Description

Technical Field

[0001] This application relates to the field of power system technology, and in particular to a method, apparatus, computer equipment, storage medium and computer program product for optimizing power grid design models. Background Technology

[0002] With the expansion of power system scale and the increase in the proportion of new energy sources connected to the power system, the complexity of power grid design has increased dramatically. Traditional power grid design relies on engineers to manually complete the work based on design standards and experience. From load forecasting, line route planning, equipment selection and configuration to electrical calculation verification, it usually requires crossing multiple professional software and platforms. The data integration and verification accuracy in the design process pose significant challenges.

[0003] Currently, automated power grid design systems typically generate some design drawings using predefined rules or templates. However, their automation level is limited, requiring a large amount of manual parameter input and lacking intelligent verification. This leads to design optimization relying on manual judgment, a lack of dynamic testing and feedback mechanisms, and consequently, reduced efficiency in power grid design. Therefore, existing power grid design models suffer from low 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 optimizing power grid design models to address the technical problem of low power grid design efficiency in existing power grid design models.

[0005] Firstly, this application provides a method for optimizing a power grid design model, including:

[0006] The target condition data for the power grid to be optimized is input into a preset power grid design model to obtain the target data for power grid design output by the power grid design model; the power grid design model processes the target condition data based on built-in preset electrical specifications and preset constraint optimization algorithms; the target data includes the main wiring scheme, line route, substation site selection and basic structure design parameters of the power grid to be optimized.

[0007] The automated testing engine is invoked to verify and evaluate the initial optimized power grid based on the target data;

[0008] The results of the verification and evaluation are fed back into the power grid design model to optimize the power grid design model.

[0009] In one embodiment, the method further includes: collecting electrical topology data, geospatial data, load forecast data, equipment characteristic data, and civil design data of the power grid to be optimized; and converting the electrical topology data, geospatial data, load forecast data, equipment characteristic data, and civil design data into a unified representation method through a multimodal coding network to obtain target condition data of the power grid to be optimized.

[0010] In one embodiment, the step of transforming the electrical topology data, the geospatial data, the load forecast data, the equipment characteristic data, and the civil engineering design data into a unified representation method through a multimodal coding network includes: transforming the electrical topology data, the geospatial data, the load forecast data, the equipment characteristic data, and the civil engineering design data into a unified representation method by combining a multimodal coding network with a fusion attention mechanism and a graph neural network.

[0011] In one embodiment, the step of inputting the target condition data for the power grid to be optimized into a preset power grid design model includes: inputting the access capacity, load density, terrain constraints, and investment ceiling, which are included in the target condition data for the power grid to be optimized, into the preset power grid design model.

[0012] In one embodiment, the results of the verification and evaluation are fed back to the power grid design model to optimize the power grid design model, including: generating multi-scenario test cases based on the results of the verification and evaluation; the multi-scenario test cases include power flow calculation, short-circuit analysis, and transient stability simulation; performing simulation tests on the multi-scenario test cases to obtain simulation results; comparing the simulation results with preset design standards, optimizing abnormal data and performance deviation data, and optimizing the power grid design model based on the optimized abnormal data and performance deviation data.

[0013] In one embodiment, the step of comparing the simulation results with preset design standards to optimize abnormal data and performance deviation data includes: comparing the simulation results with preset design standards, collecting abnormal data and performance deviation data of the design model corresponding to the power grid to be optimized; and correcting the weight parameters corresponding to the abnormal data and performance deviation data of the design model through a reinforcement learning algorithm to optimize the design model.

[0014] Secondly, this application also provides a power grid design model optimization device, comprising:

[0015] The data generation module is used to input the target condition data of the power grid to be optimized into a preset power grid design model to obtain the target data of the power grid design output by the power grid design model; the power grid design model processes the target condition data based on the built-in preset electrical specifications and preset constraint optimization algorithms; the target data includes the main wiring scheme, line route, substation site selection and basic structure design parameters of the power grid to be optimized.

[0016] The verification module is used to call the automated testing engine to verify and evaluate the initial optimized power grid based on the target data;

[0017] The model optimization module is used to feed back the results of the verification and evaluation to the power grid design model in order to optimize the power grid design model.

[0018] 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:

[0019] The target condition data for the power grid to be optimized is input into a preset power grid design model to obtain the target data for power grid design output by the power grid design model; the power grid design model processes the target condition data based on built-in preset electrical specifications and preset constraint optimization algorithms; the target data includes the main wiring scheme, line route, substation site selection and basic structure design parameters of the power grid to be optimized.

[0020] The automated testing engine is invoked to verify and evaluate the initial optimized power grid based on the target data;

[0021] The results of the verification and evaluation are fed back into the power grid design model to optimize the power grid design model.

[0022] 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:

[0023] The target condition data for the power grid to be optimized is input into a preset power grid design model to obtain the target data for power grid design output by the power grid design model; the power grid design model processes the target condition data based on built-in preset electrical specifications and preset constraint optimization algorithms; the target data includes the main wiring scheme, line route, substation site selection and basic structure design parameters of the power grid to be optimized.

[0024] The automated testing engine is invoked to verify and evaluate the initial optimized power grid based on the target data;

[0025] The results of the verification and evaluation are fed back into the power grid design model to optimize the power grid design model.

[0026] Fifthly, this application also provides a computer program product, which includes a computer program that, when executed by a processor, performs the following steps:

[0027] The target condition data for the power grid to be optimized is input into a preset power grid design model to obtain the target data for power grid design output by the power grid design model; the power grid design model processes the target condition data based on built-in preset electrical specifications and preset constraint optimization algorithms; the target data includes the main wiring scheme, line route, substation site selection and basic structure design parameters of the power grid to be optimized.

[0028] The automated testing engine is invoked to verify and evaluate the initial optimized power grid based on the target data;

[0029] The results of the verification and evaluation are fed back into the power grid design model to optimize the power grid design model.

[0030] The aforementioned power grid design model optimization method, device, computer equipment, storage medium, and computer program product, in the process of power grid design model optimization, firstly input the target condition data of the power grid to be optimized into the preset power grid design model to obtain the target data of the power grid design output by the power grid design model; the power grid design model processes the target condition data based on the built-in preset electrical specifications and preset constraint optimization algorithms; the target data includes the main wiring scheme, line route, substation site selection, and basic structure design parameters of the power grid to be optimized; the automated test engine is called to verify and evaluate the initial optimized power grid based on the target data; the verification and evaluation results are fed back to the power grid design model to optimize the power grid design model. In the above process, target data is generated by using a pre-set power grid design model and built-in electrical specifications, combined with constraint optimization algorithms. After automated testing and verification, the data is fed back to optimize the design model. This eliminates the need for manual completion of complex designs such as main wiring schemes and substation site selection, significantly reducing manual intervention in parameter input and scheme adjustments, shortening the design cycle. Furthermore, the built-in specifications and constraint optimization algorithms of the pre-set power grid design model ensure that the target data meets the requirements of electrical safety, economy, and construction feasibility, avoiding scheme defects caused by human experience bias. In addition, the verification stage completes multi-dimensional evaluation through an automated testing engine, improving verification efficiency and result accuracy. Therefore, the above process improves the power grid design efficiency of existing power grid design models. Attached Figure Description

[0031] 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.

[0032] Figure 1 This is a flowchart illustrating a power grid design model optimization method in one embodiment;

[0033] Figure 2 This is a flowchart illustrating the optimization steps of a power grid design model in one embodiment;

[0034] Figure 3 This is a flowchart illustrating the power grid design model optimization method in another embodiment;

[0035] Figure 4 This is a structural block diagram of a power grid design model optimization device in one embodiment;

[0036] Figure 5 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0037] 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.

[0038] With the expansion of power system scale and the increase in the proportion of renewable energy integration, the complexity of power grid design has risen sharply. Traditional power grid design relies on engineers to manually complete the work based on design standards and experience. From load forecasting, line route planning, equipment selection and configuration to electrical calculation verification, it usually requires the use of multiple professional software and platforms. Moreover, under the background of "dual carbon" goals and smart grid construction, power grid design must not only meet electrical safety and economic requirements, but also take into account multi-dimensional indicators such as distributed energy integration, flexible dispatch, power supply reliability, and life cycle economics. This places higher demands on design automation, data integration, and verification accuracy. However, current power grid design automation systems generate some design drawings, such as main wiring diagrams, protection configurations, or distribution schemes, through predefined rules or templates. However, the degree of automation is limited, and a large number of parameters still need to be manually input. Intelligent verification is not possible, and the optimization of design schemes depends on manual judgment, lacking dynamic testing and feedback mechanisms. Another related technology uses deep learning or graph neural network models to predict the power grid operating status, fault risk, or power flow distribution to provide auxiliary decision-making for operation and maintenance. However, it is mainly applied to operation analysis and operation and maintenance prediction, and fails to cover the automated generation and verification of the design stage. Therefore, existing power grid design models have low power grid design efficiency.

[0039] To address the technical problem of low power grid design efficiency in existing power grid design models, in an exemplary embodiment, such as... Figure 1 As shown, a power grid design model optimization method is provided. Taking the application of this method to a terminal as an example, the method includes the following steps S102 to S106. Wherein:

[0040] Step S102: Input the target condition data of the power grid to be optimized into the preset power grid design model to obtain the target data of the power grid design output by the power grid design model; The power grid design model processes the target condition data based on the built-in preset electrical specifications and preset constraint optimization algorithms; The target data includes the main wiring scheme, line route, substation site selection and basic structure design parameters of the power grid to be optimized.

[0041] The pre-set power grid design model is trained using a large amount of power industry data and incorporates semantic understanding and logical reasoning capabilities specific to the power industry. This model can accurately identify the needs in power grid design, such as automatically associating design specifications for substations of different voltage levels. The built-in electrical specifications are power industry technical standards and design guidelines pre-integrated into the power grid design model, including specific clauses such as electrical safety distances, equipment selection parameters, short-circuit current limits, and insulation level requirements. The constrained optimization algorithm is a mathematical algorithm used to find the optimal solution under multi-objective conditions. The user input conditions are key parameters provided by the user based on specific power grid design requirements. The target data can include the main wiring scheme of the power grid to be optimized, line routes, substation site selection, and basic structure design parameters.

[0042] For example, in the optimization process of the power grid to be optimized, the constraints of the constrained optimization algorithm include the upper limit of investment cost, terrain restrictions, power supply reliability requirements, construction difficulty, etc. The constrained optimization algorithm can calculate the construction cost and power supply radius of substation sites for different power grid systems, and select the scheme with the lowest cost and the widest load coverage. The user input conditions may include the area of ​​the power grid to be optimized, the access capacity, and the load density, etc.

[0043] Step S104: Call the automated test engine to verify and evaluate the initial optimized power grid based on the target data.

[0044] Among them, the automated testing engine is software with the functions of automatically executing test processes, data acquisition and result analysis. The automated testing engine can automatically call simulation tools, generate test scripts in batches, monitor the test process in real time, and automatically summarize test data to shorten the verification cycle and avoid errors caused by manual operation. Verification and evaluation is a process of comprehensively testing the accuracy of the target data corresponding to the power grid to be optimized, which can include electrical performance verification, civil engineering safety assessment and compliance inspection.

[0045] More often, the initial optimized power grid is based on the target data, which may consist solely of the target data, including the main wiring scheme, line routes, substation locations, and basic structural design parameters of the power grid to be optimized; or it may consist of the target data, including the main wiring scheme, line routes, substation locations, and basic structural design parameters of the power grid to be optimized, as well as data that does not require optimization.

[0046] Step S106: Feed the verification and evaluation results back to the power grid design model to optimize the power grid design model.

[0047] The verification and evaluation results are the detection conclusions output by the automated testing engine, including whether the target data meets the standards and the deviation values ​​when it does not meet the standards; the design model corresponding to the power grid to be optimized may include the parameter configuration of the preset power grid design model, the weight settings of the constraint optimization algorithm, the data processing logic, etc.; the optimization design model is the process of adjusting and upgrading the design model based on the verification and evaluation results, which may include: correcting parameters in the design model that do not match reality, supplementing uncovered design scenarios, and optimizing the constraint weights, etc.

[0048] In the aforementioned power grid design model optimization method, the target condition data for the power grid to be optimized are first input into a preset power grid design model to obtain the target data of the power grid design output by the power grid design model. The power grid design model processes the target condition data based on built-in preset electrical specifications and preset constraint optimization algorithms. The target data includes the main wiring scheme, line route, substation site selection, and basic structure design parameters of the power grid to be optimized. An automated testing engine is called to verify and evaluate the initial optimized power grid based on the target data. The results of the verification and evaluation are fed back to the power grid design model to optimize the power grid design model. In the above process, target data is generated by using a pre-set power grid design model and built-in electrical specifications, combined with constraint optimization algorithms. After automated testing and verification, the data is fed back to optimize the design model. This eliminates the need for manual completion of complex designs such as main wiring schemes and substation site selection, significantly reducing manual intervention in parameter input and scheme adjustments, shortening the design cycle. Furthermore, the built-in specifications and constraint optimization algorithms of the pre-set power grid design model ensure that the target data meets the requirements of electrical safety, economy, and construction feasibility, avoiding scheme defects caused by human experience bias. In addition, the verification stage completes multi-dimensional evaluation through an automated testing engine, improving verification efficiency and result accuracy. Therefore, the above process improves the power grid design efficiency of existing power grid design models.

[0049] In one exemplary embodiment, such as Figure 2 As shown, the power grid design model optimization method further includes: step S202, collecting electrical topology data, geospatial data, load forecast data, equipment characteristic data, and civil design data of the power grid to be optimized; step S204, converting the electrical topology data, geospatial data, load forecast data, equipment characteristic data, and civil design data into a unified representation method through a multimodal coding network to obtain the target condition data of the power grid to be optimized.

[0050] Among them, electrical topology data refers to the connection relationships and structural information of various components in the power grid to be optimized, such as generators, transformers, and circuit breakers, which may include component models, connection methods, and voltage levels; geospatial data refers to geographic information data related to the construction area of ​​the power grid to be optimized, which may include topography, such as plains and mountains; land use types, such as arable land and construction land; transportation conditions, such as proximity to highways and distance from railways; and environmentally sensitive areas, such as nature reserves and residential areas; load forecast data refers to the predicted electricity load of the power supply area of ​​the power grid to be optimized for a period of time in the future; equipment characteristic data refers to the performance indicators of various electrical and civil engineering equipment required for the power grid to be optimized; and civil engineering design data refers to the design parameters and requirements related to the civil engineering of the power grid to be optimized.

[0051] More specifically, multimodal coding networks are neural network models that can process various types of data and transform them into a unified format. They can extract modal features through coding subnetworks specific to different modalities, and then capture the correlation between different modal data through cross-modal attention mechanisms, ultimately outputting a feature vector of a unified dimension to achieve semantic alignment of heteromodal data. The unified representation method transforms multimodal power grid design data with different formats, types, and semantics into a form with a unified structure.

[0052] In this embodiment, by collecting multi-dimensional power grid design data and transforming it into a unified representation, the foundation is laid for the intelligent design and verification of the power grid design model to be optimized. By using a multimodal coding network to transform different data into a unified representation, the preset power grid design model can efficiently call the full amount of data for collaborative reasoning without the need for manual and repeated conversion of data formats, thus improving data processing efficiency.

[0053] In one embodiment, electrical topology data, geospatial data, load forecast data, equipment characteristic data, and civil design data are transformed into a unified representation through a multimodal coding network. This includes: transforming electrical topology data, geospatial data, load forecast data, equipment characteristic data, and civil design data into a unified representation through a multimodal coding network combined with a fusion attention mechanism and a graph neural network.

[0054] Among them, drawing data is a visual representation of the physical structure and electrical connections of the power grid to be optimized, which may include: line route diagrams, substation general layout plans, and civil engineering structure diagrams, etc.; geographic information data is spatial geographic data related to the construction area of ​​the power grid to be optimized, covering topography, land attributes, environmental constraints, and infrastructure distribution, etc.; operational data is the actual operating parameters of the existing or similar power grids to be optimized, which is an important basis for evaluating the operating status of the power grid to be optimized and optimizing the design scheme, mainly including: historical load data, equipment operating parameters, etc.; civil engineering parameter data is the technical parameters and design requirements related to the civil engineering of the power grid; graph neural networks are neural network models specifically for processing data with graph structures, suitable for the correlation modeling of multi-source data of the power grid, and for mining deep spatial correlations and logical relationships between data; unified data standards are the process of standardizing multi-source data with inconsistent formats, units, and precision into a unified format, including format unification, unit unification, and precision unification; unified semantic expression is the process of standardizing information with the same meaning but different expressions from different sources into a unified semantic.

[0055] In this embodiment, a unified representation of multi-source power grid data is achieved through a multimodal coding network, providing high-quality data support for subsequent intelligent design. By acquiring core data such as drawing data, geographic information data, operational data, and civil engineering parameter data, design deviations caused by data loss are avoided, reducing later operation and maintenance adjustment costs. Furthermore, in terms of data collaborative utilization, attention mechanisms and graph neural networks are integrated, which not only solves the problems of heterogeneous formats and semantic differences of multi-source data, but also captures deep correlations between data, laying the foundation for intelligent power grid design.

[0056] In an exemplary embodiment, inputting target condition data for the power grid to be optimized into a preset power grid design model includes: inputting the access capacity, load density, terrain constraints, and investment ceiling, among other target condition data, into the preset power grid design model.

[0057] Among them, the access capacity is the total power supply capacity that the grid to be optimized can accept, as specified by the user; the load density is the electrical load per unit area within the power supply area of ​​the grid to be optimized; the terrain constraints are the design restrictions proposed by the user based on the terrain characteristics of the construction area of ​​the grid to be optimized, including terrain type and terrain risk; the investment limit is the maximum amount of funds set by the user for the design project of the grid to be optimized; the preset semantic understanding big model is an artificial intelligence model trained for power field design scenarios, which has the ability to understand the user's natural language input, extract professional information, and accurately identify professional expressions related to grid design; the target data are the core technical parameters that are output after semantic understanding and quantitative transformation and directly guide the grid design.

[0058] In this embodiment, the pre-defined semantic understanding model in the power field can transform users' non-standardized expressions into quantitative indicators, eliminating the need for repeated manual communication and calibration, shortening the demand transformation cycle, and improving design efficiency. At the same time, the generated target data can ensure that the final design scheme not only meets the user's personalized needs but also complies with technical requirements such as electrical safety and economy, achieving efficient connection between user needs and professional design.

[0059] Furthermore, in one embodiment, the results of the verification and evaluation are fed back to the power grid design model to optimize the power grid design model, including: generating multi-scenario test cases based on the results of the verification and evaluation; the multi-scenario test cases include power flow calculation, short-circuit analysis, and transient stability simulation; performing simulation tests on the multi-scenario test cases to obtain simulation results; comparing the simulation results with preset design standards, optimizing abnormal data and performance deviation data, and optimizing the power grid design model based on the optimized abnormal data and performance deviation data.

[0060] Among them, the verification and evaluation results are the test conclusions output after the automated verification of the power grid design scheme; the test set designed to comprehensively verify the feasibility of the power grid design scheme under different operating conditions through multi-scenario test cases; power flow calculation is one of the core scenarios of power grid simulation testing; short-circuit analysis is a test scenario that simulates the changes in electrical parameters when the power grid experiences a short-circuit fault, and evaluates the equipment tolerance and the reliability of the protection system; transient stability simulation is a simulation of the ability of the power grid to recover stable operation after being disturbed; simulation testing is the process of simulating the operation of the power grid design scheme using power simulation tools based on multi-scenario test cases; simulation results are the quantitative data and analysis conclusions output by the simulation test, including electrical parameter curves and performance index values ​​under each scenario.

[0061] More often, abnormal data refers to simulation data that exceeds the preset design standards or does not conform to the operating rules of the power grid; performance deviation data refers to simulation data that does not fully achieve the design optimization goals but does not exceed the safety threshold.

[0062] In this embodiment, by using core scenarios such as power flow calculation and short-circuit analysis, different operating conditions of the power grid can be simulated, avoiding the omission of potential risks in single-scenario testing and reducing the probability of actual operation failures. Furthermore, by comparing simulation results with preset design standards, abnormal data and performance deviations can be quickly identified, reducing rectification costs.

[0063] More specifically, in one embodiment, the abnormal data and performance deviation data are optimized by comparing the simulation results with preset design standards, including: comparing the simulation results with preset design standards, collecting abnormal data and performance deviation data of the design model corresponding to the power grid to be optimized; and correcting the weight parameters corresponding to the abnormal data and performance deviation data of the design model through reinforcement learning algorithms to optimize the design model.

[0064] The simulation results are quantitative data and analysis conclusions output after running multi-scenario test cases, such as power flow calculation, short-circuit analysis, and transient stability simulation, covering power grid electrical parameters, performance indicators, and scenario compliance. Abnormal data are data in the simulation results that exceed the preset design standard safety threshold and may lead to power grid operation risks. Performance deviation data are data in the simulation results that do not meet the design optimization target but do not exceed the safety threshold. The optimized design model is corrected through weight parameters.

[0065] In this embodiment, by comparing simulation results with design standards, abnormal data and performance deviation data are distinguished to avoid overlooking design defects, providing a clear direction for design model optimization and preventing resource waste caused by blind adjustments. Furthermore, the reinforcement learning algorithm can specifically correct the weight parameters corresponding to anomalies and deviations, ensuring the optimization efficiency of the design model. At the same time, the accuracy and compliance of the optimized design model generation scheme are greatly improved, which can not only reduce repeated iterations of subsequent simulation verification, but also provide better model support for similar power grid designs.

[0066] This application provides a power grid design model optimization method. To better understand the process of the above power grid design model optimization method, combined with... Figure 3 As shown below, the specific process of a power grid design model optimization method according to this application is described in detail, including the following steps:

[0067] Step S302: The multi-source data, including electrical, geographical, meteorological, and civil engineering data, required for the design of the power grid (power grid to be optimized) are fused and standardized to form a unified input that can be used for large model inference.

[0068] Step S304: The large model (the preset power grid design model) performs design scheme reasoning based on domain knowledge and historical engineering semantics, and introduces a constraint optimization algorithm in the generation stage to generate the output scheme (target data).

[0069] Step S306: Perform automated simulation and verification on the output scheme, and adaptively optimize the design model corresponding to the power grid to be optimized based on the feedback results of the automated simulation and verification.

[0070] The overall architecture of the system (the power grid to be optimized) comprises five layers: a data perception layer, an intelligent modeling layer, a design generation layer, a testing and verification layer, and an optimization feedback layer. These layers are interconnected via a data bus, enabling cross-domain integration of electrical and civil engineering design. In terms of workflow, the system first integrates and standardizes multi-source data (electrical, geographical, meteorological, civil engineering, etc.) required for power grid design, forming a unified input for large-scale model inference. Subsequently, the large model infers design schemes based on domain knowledge and historical engineering semantics, and introduces constrained optimization algorithms during the generation phase to ensure that the output scheme meets technical specifications, economic requirements, and safety requirements. After automated simulation and verification, the system adaptively optimizes the model based on feedback results, forming a continuous improvement mechanism.

[0071] Furthermore, the multimodal data fusion modeling module transforms multi-dimensional design inputs into a unified, computable representation, including electrical topology, geospatial information, load forecasting data, equipment characteristics, and civil design parameters. It can employ a multimodal coding network to integrate structured tabular data, graphical wiring layouts, and CAD (Computer Aided) data. The design (computer-aided design) files and natural language documents are jointly embedded in the semantic space of the large model. By integrating attention mechanisms and graph neural network structures, the implicit constraint relationships between different modalities are captured, achieving a unified representation of cross-domain information and providing a data foundation for subsequent automated design reasoning. The intelligent design generation module is used to automatically generate schemes based on a power industry-specific large model. The power industry-specific large model has built-in electrical specifications, equipment standards, construction constraints, and operational safety rules. It can reason based on user-input target conditions (such as access capacity, load density, terrain constraints, investment limits, etc.) to generate main wiring schemes, line routes, substation site selection, and basic structure design parameters. This module adopts a generative constraint optimization algorithm, embedding the objective function (safety, economy, constructability) into the generation process. This not only ensures that the design scheme is optimal under multiple indicators simultaneously, but also automatically adjusts the design strategy according to the context, realizing personalized and dynamic design.

[0072] Furthermore, the automated testing and verification module can perform full-scenario verification of the generated scheme. Verification content includes power flow calculation, short-circuit analysis, transient stability simulation, voltage qualification rate assessment, and civil engineering safety assessment. It can automatically generate simulation scripts based on the design scheme, call power system simulation platforms or structural analysis tools for batch testing, and evaluate whether various performance indicators meet design requirements through the result analysis module. In addition, the automated testing and verification module also introduces an evaluation mechanism based on rule-based reasoning and statistical learning to perform multi-dimensional comparison and anomaly detection of simulation results, achieving automated testing and intelligent verification. This enables real-time closed-loop processing between design and verification, significantly reducing the time required for testing. With less human intervention, the accuracy and consistency of results are improved. The feedback optimization and self-optimization modules work by feeding back test results and design deviations to the large model after design verification. Through reinforcement learning and knowledge distillation mechanisms, adaptive optimization is performed. The feedback optimization and self-optimization models automatically adjust weight parameters and design preferences based on simulation feedback, thereby improving the accuracy and rationality of subsequent generated solutions. In addition, based on the pre-introduced continuous learning mechanism, engineering cases and simulation data can be continuously accumulated during long-term operation to form a dynamically updated domain knowledge base, realizing the transformation from experience-driven to knowledge evolution-driven, and enabling the system to have the ability of "self-learning, self-correction, and self-optimization".

[0073] In addition, the entire system is based on a modular design. The modules work together through a unified data interface and a knowledge graph sharing mechanism. Users input design requirements through the interactive interface, and the system automatically calls the data perception and modeling module to generate input features. After large model reasoning and scheme generation, the testing module completes multi-dimensional verification, and the optimization module dynamically corrects the model. Finally, the system outputs the verified power grid design scheme and related engineering documents. Furthermore, the system supports multi-user collaborative design and multi-scenario parallel verification, and can adapt to various practical scenarios such as power transmission and transformation engineering, distributed energy planning, and power grid expansion.

[0074] Through the above embodiments, a closed-loop process of data acquisition, semantic modeling, scheme generation, testing and verification, and feedback optimization is realized through a multi-layer collaborative structure, with a large model as the core. This achieves intelligent and integrated power grid design and testing processes, significantly improving design efficiency and verification accuracy. Through large model inference, multi-modal data fusion, and automated simulation verification, collaborative optimization of electrical and civil engineering designs is achieved, reducing human intervention and design deviations, and enhancing the system's self-learning, self-optimization, and continuous evolution capabilities. This provides innovative technical support for the efficient planning and safe operation of smart grids.

[0075] Furthermore, through the above embodiments, an integrated method and system for power grid automation design and testing verification based on a large model was constructed. This system achieves a closed-loop fusion of power grid design, simulation verification, and model self-evolution. Through multimodal data fusion, large model semantic reasoning, and generative optimization techniques, it can automatically generate design schemes that comply with electrical codes, economic considerations, and construction constraints under limited input constraints. Simultaneously, the system embeds an automated simulation testing engine, achieving full-process automation of power flow analysis, stability verification, and civil engineering design safety assessment, significantly reducing manual intervention and improving verification accuracy and efficiency. Through reinforcement learning and knowledge distillation mechanisms, the model can adaptively adjust parameters based on test feedback, achieving self-learning and continuous optimization, giving the system dynamic evolution capabilities. Moreover, by integrating the knowledge systems of electrical design and civil engineering design within a unified large model semantic space, cross-disciplinary information barriers are broken down, enabling structural-electrical collaborative design. This provides an efficient, intelligent, and scalable new technical path for the rapid planning, accurate verification, and safe operation of future smart grids.

[0076] 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.

[0077] Based on the same inventive concept, this application also provides a power grid design model optimization device for implementing the power grid design model optimization method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more power grid design model optimization device embodiments provided below can be found in the limitations of the power grid design model optimization method described above, and will not be repeated here.

[0078] In one exemplary embodiment, such as Figure 4 As shown, a power grid design model optimization device is provided, comprising: a data generation module 401, a verification module 402, and a model optimization module 403, wherein:

[0079] The data generation module 401 is used to input the target condition data of the power grid to be optimized into the preset power grid design model to obtain the target data of the power grid design output by the power grid design model. The power grid design model processes the target condition data based on the built-in preset electrical specifications and preset constraint optimization algorithms. The target data includes the main wiring scheme, line route, substation site selection and basic structure design parameters of the power grid to be optimized.

[0080] The verification module 402 is used to call the automated test engine to verify and evaluate the initial optimized power grid based on the target data.

[0081] The model optimization module 403 is used to feed back the verification and evaluation results to the power grid design model in order to optimize the power grid design model.

[0082] Furthermore, in one embodiment, the data generation module 401 is also used to collect electrical topology data, geospatial data, load forecast data, equipment characteristic data, and civil design data of the power grid to be optimized; and to convert the electrical topology data, geospatial data, load forecast data, equipment characteristic data, and civil design data into a unified representation method through a multimodal coding network to obtain the target condition data of the power grid to be optimized.

[0083] Furthermore, in one embodiment, the data generation module 401 is also used to transform electrical topology data, geospatial data, load forecast data, equipment characteristic data, and civil design data into a unified representation method by combining a multimodal coding network with a fusion attention mechanism and a graph neural network.

[0084] Furthermore, in one embodiment, the data generation module 401 is also used to input the target condition data for the power grid to be optimized, including access capacity, load density, terrain constraints, and investment ceiling, into a preset power grid design model.

[0085] Furthermore, in one embodiment, the model optimization module 403 is also used to generate multi-scenario test cases based on the verification and evaluation results; the multi-scenario test cases include power flow calculation, short-circuit analysis, and transient stability simulation; the simulation test cases are performed to obtain simulation results; the simulation results are compared with the preset design standards, the abnormal data and performance deviation data are optimized, and the power grid design model is optimized based on the optimized abnormal data and performance deviation data.

[0086] Furthermore, in one embodiment, the model optimization module 403 is also used to compare the simulation results with the preset design standards, collect abnormal data and performance deviation data of the design model corresponding to the power grid to be optimized, and correct the weight parameters corresponding to the abnormal data and performance deviation data of the design model through reinforcement learning algorithms to optimize the design model.

[0087] Each module in the aforementioned power grid design model optimization 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 computer device's memory as software, so that the processor can call and execute the corresponding operations of each module.

[0088] 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, this 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 operating system and computer programs stored in the non-volatile storage media. The database stores power grid design model optimization data. The I / O interfaces are used for information exchange 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 power grid design model optimization method.

[0089] 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.

[0090] 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.

[0091] 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.

[0092] 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.

[0093] 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.

[0094] 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.

[0095] 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.

[0096] 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 power grid design model optimization method, characterized in that, The method includes: The target condition data for the power grid to be optimized is input into a preset power grid design model to obtain the target data for power grid design output by the power grid design model; the power grid design model processes the target condition data based on built-in preset electrical specifications and preset constraint optimization algorithms; the target data includes the main wiring scheme, line route, substation site selection and basic structure design parameters of the power grid to be optimized. The automated testing engine is invoked to verify and evaluate the initial optimized power grid based on the target data; The results of the verification and evaluation are fed back into the power grid design model to optimize the power grid design model.

2. The method according to claim 1, characterized in that, The method further includes: Collect electrical topology data, geospatial data, load forecast data, equipment characteristic data, and civil design data of the power grid to be optimized; The electrical topology data, geospatial data, load forecast data, equipment characteristic data, and civil engineering design data are transformed into a unified representation through a multimodal coding network to obtain the target condition data of the power grid to be optimized.

3. The method according to claim 2, characterized in that, The process of transforming the electrical topology data, geospatial data, load forecast data, equipment characteristic data, and civil engineering design data into a unified representation method through a multimodal coding network includes: By combining a multimodal coding network with a fusion attention mechanism and a graph neural network, the electrical topology data, the geospatial data, the load forecast data, the equipment characteristic data, and the civil engineering design data are transformed into a unified representation method.

4. The method according to claim 1, characterized in that, The step of inputting the target condition data for the power grid to be optimized into a preset power grid design model includes: The target conditions data for the power grid to be optimized, including access capacity, load density, terrain constraints, and investment limits, are input into the preset power grid design model.

5. The method according to claim 1, characterized in that, The results of the verification and evaluation are fed back to the power grid design model to optimize the power grid design model, including: Based on the results of the verification and evaluation, multi-scenario test cases are generated; the multi-scenario test cases include power flow calculation, short-circuit analysis, and transient stability simulation. Simulation tests were performed on the multi-scenario test cases to obtain simulation results; By comparing the simulation results with the preset design standards, the abnormal data and performance deviation data are optimized, and the power grid design model is optimized based on the optimized abnormal data and performance deviation data.

6. The method according to claim 5, characterized in that, The comparison of the simulation results with the preset design standards, and the optimization of abnormal data and performance deviation data, include: By comparing the simulation results with the preset design standards, abnormal data and performance deviation data of the design model corresponding to the power grid to be optimized are collected. The design model is optimized by correcting the weight parameters corresponding to abnormal data and performance deviation data of the design model through reinforcement learning algorithms.

7. A power grid design model optimization device, characterized in that, The device includes: The data generation module is used to input the target condition data of the power grid to be optimized into a preset power grid design model to obtain the target data of the power grid design output by the power grid design model; the power grid design model processes the target condition data based on the built-in preset electrical specifications and preset constraint optimization algorithms; the target data includes the main wiring scheme, line route, substation site selection and basic structure design parameters of the power grid to be optimized. The verification module is used to call the automated testing engine to verify and evaluate the initial optimized power grid based on the target data; The model optimization module is used to feed back the results of the verification and evaluation to the power grid design model in order to optimize the power grid design model.

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.

Images