A method and system for quantitative assessment of microgrid boundaries using multidimensional analysis and dynamic verification

Abstract

This invention relates to the field of power system planning, and more particularly to a method and system for quantitative assessment of microgrid boundaries using multidimensional analysis and dynamic verification. The method includes: acquiring distribution network and microgrid scheme data, performing scenario identification and decoupling modeling after standardization processing; extracting end-point supply guarantee scenario data and analyzing cost elements, generating a critical cost threshold table through normalization processing and threshold approximation algorithms; constructing a multidimensional boundary index model based on this table through multidimensional parameter correlation analysis and ensemble learning training; performing benefit matching calculations and green value accounting to generate an economic benefit decomposition structure; generating an economic boundary index set by combining real-time streaming data processing and stability verification; and finally generating a standardized assessment file through matrix mapping and dynamic weight adjustment. This invention achieves dynamic quantitative assessment of the economic boundary of microgrids, effectively integrating capacity substitution benefits and green value, and providing accurate basis for planning decisions.

Description

Technical Field

[0001] This invention relates to the field of power system planning, and in particular to a method and system for quantitative assessment of microgrid boundaries using multidimensional analysis and dynamic verification. Background Technology

[0002] In the field of power system planning, existing solutions related to a multi-dimensional analysis and dynamic verification method for quantitative assessment of microgrid boundaries typically rely on static cost-benefit models to evaluate microgrid economics. These solutions suffer from limitations such as high scenario coupling, coarse-grained benefit decomposition, and insufficient dynamic quantification capabilities. Existing methods often employ a unified boundary index framework to handle different application scenarios. In distribution-microgrid collaborative planning, confusion easily arises between critical thresholds for end-point supply assurance scenarios and capacity economic substitution scenarios, making it difficult to achieve stable and accurate economic boundary delineation. Regarding the joint processing of cost data and environmental benefit data, existing technologies generally lack scenario decoupling modeling mechanisms and multi-dimensional index correlation analysis capabilities. This makes it difficult to establish a consistent process of cost decomposition—benefit matching—dynamic comparison—index generation in microgrid economic boundary assessment scenarios. Consequently, the cost-benefit ratio calculation deviations between traditional distribution network transformation schemes and microgrid schemes accumulate, and boundary indexes become blurred. Summary of the Invention

[0003] This invention provides a multidimensional analysis and dynamic verification method and system for quantitative assessment of microgrid boundaries, to solve the problem of how to dynamically quantify cost-benefit in the economic boundary assessment scenario of microgrids based on the cost data to be decomposed and the environmental benefit data, through an improved Shannon entropy algorithm and the analytic hierarchy process.

[0004] To address the aforementioned technical problems, this invention provides a multi-dimensional analysis and dynamic verification method for quantitative assessment of microgrid boundaries, comprising:

[0005] Acquire data on power grid transformation schemes and microgrid schemes, perform format standardization processing, extract load characteristic parameters, identify scenario types, and analyze topology to generate a scenario decoupling modeling structure;

[0006] Data on end-point supply guarantee scenarios are extracted from the scenario decoupling modeling structure, and cost element analysis, line expansion cost separation, normalization processing, and threshold approximation algorithm are performed to generate a critical cost threshold table.

[0007] Obtain the critical cost threshold table, perform multidimensional parameter correlation analysis including principal component analysis, factor analysis and correlation coefficient matrix calculation, perform regression modeling processing including multiple linear regression or nonlinear regression, and train the model using ensemble learning methods including random forest, gradient boosting tree and multilayer perceptron network to generate a multidimensional boundary index model.

[0008] Based on the multidimensional boundary index model, the elements of equipment list, procurement contract and construction plan are decomposed and processed. The benefit matching calculation is performed to match the capacity substitution potential with the line expansion demand through the capacity substitution model. Green value accounting including environmental protection, carbon emission reduction and sustainable development contribution is implemented. The analytic hierarchy process and fuzzy comprehensive evaluation method are used to generate economic benefit decomposition structure.

[0009] Obtain the critical cost threshold table and economic benefit decomposition structure, perform real-time streaming data processing, proportional relationship analysis and stability verification, and generate a set of economic boundary indicators;

[0010] The system receives a multidimensional boundary indicator model and a set of economic boundary indicators, performs matrix mapping calculations, extracts key evaluation parameters, dynamically adjusts weights, and performs encryption encoding to generate standardized evaluation files.

[0011] Furthermore, the process of generating the scene decoupling modeling structure also includes:

[0012] The system acquires data on power distribution network upgrade plans and microgrid plans, maps fields using a predefined data dictionary and marks missing values, and performs format standardization processing, including unified field naming, data type conversion, timestamp synchronization, and unit conversion, to obtain standard plan data.

[0013] Load characteristic parameters, including load capacity, load curve shape, load volatility, load correlation and load response capability, are extracted from standard scheme data. Scene type identification is performed based on support vector machine, random forest and deep neural network. The scene type is determined according to spatial distribution and temporal dynamic characteristics, and scene classification labels are generated.

[0014] The topology analysis is performed by combining scene classification labels with distribution network node topology information. This includes constructing a node-line graph model and calculating connectivity, path length and network centrality indices based on graph theory algorithms. The topology graph is then divided into several independent subgraphs according to the scene classification labels to generate a scene decoupling modeling structure.

[0015] Furthermore, the data for the power distribution network upgrade plan includes:

[0016] The data for the power distribution network renovation plan specifically includes the power distribution network topology, line parameters such as line type, length, impedance, rated capacity, node load data including historical and predicted load curves, load characteristic parameters, existing electrical equipment configuration such as transformers, switchgear type, capacity and operating status, system operating constraints such as node voltage upper and lower limits, line transmission power limits, power supply reliability indicators, and planned renovation requirements such as identification of lines to be expanded, expected load growth data, and reliability improvement targets.

[0017] Furthermore, the microgrid solution data includes:

[0018] The microgrid solution data specifically includes the microgrid's internal topology, distributed power source configuration such as photovoltaic and wind turbine renewable energy types, rated capacity, output characteristic curves, energy storage system configuration such as battery type, capacity, charge and discharge efficiency, cycle life, load control parameters such as adjustable load capacity, response characteristics, priority settings, grid-connected / islanded operation control strategies, power constraints with the main grid, and economic parameters such as equipment investment costs, operation and maintenance costs, and policy subsidy information.

[0019] Furthermore, the process of generating the critical cost threshold table also includes:

[0020] Data on end-point supply guarantee scenarios, which are mainly aimed at ensuring continuous power supply to end-point loads, are extracted from the scenario decoupling modeling structure. Cost elements are analyzed, including extracting the capacity, type, construction period, unit price and installation cost of power generation equipment from the configuration of power generation nodes and energy storage nodes. Weighted average calculation is performed by calling the equipment cost database to obtain the power construction cost components.

[0021] The line expansion cost parameter, which includes line capacity expansion cost, equipment upgrade cost, construction and maintenance cost, is separated from the power construction cost component. Normalization processing, including minimum-maximum normalization and Z-score normalization, is performed, and the normalization method is automatically selected based on the data distribution characteristics to generate an adjustable load transformation sequence.

[0022] A threshold approximation algorithm based on gradient descent or heuristic search strategy is executed on the adjustable load modification sequence, including iterative optimization and cost critical point calculation. The threshold is dynamically adjusted in combination with scenario boundary conditions with cost minimization as the objective function, and a critical cost threshold table is generated.

[0023] Furthermore, the process of generating a multidimensional boundary index model also includes:

[0024] Obtain the critical cost threshold table, perform multidimensional parameter correlation analysis including principal component analysis, factor analysis and correlation coefficient matrix calculation, combine with the economic evaluation rules of the expert knowledge base, and use the entropy weight method to evaluate the rationality of the weight distribution to generate indicator weight coefficients.

[0025] Based on the index weight coefficients, regression modeling processing, including multiple linear regression or nonlinear regression, is performed. The functional relationship is fitted by the least squares method or gradient descent algorithm, and the reliability premium factor is extracted to build the model and generate the boundary dimension matrix.

[0026] The weight coefficients of the indicators are obtained, and the model is trained using an ensemble learning method including random forest, gradient boosting tree and multilayer perceptron network. Cross-validation is used to optimize the model parameters to prevent overfitting and generate a multidimensional boundary indicator model.

[0027] Furthermore, the process of generating the economic benefit decomposition structure also includes:

[0028] Based on the multidimensional boundary index model, the elements including the equipment list, procurement contract and construction plan are decomposed into equipment procurement cost, transportation cost, installation and commissioning cost and equipment depreciation cost. The cost components are adjusted according to the principle of matching load demand forecast with power generation capacity to obtain the power generation construction cost components.

[0029] Based on the power construction cost component, a benefit matching calculation is performed to match the capacity substitution potential with the line expansion demand through a capacity substitution model. The impact of microgrid capacity substitution is evaluated based on the capacity substitution coefficient matrix, and a capacity substitution benefit value is generated.

[0030] To obtain the capacity substitution benefit value, implement green value accounting including contributions to environmental protection, carbon emission reduction and sustainable development, adopt the analytic hierarchy process and fuzzy comprehensive evaluation method, and form a three-dimensional weighting system of capacity substitution benefit, reliability premium benefit and green value benefit to generate an economic benefit decomposition structure.

[0031] Furthermore, the process of generating the set of economic boundary indicators also includes:

[0032] The system obtains the critical cost threshold table and economic benefit decomposition structure, performs real-time streaming data processing including dynamic operation data accessed from the distribution network operation monitoring system and microgrid dispatching system, performs noise filtering, anomaly detection and missing value imputation, and uses rule engine and event-driven mechanism to filter key economic boundary events and generate dynamic calculation queues.

[0033] Based on the cost threshold parameters and economic benefit indicators in the dynamic calculation queue, the proportional relationship analysis is performed, including statistical regression methods and dynamic weighting algorithms. The sliding window technique is used to achieve continuous updates, and covariance analysis is used to identify abnormal ratios and generate a real-time ratio sequence.

[0034] The system acquires real-time ratio sequences, performs stability checks including variance analysis, trend detection, and autocorrelation function verification, and handles anomalies by isolating abnormal data, diagnosing the causes of anomalies, and correcting data. It also uses interpolation, regression prediction, and historical data substitution techniques to correct abnormal data and generate a set of economic boundary indicators.

[0035] Furthermore, the process of generating standardized assessment files also includes:

[0036] The system receives a multidimensional boundary indicator model and a set of economic boundary indicators, performs matrix mapping calculations including multiplication of weight matrix and indicator vector matrix, and dynamically considers the correlation between indicators according to preset mapping rules and weight adjustment mechanism to generate an initial decision matrix.

[0037] Key evaluation parameters, including capacity substitution benefit weight, reliability premium weight, green value contribution and cost threshold sensitivity, are extracted from the initial decision matrix. Dynamic weight adjustment is performed using genetic algorithms and particle swarm optimization. The weight distribution is iteratively updated based on multi-objective optimization algorithms and adaptive weight adjustment strategies to generate the basis for optimal decision-making.

[0038] The selection criteria are encrypted using advanced encryption standards and RSA algorithms, and standardized processing is performed, including unified naming of data fields, structured storage format, version control, and metadata appending. A hash algorithm is used to generate a digital digest to ensure the integrity of the archive, thus generating a standardized evaluation archive.

[0039] Furthermore, a multi-dimensional analysis and dynamic verification microgrid boundary quantitative assessment system, applied to any one of the methods described above, includes:

[0040] Install a calibration module to acquire and verify the installation parameters of microgrid equipment;

[0041] The data preprocessing module is used to perform real-time running data filtering and historical load data fitting;

[0042] The time alignment module is used to perform timestamp synchronization processing on multi-source data;

[0043] The arbitration decision module is used to generate optimized control instructions based on a dynamic threshold table;

[0044] The instruction execution module is used to drive the energy storage converter to complete power adjustment;

[0045] The strategy update module is used to record operation logs and update the economic threshold table.

[0046] The key innovations of this invention include:

[0047] (1) Through the scenario decoupling modeling mechanism, the critical cost threshold is quantified for the end-of-line supply guarantee scenario, which involves the joint processing of cost data to be decomposed and environmental benefit data.

[0048] (2) Construct a multi-dimensional boundary indicator model for capacity economy substitution scenarios, involving green value accounting of benefits to be integrated.

[0049] (3) Innovate the economic benefit decomposition algorithm, and perform benefit matching calculation for the separate line expansion cost parameters in the cost to be decomposed, and generate capacity substitution benefit value.

[0050] The following are its main beneficial effects:

[0051] (1) Through the scenario decoupling modeling mechanism, it is possible to effectively distinguish between end-of-line supply guarantee scenarios and other scenarios, ensuring that the quantification results of critical cost thresholds are accurately applied in different scenarios and avoiding confusion of critical thresholds in common problems.

[0052] (2) The construction of the multidimensional boundary indicator model enables the generation of economic benefit decomposition structure through green value accounting in capacity economy substitution scenarios, ensuring the accurate generation of economic boundary indicators.

[0053] (3) Through innovative economic benefit decomposition algorithm, capacity substitution benefit value can be generated, ensuring the accuracy of dynamic calculation of benefit-cost ratio and solving the problem of accumulated deviation in benefit-cost ratio calculation in traditional methods. Attached Figure Description

[0054] Figure 1 A flowchart illustrating a multidimensional analysis and dynamic verification method for quantitative evaluation of microgrid boundaries provided in this application embodiment;

[0055] Figure 2 This is a structural block diagram of a microgrid boundary quantitative assessment system for multidimensional analysis and dynamic verification, provided in an embodiment of this application. Detailed Implementation

[0056] Example 1: Refer to Figure 1 This is a flowchart illustrating a multi-dimensional analysis and dynamic verification method for quantitative assessment of microgrid boundaries provided by an embodiment of the present invention. The process may include at least steps S100-S600:

[0057] S100: From the data of the power distribution network transformation plan and the microgrid plan, perform format standardization processing, load characteristic parameter extraction, scene type identification and topology analysis to generate a scene decoupling modeling structure;

[0058] S200: Extract end-point supply guarantee scenario data from the scenario decoupling modeling structure, perform cost element analysis, line expansion cost separation, normalization processing and threshold approximation algorithm to generate critical cost threshold table;

[0059] S300: Obtain the critical cost threshold table, perform multidimensional parameter correlation analysis including principal component analysis, factor analysis and correlation coefficient matrix calculation, perform regression modeling processing including multiple linear regression or nonlinear regression, and train using ensemble learning methods including random forest, gradient boosting tree and multilayer perceptron network to generate a multidimensional boundary index model.

[0060] S400, based on the multidimensional boundary index model, performs element decomposition processing including the analysis of equipment list, procurement contract and construction plan, performs benefit matching calculation by matching capacity substitution potential with line expansion demand through capacity substitution model, and implements green value accounting including environmental protection, carbon emission reduction and sustainable development contribution. It uses the analytic hierarchy process and fuzzy comprehensive evaluation method to generate economic benefit decomposition structure.

[0061] S500: Obtain the critical cost threshold table and economic benefit decomposition structure, perform real-time streaming data processing, proportional relationship analysis and stability verification, and generate a set of economic boundary indicators;

[0062] S600 receives a multidimensional boundary indicator model and a set of economic boundary indicators, performs matrix mapping calculations, extracts key evaluation parameters, dynamically adjusts weights, and performs encryption encoding to generate standardized evaluation files.

[0063] Step S100 includes at least steps S110-S130:

[0064] S110. Obtain data on power grid renovation schemes and microgrid schemes, perform format standardization processing, and obtain standard scheme data.

[0065] Specifically, the data for the distribution network upgrade plan and the data for the microgrid plan are respectively sourced from the distribution network planning system and the microgrid design system. Both contain multi-dimensional information such as grid structure parameters, load distribution information, equipment configuration details, and operational constraints. Specifically, the distribution network upgrade plan data includes the distribution network topology, line parameters such as line type, length, impedance, and rated capacity, node load data including historical and predicted load curves, load characteristic parameters, existing electrical equipment configuration such as transformer and switchgear type, capacity, and operating status, system operating constraints such as node voltage upper and lower limits, line transmission power limits, power supply reliability indicators, and planned upgrade requirements such as identification of lines to be expanded, expected load growth data, and reliability improvement targets.

[0066] The microgrid scheme data specifically includes the microgrid's internal topology, distributed power source configuration such as photovoltaic and wind turbine renewable energy types, rated capacity, output characteristic curves, energy storage system configuration such as battery type, capacity, charge and discharge efficiency, cycle life, load control parameters such as adjustable load capacity, response characteristics, priority settings, grid-connected / islanded operation control strategies, power constraints with the main grid, and economic parameters such as equipment investment costs, operation and maintenance costs, and policy subsidy information.

[0067] First, the system receives data from the power distribution network upgrade plan and microgrid plan through the interface module. Addressing the format differences between the data sources, it performs format standardization processing. This standardization includes, but is not limited to, unified field naming, data type conversion, timestamp synchronization, and unit conversion, ensuring the data meets the requirements of a unified data model. Specifically, a predefined data dictionary is used to map each field, missing values ​​are marked for missing fields, and abnormal data is recorded for subsequent review. The format standardization processing is automatically triggered by a rule engine, combined with a verification module to check data integrity and consistency. When format discrepancies or data anomalies are detected, an error report is automatically generated, and a preset anomaly handling process is executed, such as requesting re-upload or manual confirmation. Further, the standard plan data undergoes a data cleaning module to remove redundant information and merge duplicate records, ensuring data uniqueness and validity. Data cleaning also includes outlier identification and correction, adjusting outlier data to a reasonable range based on statistical analysis and comparison with historical data. After the standard plan data is constructed, the system automatically generates a data index and version control information for easy traceability and management. Finally, the standard scheme data, as the output field of this step, is passed to the "data to be decoupled" input in the subsequent step S120 for use in scenario type identification and decoupling modeling. The standard scheme data includes distribution network node information, line parameters, load characteristics, microgrid generation side and energy storage device configuration, etc., to meet the needs of subsequent multi-dimensional scenario analysis, ensuring closed-loop data links and no information loss.

[0068] S120. Extract load characteristic parameters from standard scheme data, identify scene types, and generate scene classification labels;

[0069] Specifically, for the standard scheme data, the system first filters load-related parameters through a feature extraction module. These load characteristic parameters include load capacity, load curve shape, load volatility, load correlation, and load response capability. Load capacity refers to the maximum electricity demand of each distribution network node. The load curve shape is extracted using time series analysis technology to identify typical load curve features. Load volatility is calculated using statistical methods to determine the magnitude of load changes. Load correlation reflects the correlation coefficient between loads at different nodes, and load response capability assesses the speed and magnitude of the load's response to adjustment signals. The system employs a multi-dimensional feature fusion algorithm to normalize the above load characteristic parameters before inputting them into the scene recognition model. The scene recognition model is based on machine learning classification algorithms, combining multiple algorithms such as Support Vector Machine (SVM), Random Forest, and Deep Neural Network (DNN). Through iterative optimization of model parameters using training and validation sets, it achieves accurate differentiation between end-point supply guarantee scenarios and capacity economic substitution scenarios. Specifically, based on the spatial distribution and temporal dynamic characteristics of the load characteristic parameters, the system determines the scenario type of each node and its associated lines and generates corresponding scenario classification labels. The scene classification labels include scene type identifiers, scene weights, a list of associated nodes, and classification confidence levels. The system further performs consistency checks on the scene classification labels, comparing them with a historical scene database to ensure their rationality and stability. Abnormal or low-confidence labels are reviewed and corrected by a manual review module. The scene classification labels, as output fields of this step, are passed to the "to-be-structured data" input in subsequent step S130 for topology analysis and scene decoupling modeling. These scene classification labels provide fundamental data support for subsequent critical cost threshold quantification and multi-dimensional boundary indicator modeling, enabling precise scene-level segmentation and management.

[0070] S130. Perform topological analysis on scene classification labels to generate scene decoupling modeling structure;

[0071] Specifically, the system receives the scenario classification labels and, combined with the distribution network node topology information, performs topology analysis. The topology analysis module first constructs a node-line graph model of the distribution network and microgrid. Nodes include load nodes, generation nodes, and energy storage nodes, and lines represent the power connections between nodes. Based on graph theory algorithms, the system calculates the connectivity, path length, and network centrality indices between nodes, identifying key nodes and bottleneck lines. Further, according to the scenario classification labels, the system divides the distribution network topology into several subgraphs, each corresponding to an independent scenario, thus completing scenario decoupling. Specifically, the scenario decoupling modeling structure is defined as a structured data set containing scenario identifiers, subgraph node sets, subgraph line sets, scenario boundary conditions, and scenario operational constraints. The system uses a rule engine to perform boundary condition identification on each subgraph, clarifying the power exchange interfaces and constraints between scenarios, ensuring the accuracy and completeness of scenario decoupling. The scenario decoupling modeling structure also includes detailed descriptions of load distribution, generation configuration, and energy storage status within the scenario, supporting subsequent cost threshold quantification and indicator modeling. The system performs integrity verification on the scene decoupling modeling structure, confirming the unique attribution of nodes and lines and the absence of overlapping conflicts between scenes. Abnormal situations are automatically triggered to initiate a correction process. This scene decoupling modeling structure, as an output field of this step, is passed to the "Scene Classification Data" input of the subsequent step S210 for use in quantifying critical cost thresholds. This structure provides a data foundation for cross-step connections, ensuring that S200 and subsequent modules can conduct quantitative assessments of economic boundaries based on accurate scene classification.

[0072] Step S200 includes at least steps S210-S230:

[0073] S210. Extract end-point supply guarantee scenario data from the scenario decoupling modeling structure, perform cost element analysis, and obtain the power construction cost components.

[0074] Specifically, the system receives the scenario decoupling modeling structure from step S130 as input. This structure includes multiple scenario identifiers and their corresponding node sets, line sets, boundary conditions, and operational constraints. For the scenario decoupling modeling structure, the system first identifies end-point supply guarantee scenario data through a scenario filtering module. End-point supply guarantee scenarios are defined as scenarios in the distribution network whose primary objective is to ensure continuous power supply to end-point loads. Their characteristics include high reliability requirements for load nodes, low capacity substitution potential, and clearly defined power supply boundaries. Specifically, based on scenario identifiers and load characteristic parameters, combined with scenario weights and a list of associated nodes, the system filters out subgraph structures that meet the characteristics of end-point supply guarantee scenarios, forming an end-point supply guarantee scenario data set.

[0075] Furthermore, the system performs cost element analysis processing on the data from the end-point supply guarantee scenario. This analysis process first extracts power construction-related parameters from the configuration of power generation and energy storage nodes within the scenario, specifically including power generation equipment capacity, type, construction period, equipment unit price, and installation costs. The system calls the equipment cost database through an interface module, combining equipment technical parameters and market price information to calculate the power construction cost components. During the calculation process, the system considers the economies of scale and regional differences, using a weighted average method to comprehensively evaluate the construction costs of different equipment types. In addition, the system performs a matching analysis of the capacity configuration of power generation nodes with load demand, adjusting the power construction cost components to reflect the rationality of the actual power supply demand and the construction scale.

[0076] During cost element analysis, the system employs a rule engine to detect abnormal data. Issues such as abnormal equipment unit prices or unreasonable capacity configurations automatically trigger an early warning mechanism and record anomaly logs for subsequent manual review. The power supply construction cost component serves as the output field for this step and is passed to the input of the subsequent step S220 for the separation and normalization of line expansion cost parameters. This output field contains a detailed cost composition of power supply construction in end-point supply guarantee scenarios, supporting precise calculation of subsequent critical cost threshold quantification and forming a cross-step data transmission link.

[0077] S220. Separate the line expansion cost parameters from the power construction cost component, perform normalization processing, and generate an adjustable load transformation sequence.

[0078] Specifically, the system receives the power construction cost component output in step S210 as input. This set includes the power construction cost component and other relevant cost parameters. The system further subdivides the power construction cost component through a cost classification module, focusing on separating the line expansion cost parameters. These line expansion cost parameters include line capacity increase costs, equipment upgrade costs, construction and maintenance costs, etc., reflecting the economic investment in increasing line capacity to meet load growth or renovation needs in the distribution network. Based on the line type, length, current carrying capacity, and existing equipment status, the system extracts the corresponding expansion unit price and construction costs from the cost database, and calculates the line expansion cost in conjunction with specific line parameters.

[0079] Furthermore, the system performs normalization processing on the extracted line expansion cost parameters. The normalization module uses a standardization algorithm to convert different line expansion costs into dimensionless indicators according to a unified metric, eliminating data biases caused by equipment type, regional differences, and economies of scale. The normalization process includes two methods: min-max normalization and Z-score normalization. The system automatically selects the appropriate method based on the data distribution characteristics to ensure the stability and representativeness of the normalization results. The normalized line expansion cost parameters are integrated into an adjustable load modification sequence, which reflects the changing trend of line expansion costs under different load adjustment schemes and their impact on the overall economic efficiency of the modification scheme.

[0080] To ensure data integrity and accuracy, the system monitors abnormal fluctuations in the input data in real time during the normalization process, marking and correcting outliers. It also performs trend regression using historical cost data to improve the reliability of the normalization results. The adjustable load modification sequence, as an output field of this step, is passed to the "data to be quantified" input of the subsequent step S230, allowing the threshold approximation algorithm to generate the critical cost threshold table. This output field provides crucial load modification cost sequence data for critical cost threshold quantification, and it is integrated throughout the S200 module and its interactions with subsequent modules.

[0081] S230. Perform a threshold approximation algorithm on the adjustable load modification sequence to generate a critical cost threshold table;

[0082] Specifically, the system receives the data to be quantized from step S220 as input. This data mainly consists of a normalized adjustable load modification sequence. Based on this data, the system calls the threshold approximation algorithm module to perform quantization calculations of the critical cost threshold. The threshold approximation algorithm relies on an iterative optimization mechanism. By setting initial threshold parameters and combining the load modification cost sequence with the power supply construction cost component, it calculates the cost critical point under different modification schemes. Specifically, the algorithm uses cost minimization as the objective function, and dynamically adjusts the threshold parameters based on the scene boundary conditions and operational constraints in the scene decoupling modeling structure to approximate the optimal critical cost level.

[0083] During algorithm execution, the system first preprocesses the input data, including data smoothing, outlier removal, and time series alignment, to improve computational stability. Then, based on gradient descent or a heuristic search strategy, the system iteratively updates the critical cost threshold, gradually narrowing the threshold range until convergence conditions are met or the preset number of iterations is reached. In each iteration, the algorithm evaluates the economic boundary performance under the current threshold, compares cost input with benefit output, and adjusts the threshold to optimize the balance point. The system records key parameters and calculation results in real time during the iteration process, forming a complete computational trajectory for subsequent analysis and verification.

[0084] Furthermore, the system performs consistency checks on the threshold approximation results, combining historical critical cost data and expert experience rules to determine the reasonableness of the threshold. Abnormal or significantly deviating thresholds trigger an automatic adjustment mechanism, re-executing the approximation process to ensure the accuracy and applicability of the output results. The critical cost threshold table, as an output field of this step, is passed to the "threshold input parameter" input terminal of the subsequent step S310 for multi-dimensional boundary indicator modeling. This output field systematically reflects the cost critical point under end-point supply guarantee scenarios, constructing the basic data for quantitative assessment of economic boundaries and providing crucial support for cross-step economic benefit analysis.

[0085] Step S300 includes at least steps S310-S330:

[0086] S310. Obtain the critical cost threshold table, perform multi-dimensional parameter correlation analysis, and generate indicator weight coefficients.

[0087] Specifically, the system receives a critical cost threshold table from step S230 as input. This table contains cost critical point data for end-point supply guarantee scenarios. Combined with the capacity economic substitution scenario identifier and its associated node information from the scenario decoupling modeling structure output in step S130, a capacity economic substitution scenario feature vector is formed. This feature vector encompasses multi-dimensional data such as load capacity, load variation trends, power generation and energy storage configuration parameters, historical operating data, and economic indicators, reflecting capacity substitution potential and economic characteristics. The system automatically integrates the aforementioned multi-source heterogeneous data through a data acquisition module, performs data preprocessing including missing value imputation, outlier removal, and time synchronization, forming a structured feature vector set.

[0088] Furthermore, the system performs multi-dimensional parameter correlation analysis on the feature vector of the capacity economic substitution scenario. This analysis process is driven by the association rule mining module and employs statistical methods such as Principal Component Analysis (PCA), Factor Analysis, and correlation coefficient matrix calculation to identify the intrinsic relationships and influence strength among the feature parameters. Specifically, the system calculates the correlation coefficient matrix of the feature parameters, filters out key parameters that significantly affect the capacity economic substitution effect, eliminates redundant or low-correlation parameters, and optimizes the feature space dimension. Simultaneously, the system combines economic evaluation rules from the expert knowledge base to assign preliminary weights to the key parameters, forming a weight vector.

[0089] Subsequently, the system employs a weighted comprehensive algorithm to combine statistical analysis results with expert weighting to generate indicator weight coefficients. These coefficients reflect the contribution of each characteristic parameter to the economic boundary of the capacity economic substitution scenario, specifically including the weight distribution of indicators such as capacity substitution potential, load response capability, generation cost sensitivity, and energy storage regulation efficiency. The system performs consistency checks on the weight coefficients, using the entropy weight method to evaluate the rationality and stability of the weight distribution. Abnormal weights are adjusted by the rule engine. The weight coefficients are stored in vector or matrix form, containing parameter identifiers, weight values, and confidence information.

[0090] The indicator weight coefficients, as output fields of this step, are passed to the "model construction parameters" input field of the subsequent step S320 for use in constructing the multidimensional boundary indicator model. This output field systematically expresses the relative importance of each economic parameter in the capacity economy substitution scenario, supports cross-step data flow and model optimization, and forms a continuous data link from cost threshold to indicator weights.

[0091] S320. Based on the index weight coefficients, perform regression modeling to generate the boundary dimension matrix;

[0092] Specifically, the system receives the index weighting coefficients output in step S310 as input, and combines them with the capacity economic substitution scenario data from the scenario decoupling modeling structure in step S130 to extract the reliability premium factor. The reliability premium factor is defined as reflecting the economic added value generated by the power system in the process of capacity substitution due to improved power supply reliability, encompassing aspects such as reduced power outage risk, increased reserve capacity, and enhanced system stability. The system calculates reliability indicators based on historical fault data, load fluctuation characteristics, and equipment maintenance records through the reliability assessment module, thereby quantifying the reliability premium factor.

[0093] Furthermore, the system integrates the indicator weight coefficients and reliability premium factors to form a set of model building parameters. This set includes key economic indicators for capacity substitution and economic compensation parameters for reliability, serving as the basic input for constructing the multidimensional boundary indicator model. The system performs multiple linear regression or nonlinear regression modeling on the model building parameters. The regression model fits the functional relationship between economic benefits and each indicator using the least squares method or gradient descent algorithm, quantifying the combined impact of indicator weights and reliability premiums on the economic boundary.

[0094] In the regression modeling process, the system first normalizes the input parameters to eliminate the influence of different units and magnitudes, improving model training performance. Then, the system divides the sample dataset into training and validation sets and uses cross-validation to optimize model parameters and prevent overfitting. After model training is complete, the system evaluates the goodness of fit (R²), mean squared error, and parameter significance levels of the regression model to ensure its statistical validity.

[0095] Based on the regression model output, the system generates a boundary dimension matrix. This matrix, centered on the economic boundary, includes the weights of capacity substitution indicators, reliability premium factors, and their interaction coefficients, forming a multi-dimensional quantitative structure. The rows and columns of the matrix correspond to different economic parameters and reliability indicators, respectively, and the element values ​​reflect the weighted relationships and influence strengths between the parameters. The system stores the boundary dimension matrix in a structured database, supporting subsequent machine learning training and model optimization.

[0096] The boundary dimension matrix, as the output field of this step, is passed to the input of the "model to be optimized" in the subsequent step S330 for further training and improvement of the multi-dimensional boundary index model. This output field realizes the integrated expression of economic weight and reliability value, constructs the core data structure of the economic boundary of capacity economic substitution scenario, and promotes cross-step data closure and model iteration.

[0097] S330. Obtain the indicator weight coefficients, select an ensemble learning method for training, and generate a multidimensional boundary indicator model.

[0098] Specifically, the system receives the boundary dimension matrix output in step S320 as input, and combines it with the capacity economic substitution scenario decoupling modeling structure provided in step S130 and the critical cost threshold table from step S230 to form a complete training dataset. The training dataset includes multi-dimensional economic parameters, reliability premium factors, and corresponding economic boundary labels, reflecting the economic boundary performance under the capacity substitution scenario. The system uses a data preprocessing module to perform data cleaning, outlier detection, and feature engineering to improve data quality and model training effectiveness.

[0099] Furthermore, the system trains the model using supervised machine learning algorithms based on the aforementioned training dataset. Specifically, the system selects ensemble learning methods, such as Random Forest and Gradient Boosting Trees, and deep learning methods, such as Multi-Layer Perceptron (MLP) networks, to model the economic boundary. During training, the system adjusts the model parameters through iterative optimization algorithms to minimize prediction errors and improve the model's generalization ability.

[0100] During training, the system performs model performance evaluation using metrics including Root Mean Square Error (RMSE), Coefficient of Determination (R²), and Mean Absolute Error (MAE), and conducts multiple rounds of cross-validation. The system incorporates an early stopping strategy to prevent overfitting and ensure the model's predictive accuracy on unseen data. After training, the system uses feature importance analysis to identify the parameters that have the greatest impact on the economic boundary, thus aiding in model interpretability.

[0101] Furthermore, the system performs version management and model compression on the trained models to generate multidimensional boundary index models suitable for online real-time computation. This model can take multidimensional parameters from capacity-economic substitution scenarios as input and output corresponding economic boundary indices, providing accurate support for subsequent economic benefit decomposition and dynamic calculation. The system stores the trained multidimensional boundary index models in a model library, supporting retrieval and updates.

[0102] The multidimensional boundary indicator model, as an output field of this step, is passed to the "Indicator Model Input" input terminal of the subsequent step S410 for economic benefit decomposition processing. This output field realizes the mapping from multidimensional economic parameters to economic boundary indicators, constructs the core evaluation model of capacity economic substitution scenario, and promotes the closed-loop connection between cross-step data and models.

[0103] Step S400 includes at least steps S410-S430:

[0104] S410. Based on the multidimensional boundary index model, the elements are decomposed to obtain the power construction cost components.

[0105] Specifically, the system receives the multi-dimensional boundary index model output from step S330 and traditional scheme cost data from the external distribution network planning system and microgrid design system as input. The traditional scheme cost data includes power generation construction costs, line expansion costs, adjustable load modification costs, and other related expenses, covering multiple cost components such as equipment procurement, installation, commissioning, acceptance, and maintenance. The system first obtains the traditional scheme cost data through the data interface module and, combined with predefined cost classification standards, initially categorizes the overall cost data according to power generation construction, line expansion, and load modification. Specifically, the system uses a rule engine to match and map the input data fields, identifying power generation construction-related fields, including generator capacity, type, unit price, construction period, and related installation costs, ensuring that the data structure conforms to the unified cost model requirements.

[0106] Furthermore, the system performs element decomposition processing on the power construction-related fields. This processing involves parsing the equipment list, procurement contracts, and construction plans item by item to break down the power construction cost into multiple detailed components, such as equipment procurement costs, transportation costs, installation and commissioning costs, and equipment depreciation costs. The system combines equipment technical parameters with a market price database and uses a weighted allocation method to reasonably distribute each component, reflecting the cost contribution of different equipment types and construction stages. During the decomposition process, the system automatically detects abnormal cost data, including abnormal unit prices, quantity discrepancies, and mismatched time points, triggering abnormal log recordings and early warning notifications for manual review. This element decomposition processing, through multiple iterations and data verification, forms a structured set of power construction cost component data.

[0107] Furthermore, the system combines the operational constraints and planning objectives of traditional solutions to determine the rationality of power generation construction cost components. Specifically, based on the principle of matching load demand forecasting with generation capacity, the system adjusts the power generation construction cost components, eliminating redundant or duplicate calculations to ensure the accuracy and completeness of cost data. The power generation construction cost component data includes equipment category, capacity scale, unit price, and decomposed cost details, forming a standardized cost data structure. The power generation construction cost component, as an output field of this step, is passed to the "cost to be decomposed" input terminal of the subsequent step S420 for capacity substitution benefit calculation, supporting cross-step data transfer and economic benefit analysis.

[0108] S420: Based on the power construction cost component, perform benefit matching calculations to generate capacity substitution benefit values;

[0109] Specifically, the system receives the power construction cost component and other traditional solution cost data output from step S410 as input, focusing on separating and processing the line expansion cost parameters. These line expansion cost parameters include line capacity expansion costs, equipment upgrade costs, construction and maintenance costs, and related indirect costs, specifically covering information such as line length, conductor type, current carrying capacity, and existing equipment status. Through a cost classification module and in conjunction with the line asset management database, the system automatically identifies and extracts line expansion-related cost fields, forming a set of line expansion cost parameters.

[0110] Furthermore, the system performs benefit-matching calculations based on the aforementioned line expansion cost parameters. This calculation process first uses a capacity substitution model to match the capacity substitution potential of the microgrid scheme with the line expansion requirements of the traditional scheme. Based on the capacity substitution coefficient matrix and a multi-dimensional boundary index model, the system evaluates the impact of microgrid capacity substitution on reducing line expansion requirements and quantifies the cost savings brought about by capacity substitution. Specifically, the system proportionally maps line expansion cost parameters to capacity substitution potential, calculates the economic benefit value corresponding to capacity substitution, and reflects the contribution of the microgrid scheme in reducing line expansion investment.

[0111] During the benefit matching calculation process, the system dynamically adjusts the capacity substitution benefit value by combining operational data and historical load curves, taking into account the impact of load fluctuations and reserve capacity demand on the benefit. The system also performs consistency checks on the calculation results, using an expert rule base and historical cases to identify abnormal or significantly deviating benefit values, triggering an automatic correction mechanism to ensure the accuracy and rationality of the benefit calculation. After calculation, the capacity substitution benefit value is stored in structured data format, containing detailed information such as the benefit amount, influencing factors, and calculation basis.

[0112] The capacity substitution benefit value, as an output field of this step, is passed to the "Benefits to be Integrated" input of the subsequent step S430 for green value accounting and the generation of the economic benefit decomposition structure. This output field effectively connects the costs of traditional solutions with the benefits of microgrids, constructs a key intermediate product for economic benefit decomposition, and supports quantitative assessment of economic boundaries across steps.

[0113] S430. Obtain capacity substitution benefit value, implement green value accounting, and generate economic benefit decomposition structure;

[0114] Specifically, the system receives the capacity substitution benefit value output in step S420 as input, and combines it with microgrid solution benefit data and relevant environmental indicators to conduct green value accounting. This green value accounting aims to quantify the economic contribution of microgrid solutions to environmental protection, carbon emission reduction, and sustainable development. Indicators include carbon emission reduction, increased renewable energy utilization, energy conservation and emission reduction benefits, and socio-environmental benefits. The system acquires relevant environmental parameters and policy incentive information through an environmental data acquisition module to form the basic dataset for green value accounting.

[0115] Furthermore, based on the green value accounting model, and combining capacity substitution benefit value and reliability premium factor, the system performs multi-dimensional economic benefit decomposition. This decomposition process employs the Analytic Hierarchy Process (AHP) and fuzzy comprehensive evaluation method to refine the economic benefit structure of the microgrid scheme, dividing it into three main parts: capacity substitution benefit, reliability premium benefit, and green value benefit. Through a weight allocation mechanism, combined with expert experience and historical data, the system determines the relative contribution of each benefit component, forming a multi-dimensional weighted system for the economic benefit decomposition structure.

[0116] During the economic benefit decomposition process, the system performs consistency checks and anomaly detection on the input benefit data, and uses statistical analysis and trend regression methods to correct abnormal data, ensuring the stability and scientific validity of the decomposition results. The system stores the decomposed economic benefit data in a structured format, including the specific values, weight distribution, and calculation basis of each benefit component, supporting subsequent dynamic boundary indicator generation and decision matrix output. The economic benefit decomposition structure serves as the output field of this step and is passed to the "Benefit Input Parameters" input terminal of the subsequent step S510, running through the economic benefit analysis and dynamic calculation stages.

[0117] In another embodiment, the system receives the cost data to be decomposed output in step S410, including power source construction cost components, line expansion cost parameters, and adjustable load modification costs. The cost data to be decomposed originates from equipment procurement records, construction contracts, and operation and maintenance databases in traditional distribution network planning schemes. The system separates the line expansion cost parameters through a cost classification module, defined by formula ①:

[0118]

[0119] in:

[0120] This is an index for the cost parameters of line expansion, with values ​​ranging from 1 to... The cost data is derived from the list of line expansion expenses in the cost data to be broken down.

[0121] The total number of entries for line expansion cost parameters;

[0122] For the first The original distribution probability of the line expansion cost parameter in the traditional scheme is calculated by extracting the "line expansion cost ratio" from the "cost data to be decomposed";

[0123] For the first The corrected distribution probability of the line expansion cost parameter in the microgrid scheme is obtained by extracting the microgrid capacity substitution weight from the capacity substitution coefficient matrix;

[0124] It serves as a separation index for line expansion cost parameters, reflecting the cost distribution differences between traditional and microgrid solutions;

[0125] Calculated by formula ① As a threshold trigger condition, when At that time, the system determines that the cost parameters of line expansion can be effectively separated. Furthermore, the system simulates the benefit distribution under different load adjustment schemes, and formula ② is defined as:

[0126]

[0127] in:

[0128] This is the index for the number of Monte Carlo simulation iterations, with values ​​ranging from 1 to... ( The number of simulations is preset by the system.

[0129] Total number of Monte Carlo simulations;

[0130] For the first The weights of the load adjustment schemes in this simulation are calculated by extracting the "scheme priority" from the "adjustable load modification sequence";

[0131] For the traditional solution in the first The line expansion cost in this simulation is calculated by extracting the product of the line expansion unit price and the line length from the cost data to be decomposed.

[0132] For microgrid solutions in the first The equivalent expansion cost in this simulation is calculated by multiplying the capacity substitution coefficient and the line expansion cost extracted from the capacity substitution benefit model.

[0133] This represents the expected value of capacity substitution benefits, reflecting the average benefit of microgrid solutions compared to traditional solutions;

[0134] Formula ① As the basis for determining the separation degree, the simulation conditions driving Formula ② are used; Formula ② As the core indicator of the output field "capacity substitution benefit value" in this step, it is passed to the input of step S430.

[0135] Furthermore, the system's receiving capacity substitution benefit value Based on the environmental benefit data of the microgrid scheme, green value accounting is performed. The environmental benefit data comes from the carbon emission records of the microgrid design system, renewable energy output curves, and policy subsidy databases. The system uses the Analytic Hierarchy Process (AHP) to construct a green value weight matrix, defined by formula ③:

[0136]

[0137] in:

[0138] For the first Normalized weights for each dimension;

[0139] This is an index for the dimensions of green value evaluation, with values ​​ranging from 1 to... ( (corresponding to carbon emissions, renewable energy utilization rate, and social and environmental benefits), which comes from the definition of the green value accounting model;

[0140] To determine the index of a matrix element, the value ranges from 1 to... ( (corresponding to a 3×3 judgment matrix), derived from expert scoring data;

[0141] This represents the total number of elements in the matrix;

[0142] For the first The first dimension The elements of the judgment matrix are calculated by extracting the expert scoring table from the environmental benefit data.

[0143] This is a circular index variable, with a value range of 1 to... ;

[0144] This represents the total number of dimensions for green value evaluation.

[0145] This is the chain multiplication operator;

[0146] Furthermore, the system will use formula ③ With capacity substitution benefit value The integration generates a comprehensive economic benefit indicator, which is defined in formula ④ as follows:

[0147]

[0148] in:

[0149] For the first The adjustment coefficients for capacity substitution benefits under each dimension are calculated by extracting the carbon emission trading price from the policy incentive database.

[0150] For the first The adjustment coefficients for green value under each dimension are calculated by extracting the "unit power generation subsidy" from the "renewable energy subsidy standard".

[0151] For the first The green value quantification values ​​in each dimension are calculated by extracting carbon emission reduction, renewable energy share, and social satisfaction index from the environmental benefit data.

[0152] As a comprehensive economic benefit indicator;

[0153] Formula ③ As a weight, it is input into formula ④, formula ④ As an intermediate quantity passed to the next step, the output field "Economic Benefit Decomposition Structure" contains... and its component parameters.

[0154] Furthermore, the system is based on formula ④. The system combines dynamic load fluctuation data to perform stability checks on the economic benefit decomposition structure. The dynamic load data originates from the load curves of the distribution network real-time monitoring system and the charging and discharging records of microgrid energy storage. The system models the benefit fluctuations, and formula ⑤ is defined as:

[0155]

[0156] in:

[0157] This is a time variable, derived from the timestamps of real-time load data;

[0158] It is a partial differential operator;

[0159] The benefit propagation speed vector is calculated from the 'load volatility' and 'energy storage response rate'.

[0160] For divergence operators;

[0161] The benefit diffusion coefficient is calculated by extracting the benefit fluctuation variance from the historical economic data.

[0162] For the Laplace operator;

[0163] The solution result of formula ⑤ Used to evaluate the time stability of the economic benefit decomposition structure, where... These are time-related economic benefit indicators. Finally, the system generates a structured decomposition result, and the output field "Economic Benefit Decomposition Structure" includes capacity substitution benefits, green value benefits, and stability indicators, which are then passed to the "Benefit Input Parameters" input terminal in step S510.

[0164] The technical effect of this section is as follows: By separating the cost parameters of line expansion, combining the analytic hierarchy process (AHP) to quantify the benefits, and using fluid dynamics equations to verify stability, an economic benefit decomposition structure covering capacity substitution, green value, and dynamic fluctuations is formed, realizing a closed-loop mapping of cost benefits and cross-step data connection.

[0165] Step 500 includes at least steps S510-S530:

[0166] S510. Obtain the critical cost threshold table and economic benefit decomposition structure, perform real-time streaming data processing, and generate a dynamic calculation queue.

[0167] Specifically, the system receives the critical cost threshold table from step S230 and the economic benefit decomposition structure from step S430 as input. The critical cost threshold table contains cost threshold information for end-point supply guarantee scenarios, and the economic benefit decomposition structure covers multi-dimensional economic indicators such as capacity substitution benefits, reliability premium benefits, and green value benefits. The system first uses a data integration module to unify the format and map fields of the input data, forming a unified data format that meets the requirements of real-time stream processing. The data integration module employs a timestamp-based synchronization mechanism to coordinate the time series of different data sources, ensuring the temporal consistency and integrity of the input data.

[0168] Furthermore, the system integrates dynamic operational data from the distribution network operation monitoring system and microgrid dispatching system via a real-time streaming data processing module. This data includes load changes, power generation, energy storage status, and environmental parameters. The real-time streaming data undergoes noise filtering, anomaly detection, and missing value imputation via a preprocessing module to improve data quality. The system combines a critical cost threshold table with an economic benefit decomposition structure, utilizing a rule engine and event-driven mechanism to filter key economic boundary events that match the current operational status, forming a dynamic calculation queue. This queue is arranged in a time-series format and includes cost thresholds, benefit indicators, and related parameters to be calculated, supporting real-time response capabilities for subsequent dynamic calculations.

[0169] During the dynamic calculation queue generation process, the system monitors abnormal fluctuations in the input data, uses sliding window analysis and statistical control chart methods to identify abnormal patterns, and automatically triggers data correction processes and log recording. The dynamic calculation queue includes event identifiers, timestamps, threshold parameters, economic benefit indicators, and operational status information, constructing the basic data structure for dynamic boundary indicator calculation. As an output field of this step, the dynamic calculation queue is passed to the "data to be calculated" input in step S520 for use in extracting benefit-cost elements and analyzing proportional relationships. The dynamic calculation queue achieves dynamic integration of critical cost thresholds and economic benefit decomposition, spanning the entire process of real-time data stream and economic boundary indicator generation, supporting cross-step data closure and real-time updates.

[0170] S520: Based on the cost threshold parameters and economic benefit indicators in the dynamic calculation queue, perform proportional relationship analysis and generate a real-time ratio sequence;

[0171] Specifically, the system receives the data to be calculated output from step S510 as input. This data includes cost threshold parameters and economic benefit indicators from the dynamic calculation queue. The system uses a feature extraction module to extract key benefit cost elements from the economic benefit components and cost composition of the dynamic data stream. These benefit cost elements include, but are not limited to, power generation construction cost components, line expansion cost parameters, capacity substitution benefit values, reliability premiums, and green value benefits, covering the main influencing factors of the microgrid's economic boundary. The extraction process employs multidimensional data analysis technology, combined with time series decomposition and feature selection algorithms, to filter redundant information and highlight representative indicators.

[0172] Furthermore, the system utilizes a proportional relationship analysis module to construct a dynamic proportional model between economic benefits and cost inputs based on extracted benefit-cost elements. This proportional relationship analysis employs statistical regression methods and a dynamic weighted algorithm to calculate the benefit-cost ratio, reflecting the economic boundary performance of the microgrid scheme under different operating conditions. The system dynamically adjusts the ratio calculation weights by combining historical and real-time operating data, responding to factors such as load fluctuations, equipment status changes, and market price fluctuations. The proportional relationship analysis process utilizes sliding window technology for continuous updates, ensuring the timeliness and accuracy of the ratio sequence.

[0173] During the analysis, the system performs consistency checks on the input data, employs covariance analysis and anomaly detection algorithms to identify abnormal ratios, and triggers a data correction mechanism. The system stores the calculated real-time ratio sequence in time-series format, including ratio timestamps, ratio values, and related economic indicators. This real-time ratio sequence, as the output field of this step, is passed to the "Indicators to be Generated" input in step S530 for subsequent stability verification and generation of the economic boundary indicator set. This output field achieves a continuous expression of the dynamic mapping between economic benefits and costs, supporting dynamic assessment and real-time decision-making regarding the economic boundary.

[0174] S530. Obtain the real-time ratio sequence, perform stability verification, and generate a set of economic boundary indicators.

[0175] Specifically, the system receives the real-time ratio sequence output in step S520 as input, which reflects the dynamic proportional relationship between the economic benefits and costs of the microgrid. The system first performs statistical stability analysis on the real-time ratio sequence through a stability verification module, employing multiple methods such as analysis of variance, trend detection, and autocorrelation function to evaluate the fluctuation range, trend consistency, and periodicity of the ratio sequence. This stability verification process, combined with set threshold standards, determines the stability of the economic boundary indicators and identifies abnormal fluctuations or abrupt changes.

[0176] Furthermore, the system initiates anomaly handling mechanisms for detected abnormal data. These mechanisms include abnormal data isolation, anomaly cause diagnosis, and data correction. Abnormal data isolation is achieved through a sliding time window, marking and temporarily removing abnormal data segments. Anomaly cause diagnosis combines multi-dimensional information such as equipment operating status, load changes, and market environment for comprehensive analysis to identify possible factors contributing to the anomaly. Data correction employs techniques such as interpolation, regression prediction, and historical data substitution to adjust the abnormal data and restore the continuity and reliability of the sequence. The entire anomaly handling process is tracked in detail through log recording and event management modules, forming a complete anomaly handling archive.

[0177] After completing stability verification and anomaly handling, the system inputs the corrected real-time ratio sequence into the economic boundary index generation module. This module, based on a multi-index fusion algorithm, combines multi-dimensional economic indicators such as capacity substitution benefits, reliability premium, and green value to comprehensively construct an economic boundary index set. This index set is represented in vector or matrix form, containing the numerical values, weights, and confidence information of multiple economic boundary indicators, reflecting the comprehensive economic performance of the microgrid solution. The system further performs consistency and integrity checks on the index set to ensure reasonable logical relationships between indicators and standardized data structures.

[0178] The economic boundary index set, as the output field of this step, is passed to the "Indicator Input Parameters" input terminal of step S610 for the mapping calculation and optimization of the decision matrix. This output field realizes a closed-loop expression of the dynamic calculation of economic benefits and costs, constructs a multi-dimensional evaluation basis for the economic boundary of the microgrid, and supports cross-step data flow and decision support functions.

[0179] Step S600 includes at least steps S610-S630:

[0180] S610: Receive the multidimensional boundary indicator model and the economic boundary indicator set, perform matrix mapping calculation, and generate the initial decision matrix;

[0181] Specifically, the system first receives the multidimensional boundary index model output from step S330 and the economic boundary index set output from step S530 as inputs. The multidimensional boundary index model includes the mapping relationship of multidimensional economic parameters and their weight distribution for capacity economic substitution scenarios, while the economic boundary index set reflects the dynamic ratio of economic benefits to costs of microgrid schemes under different operating conditions. The system uses a data interface module to uniformly format both into structured data, ensuring consistency in field names, data types, and timestamps, forming an input data set available for subsequent mapping calculations.

[0182] Furthermore, the system, based on a matrix mapping calculation module, performs the mapping and fusion of a multidimensional boundary indicator model and a set of economic boundary indicators. Specifically, the system performs matrix multiplication operations on the weight matrix defined in the multidimensional boundary indicator model and the indicator vectors in the set of economic boundary indicators. The mapping process encompasses the weighted summation of indicator weights and the interactions between different economic indicators. This mapping calculation dynamically considers the correlation and complementarity between indicators according to preset mapping rules and weight adjustment mechanisms, forming an initial decision matrix. During the mapping process, the system triggers an anomaly detection module to record abnormal events and perform data imputation or weight adjustment in case of missing, abnormal, or inconsistent input data, ensuring the completeness and accuracy of the mapping results.

[0183] The mapping calculation module also incorporates scene boundary conditions from the scene decoupling modeling structure to constrain the initial decision matrix, eliminating decision elements that do not conform to scene operation constraints and enhancing the matrix's scene adaptability. The system uses a log management module to record in detail the calculation steps, weight adjustments, and anomaly handling results of the mapping process, forming a complete mapping calculation trajectory. The initial decision matrix is ​​stored in a matrix data structure, containing the mapping relationship between each scene and the economic boundary indicators, along with their corresponding weight values. This matrix serves as the output field for this step and is passed to the input of step S620 for dynamic weight adjustment and the generation of optimal decision-making criteria, ensuring the continuity and consistency of data flow across steps.

[0184] S620. Extract key evaluation parameters from the initial decision matrix, perform dynamic weight adjustment, and generate optimal decision-making basis;

[0185] Specifically, the system receives the initial decision matrix output in step S610 as input. This initial decision matrix includes the weight distribution of multi-dimensional economic boundary indicators and scenario mappings, as well as initial decision values. The system first uses a parameter extraction module to filter out key evaluation parameters. These key evaluation parameters are defined as indicators that significantly impact the microgrid's economic boundary decisions, including capacity substitution benefit weights, reliability premium weights, green value contribution, and cost threshold sensitivity. Parameter extraction is based on statistical analysis and an expert rule base, dynamically determining parameter priorities and weight ranges by combining the historical performance of the indicators with the current operating environment.

[0186] Furthermore, the system utilizes a dynamic weight adjustment module to perform dynamic weight optimization on the key evaluation parameters. This module combines real-time operational data, market price fluctuations, and load change information, employing a multi-objective optimization algorithm and an adaptive weight adjustment strategy to adjust the weights of each indicator to reflect the actual impact of the current economic boundary. Specifically, the system iteratively updates the weight distribution based on intelligent optimization methods such as genetic algorithms and particle swarm optimization, balancing economic benefits and cost inputs to improve the adaptability and accuracy of the decision matrix. During the weight adjustment process, the system incorporates scene characteristics and constraints from the scene decoupling modeling structure to ensure that the adjustment results conform to the scene's operational logic and technical specifications.

[0187] The system monitors and logs the weight adjustment process, capturing the adjustment trajectory, parameter changes, and algorithm convergence, supporting subsequent auditing and verification. The dynamic weight adjustment results are stored in structured data, including the adjusted weight values, the optimized objective function value, and confidence indices. Based on the adjusted weights, the system calculates and generates optimal decision-making criteria, specifically a comprehensive evaluation result and ranking for different scenarios and economic boundary indicators. These optimal decision-making criteria are passed as output fields to the "Results to be Stored" input in step S630 for subsequent encryption encoding and standardized evaluation file generation, achieving a closed-loop decision support across steps.

[0188] S630. Encrypt and encode the basis for the optimal decision-making process to generate a standardized evaluation file;

[0189] Specifically, the system receives the optimal decision criteria output from step S620 as input. These criteria include a comprehensive evaluation of the microgrid's economic boundary and a ranking of optimal schemes. The system first performs encryption encoding on the optimal decision criteria through a data encryption module. This encryption encoding employs a combination of symmetric and asymmetric encryption algorithms, specifically including Advanced Encryption Standard (AES) and RSA algorithms, to ensure data confidentiality and integrity. The encryption process dynamically allocates keys through a key management system, supporting key updates, revocations, and backups to meet security management requirements.

[0190] Furthermore, the system uses a standardization module to convert the encrypted decision-making criteria into evaluation files that conform to a unified format specification. The standardization process encompasses standardized naming of data fields, structured storage formats (such as XML, JSON, or dedicated database formats), version control, and metadata appending, ensuring the compatibility and traceability of the evaluation files. Combining evaluation templates and specifications, the system automatically generates complete files including a summary of the decision-making criteria, scenario descriptions, economic indicators, and weight adjustment records, supporting subsequent queries, analysis, and report generation.

[0191] In addition, the system performs integrity verification of the evaluation files, using a hash algorithm (such as SHA-256) to generate a digital digest, ensuring that the file content has not been tampered with. The verification result, along with the encrypted information, is stored in a secure database, forming a standardized evaluation file. The system uploads the standardized evaluation file to the system database through an interface module and triggers a storage confirmation mechanism, recording the storage timestamp, storage location, and access permission information. The standardized evaluation file serves as an output field of this step, used by the system database for subsequent microgrid optimization decision support, achieving closed-loop management and security assurance for the quantitative assessment of economic boundaries.

[0192] Example 2: Figure 2 A structural block diagram of a microgrid boundary quantitative assessment system based on multidimensional analysis and dynamic verification according to an embodiment of the present invention is shown. Figure 2 As shown, the structure may include:

[0193] The installation calibration module 01 is used to acquire and verify the installation parameters of microgrid equipment. Specifically, it receives equipment installation data from the microgrid design system and combines it with structural parameters provided by the distribution network planning system to complete the acquisition and verification of equipment installation parameters. The installation calibration module connects to the equipment database through a standardized interface to obtain the equipment's technical specifications and installation requirements. During the verification process, the system uses a rule engine to detect parameter consistency, identify abnormal parameters, and generate a verification report. Verified installation parameters are recorded as a set of equipment installation parameters and transmitted to the data preprocessing module as input, while verification logs are retained for subsequent traceability.

[0194] The data preprocessing module 02 is used to perform real-time operational data filtering and historical load data fitting. Specifically, it receives the equipment installation parameter set from the installation calibration module and the operational data from the real-time monitoring system, and performs data filtering and load fitting processing. The system uses filtering algorithms to suppress noise in real-time data, remove outliers, and ensure data stability. Combined with historical load data, time series analysis techniques are used for fitting to form a load prediction model. The load prediction model is passed as load characteristic data to the time series alignment module for invocation, and timestamp information is registered in the data buffer for subsequent modules to read.

[0195] The time-series alignment module 03 is used to perform timestamp synchronization processing on multi-source data. Specifically, it receives load characteristic data from the data preprocessing module, combines it with topology information from the distribution network planning system, and performs timestamp synchronization processing. The system uses a timestamp alignment algorithm to synchronize and adjust the multi-source data, ensuring the time-series consistency of each data source. After the synchronization conditions are met, a time-synchronized dataset is generated, which is then submitted to the arbitration and judgment module as input, and the correspondence between the dataset and the synchronization strategy is recorded in the data repository.

[0196] Arbitration decision module 04 is used to generate optimized control instructions based on a dynamic threshold table. Specifically, it generates control instructions based on the time-synchronized dataset from the time-series alignment module, combined with the dynamic threshold table. The system analyzes the dataset through a decision rule base, identifies the current operating state and economic boundary conditions, and generates optimized control instructions. The control instructions are output to the instruction execution module and the effective status is returned to the policy update module for registration.

[0197] The instruction execution module 05 is used to drive the energy storage converter to complete power adjustment. Specifically, it receives optimized control instructions from the arbitration and determination module and completes the power adjustment of the energy storage converter. The system sends the control instructions to the energy storage converter through the communication interface and monitors the converter's response status in real time. After the adjustment is completed, it generates adjustment confirmation information and sends it back to the strategy update module to ensure the effective execution of the control instructions.

[0198] The strategy update module 06 is used to record operation logs and update the economic threshold table. Specifically, it receives adjustment confirmation information from the instruction execution module and the effective status from the arbitration judgment module, and performs log recording and threshold table updates. The system stores the operation logs in the log system, maintaining an index relationship consistent with the time series. Combined with the adjustment confirmation information, the economic threshold table is updated to form the latest threshold configuration. The updated threshold table is returned to the installation and calibration module for parameter backfeeding, completing the refresh and status update of the data link.

Claims

1. A method for quantitative assessment of microgrid boundaries using multidimensional analysis and dynamic verification, characterized in that, include: Acquire data on power grid transformation schemes and microgrid schemes, perform format standardization processing, extract load characteristic parameters, identify scenario types, and analyze topology to generate a scenario decoupling modeling structure; Data on end-point supply guarantee scenarios is extracted from the scenario decoupling modeling structure. Cost element analysis, line expansion cost separation, normalization processing, and threshold approximation algorithms are then performed to generate a critical cost threshold table. The process of generating the critical cost threshold table also includes: Data on end-point supply guarantee scenarios, which are mainly aimed at ensuring continuous power supply to end-point loads, are extracted from the scenario decoupling modeling structure. Cost elements are analyzed, including extracting the capacity, type, construction period, unit price and installation cost of power generation equipment from the configuration of power generation nodes and energy storage nodes. Weighted average calculation is performed by calling the equipment cost database to obtain the power construction cost components. The line expansion cost parameter, which includes line capacity expansion cost, equipment upgrade cost, construction and maintenance cost, is separated from the power construction cost component. Normalization processing, including minimum-maximum normalization and Z-score normalization, is performed, and the normalization method is automatically selected based on the data distribution characteristics to generate an adjustable load transformation sequence. A threshold approximation algorithm based on gradient descent or heuristic search strategy is executed on the adjustable load modification sequence, including iterative optimization and cost critical point calculation. The threshold is dynamically adjusted in combination with scenario boundary conditions with cost minimization as the objective function to generate a critical cost threshold table. Obtain the critical cost threshold table, perform multidimensional parameter correlation analysis including principal component analysis, factor analysis and correlation coefficient matrix calculation, perform regression modeling processing including multiple linear regression or nonlinear regression, and train the model using ensemble learning methods including random forest, gradient boosting tree and multilayer perceptron network to generate a multidimensional boundary index model. Based on the multidimensional boundary index model, the elements of equipment list, procurement contract and construction plan are decomposed and processed. The benefit matching calculation is performed to match the capacity substitution potential with the line expansion demand through the capacity substitution model. Green value accounting including environmental protection, carbon emission reduction and sustainable development contribution is implemented. The analytic hierarchy process and fuzzy comprehensive evaluation method are used to generate economic benefit decomposition structure. Obtain the critical cost threshold table and economic benefit decomposition structure, perform real-time streaming data processing, proportional relationship analysis and stability verification, and generate a set of economic boundary indicators; The system receives a multidimensional boundary indicator model and a set of economic boundary indicators, performs matrix mapping calculations, extracts key evaluation parameters, dynamically adjusts weights, and performs encryption encoding to generate standardized evaluation files.

2. The method according to claim 1, characterized in that, The process of generating a scene decoupled modeling structure also includes: The system acquires data on power distribution network upgrade plans and microgrid plans, maps fields using a predefined data dictionary and marks missing values, and performs format standardization processing, including unified field naming, data type conversion, timestamp synchronization, and unit conversion, to obtain standard plan data. Load characteristic parameters, including load capacity, load curve shape, load volatility, load correlation and load response capability, are extracted from standard scheme data. Scene type identification is performed based on support vector machine, random forest and deep neural network. The scene type is determined according to spatial distribution and temporal dynamic characteristics, and scene classification labels are generated. The topology analysis is performed by combining scene classification labels with distribution network node topology information. This includes constructing a node-line graph model and calculating connectivity, path length and network centrality indices based on graph theory algorithms. The topology graph is then divided into several independent subgraphs according to the scene classification labels to generate a scene decoupling modeling structure.

3. The method according to claim 1 or 2, characterized in that, Data for the power distribution network upgrade plan includes: The data for the power distribution network renovation plan specifically includes the power distribution network topology, line parameters such as line type, length, impedance, rated capacity, node load data including historical and predicted load curves, load characteristic parameters, existing electrical equipment configuration such as transformers, switchgear type, capacity and operating status, system operating constraints such as node voltage upper and lower limits, line transmission power limits, power supply reliability indicators, and planned renovation requirements such as identification of lines to be expanded, expected load growth data, and reliability improvement targets.

4. The method according to claim 1 or 2, characterized in that, Microgrid solution data includes: The microgrid solution data specifically includes the microgrid's internal topology, distributed power source configuration such as photovoltaic and wind turbine renewable energy types, rated capacity, output characteristic curves, energy storage system configuration such as battery type, capacity, charge and discharge efficiency, cycle life, load control parameters such as adjustable load capacity, response characteristics, priority settings, grid-connected / islanded operation control strategies, power constraints with the main grid, and economic parameters such as equipment investment costs, operation and maintenance costs, and policy subsidy information.

5. The method according to claim 1, characterized in that, The process of generating a multidimensional boundary index model also includes: Obtain the critical cost threshold table, perform multidimensional parameter correlation analysis including principal component analysis, factor analysis and correlation coefficient matrix calculation, combine with the economic evaluation rules of the expert knowledge base, and use the entropy weight method to evaluate the rationality of the weight distribution to generate indicator weight coefficients. Based on the index weight coefficients, regression modeling processing, including multiple linear regression or nonlinear regression, is performed. The functional relationship is fitted by the least squares method or gradient descent algorithm, and the reliability premium factor is extracted to build the model and generate the boundary dimension matrix. The weight coefficients of the indicators are obtained, and the model is trained using an ensemble learning method including random forest, gradient boosting tree and multilayer perceptron network. Cross-validation is used to optimize the model parameters to prevent overfitting and generate a multidimensional boundary indicator model.

6. The method according to claim 1, characterized in that, The process of generating the economic benefit decomposition structure also includes: Based on the multidimensional boundary index model, the elements including the equipment list, procurement contract and construction plan are decomposed into equipment procurement cost, transportation cost, installation and commissioning cost and equipment depreciation cost. The cost components are adjusted according to the principle of matching load demand forecast with power generation capacity to obtain the power generation construction cost components. Based on the power construction cost component, a benefit matching calculation is performed to match the capacity substitution potential with the line expansion demand through a capacity substitution model. The impact of microgrid capacity substitution is evaluated based on the capacity substitution coefficient matrix, and a capacity substitution benefit value is generated. To obtain the capacity substitution benefit value, implement green value accounting including contributions to environmental protection, carbon emission reduction and sustainable development, adopt the analytic hierarchy process and fuzzy comprehensive evaluation method, and form a three-dimensional weighting system of capacity substitution benefit, reliability premium benefit and green value benefit to generate an economic benefit decomposition structure.

7. The method according to claim 1, characterized in that, The process of generating the economic boundary indicator set also includes: The system obtains the critical cost threshold table and economic benefit decomposition structure, performs real-time streaming data processing including dynamic operation data accessed from the distribution network operation monitoring system and microgrid dispatching system, performs noise filtering, anomaly detection and missing value imputation, and uses rule engine and event-driven mechanism to filter key economic boundary events and generate dynamic calculation queues. Based on the cost threshold parameters and economic benefit indicators in the dynamic calculation queue, the proportional relationship analysis is performed, including statistical regression methods and dynamic weighting algorithms. The sliding window technique is used to achieve continuous updates, and covariance analysis is used to identify abnormal ratios and generate a real-time ratio sequence. The system acquires real-time ratio sequences, performs stability checks including variance analysis, trend detection, and autocorrelation function verification, and handles anomalies by isolating abnormal data, diagnosing the causes of anomalies, and correcting data. It also uses interpolation, regression prediction, and historical data substitution techniques to correct abnormal data and generate a set of economic boundary indicators.

8. The method according to claim 1, characterized in that, The process of generating standardized assessment files also includes: The system receives a multidimensional boundary indicator model and a set of economic boundary indicators, performs matrix mapping calculations including multiplication of weight matrix and indicator vector matrix, and dynamically considers the correlation between indicators according to preset mapping rules and weight adjustment mechanism to generate an initial decision matrix. Key evaluation parameters, including capacity substitution benefit weight, reliability premium weight, green value contribution and cost threshold sensitivity, are extracted from the initial decision matrix. Dynamic weight adjustment is performed using genetic algorithms and particle swarm optimization. The weight distribution is iteratively updated based on multi-objective optimization algorithms and adaptive weight adjustment strategies to generate the basis for optimal decision-making. The selection criteria are encrypted using advanced encryption standards and RSA algorithms, and standardized processing is performed, including unified naming of data fields, structured storage format, version control, and metadata appending. A hash algorithm is used to generate a digital digest to ensure the integrity of the archive, thus generating a standardized evaluation archive.

9. A multidimensional analysis and dynamic verification microgrid boundary quantitative assessment system, applied to the method of any one of claims 1-8, characterized in that, include: Install a calibration module to acquire and verify the installation parameters of microgrid equipment; The data preprocessing module is used to perform real-time running data filtering and historical load data fitting; The time alignment module is used to perform timestamp synchronization processing on multi-source data; The arbitration decision module is used to generate optimized control instructions based on a dynamic threshold table; The instruction execution module is used to drive the energy storage converter to complete power adjustment; The strategy update module is used to record operation logs and update the economic threshold table.

Images